THE COMMUNITY RESOURCE
DIRECTORY
Small decisions.
Small decisions.
Endless possibilities.
Find what people build with Jev. Explore tools, libraries, experiments and ideas from across the TypeSafe ecosystem.
11,066unique resources
29clear categories
220focused subcategories
Yourssave a personal reading list
YOUR NEXT BUILD STARTS HERE
Download catalog
Explore the collection
Browse groups
Agents & orchestration1,475 of 1,475 matches
Model & task routing139 of 139 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| A3M Router — npmwww.npmjs.com | — | Adaptive multi-model LLM router across 80+ providers with pheromone-trail failover and ensemble merging, where model=jev-auto lets Jev route each request to a suitable provider in a single pass. | Agents & orchestrationModel & task routing | |
| Agent router testwww.reddit.com | — | Uses Jev as the router for a personal knowledge app's agents, correctly picking the recipe or scraper agent from an Instagram reel transcript in 145-271 ms. | Agents & orchestrationModel & task routing | |
| Agent with a Jev model routerx.com | — | Agent build in which Jev acts as the model router, choosing which model handles each request. | Agents & orchestrationModel & task routing | |
| agent-routergithub.com | 100 | CLI that filters your coding-agent subscriptions by quota rules, asks Jev to rank the eligible agents and models and pick a reasoning effort, then launches Cursor, Claude Code, Codex, or OpenCode in a Herdr pane. | Agents & orchestrationModel & task routing | |
| ai_sdkgithub.com | 2 | jcpsimmons/jev-model-router-demo — Throwaway Jev demo: route coding tasks to Grok Build or Codex Astra · ai_sdk · JavaScript | Agents & orchestrationModel & task routing | |
| Ailerixgithub.com | 2 | OpenRouter-style model router with a single ailerix/auto slug: Jev classifies each request into a typed catalog route, then a cost-per-task Pareto chain picks the provider. | Agents & orchestrationModel & task routing | |
| Ailerix — appailerix.vercel.app | — | OpenRouter-style model router with a single ailerix/auto slug: Jev classifies each request into a typed catalog route, then a cost-per-task Pareto chain picks the provider. | Agents & orchestrationModel & task routing | |
| claude-routergithub.com | 0 | Claude Code plugin that auto-picks the right model for each turn — micro, low, medium, or high tier — using Jev routing. | Agents & orchestrationModel & task routing | |
| codex-jevgithub.com | 0 | Route each Codex prompt to the right effort with TypeSafe's Jev | Agents & orchestrationModel & task routing | |
| codex-subagent-routergithub.com | 0 | AI model routing for Codex subagents. Let Jev choose the model and reasoning effort for each mission, using your own task profiles. | Agents & orchestrationModel & task routing | |
| Corent model routingx.com | — | Corent's router uses Jev to decide what each request needs across 1000+ models, how confident that call is, and whether to take the route or fall back, with stricter confidence for costly jobs like a $2 video clip. | Agents & orchestrationModel & task routing | |
| Corent model routing — appcorent.tech | — | Corent's router uses Jev to decide what each request needs across 1000+ models, how confident that call is, and whether to take the route or fall back, with stricter confidence for costly jobs like a $2 video clip. | Agents & orchestrationModel & task routing | |
| Criteria-based model routing in evex.com | — | Experimental autoModel option in the eve agent framework that uses Jev to route each request between models described by plain-language criteria. | Agents & orchestrationModel & task routing | |
| Crucible system_one routegithub.com | — | Goal-directed research loop engine with an example domain where a decision role routes tickets through any POST /v1/systemone server, either hosted Jev or a vLLM DiffusionGemma server. | Agents & orchestrationModel & task routing | |
| Crucible system_one route — appneuralmagic.github.io | — | Goal-directed research loop engine with an example domain where a decision role routes tickets through any POST /v1/systemone server, either hosted Jev or a vLLM DiffusionGemma server. | Agents & orchestrationModel & task routing | |
| Crucible system_one route — repogithub.com | — | Goal-directed research loop engine with an example domain where a decision role routes tickets through any POST /v1/systemone server, either hosted Jev or a vLLM DiffusionGemma server. | Agents & orchestrationModel & task routing | |
| ddfeyes/jev-modegithub.com | 3 | I kept watching coding agents burn context on decisions that aren't hard - triage 400 tickets, tag 600 files, route to one of six teams. jev-mode moves those verdicts to a typed-judgment model. I A/B'd it: 78% fewer tokens, 16x less work-attributable input, accuracy 96.1% vs 93.7%. Python, no deps, MIT. | Agents & orchestrationModel & task routing | |
| dsh-jev-routergithub.com | 0 | DSH 插件 · 用 Jev（TypeSafe System One）判定推理强度：默认只切思考强度（零缓存代价），自动模型路由出厂关闭 + 硬门禁 + 成本闸。作者 chenshi.ai | Agents & orchestrationModel & task routing | |
| duet-agent Jev model routinggithub.com | — | Model-routing layer in the duet-agent harness whose default classifier is Jev via AI Gateway, re-checking every 5 assistant steps which model tier should handle the work and switching when the task changes domain. | Agents & orchestrationModel & task routing | |
| duet-agent Jev model routing — repogithub.com | — | Model-routing layer in the duet-agent harness whose default classifier is Jev via AI Gateway, re-checking every 5 assistant steps which model tier should handle the work and switching when the task changes domain. | Agents & orchestrationModel & task routing | |
| DurinDoor Jev routing classifiergithub.com | — | Self-hosted AI gateway over 236 providers that can ask Jev to classify each request as simple, medium, complex or reasoning and route it to a combo member of matching task level. | Agents & orchestrationModel & task routing | |
| DurinDoor Jev routing classifier — appdurindoor.vercel.app | — | Self-hosted AI gateway over 236 providers that can ask Jev to classify each request as simple, medium, complex or reasoning and route it to a combo member of matching task level. | Agents & orchestrationModel & task routing | |
| DurinDoor Jev routing classifier — repogithub.com | — | Self-hosted AI gateway over 236 providers that can ask Jev to classify each request as simple, medium, complex or reasoning and route it to a combo member of matching task level. | Agents & orchestrationModel & task routing | |
| effort-routergithub.com | 0 | Picks the reasoning effort for each turn of Claude Code with a System One classifier | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 3 | BeLazy167/typesafe-mod — Claude Code mod that routes decisions to TypeSafe's Jev model: ranks installed skills per prompt, and answers the agent's own this-or-that questions when confident. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 330 | BillionsBobby/JevRouter — A lightweight Jev-powered router for models, tools, and subagents · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 14 | DECRUX9812/typesafe-skill-router — TypeSafe (Jev) skill routing for Hermes Agent: names the one skill worth loading, before the model call. Opt-in, stdlib only, ~$0.001 per routed turn. · endpoint · Python | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 23 | GodsBoy/jev-agent-skill-router — Typed, confidence-aware agent skill routing with TypeSafe Jev. · endpoint · Python | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 1 | Mandrilsquad1441/jev-model-router — Pick the best AI model and reasoning effort for any task in ~1s. Plugin for Claude Code, Claude Desktop and Codex, powered by TypeSafe's Jev decision model and live OpenRouter pricing. Balance intelligence, speed and cost, or choose your priority. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 3 | MongLong0214/jev-gate — Not every coding task needs your best model. Experimental Jev-powered model routing for Claude Code — V3 prototype runs today, V4 routes at the task boundary. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 5 | bestagentkits/jev-skillful — Per-prompt capability router for coding agents: resolves installed skills, MCP servers, agents and commands against your prompt via TypeSafe Jev, and measures whether the injection actually helps. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 3 | ddfeyes/jev-mode — I kept watching coding agents burn context on decisions that aren't hard - triage 400 tickets, tag 600 files, route to one of six teams. jev-mode moves those verdicts to a typed-judgment model. I A/B'd it: 78% fewer tokens, 16x less work-attributable input, accuracy 96.1% vs 93.7%. Python, no deps, MIT. · endpoint · Python | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 2 | gholtzap/jev-codex-model-and-effort-router — Copy and paste this into your coding agent: · endpoint · Python | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 3 | miniLV/Jev-Auto-Router — Codex 可恢复委派：Capsule → Jev 选路 → Policy Guard（仅 ALLOW/DENY）→ Root 验收。自动委派默认关闭。Recoverable Codex delegation via Jev + Policy Guard; Root keeps acceptance. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 1 | purplesmoke05/opencode-plugin-jev-auto-model-router — Opt-in Auto (Jev) model routing for OpenCode with a configurable model allowlist · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| endpointgithub.com | 1 | redrossa/pi-model-router — Pi extension that classifies each prompt with TypeSafe's Jev and routes it to the best model you're logged into. · endpoint · TypeScript | Agents & orchestrationModel & task routing | |
| exemplo-hev-roteamentogithub.com | 0 | Roteamento automático de incidentes para o time certo usando o modelo Jev (TypeSafe AI) via OpenRouter, com benchmark de 1000 requisições. | Agents & orchestrationModel & task routing | |
| Frostgithub.com | 0 | Go CLI that takes a prompt or markdown doc and recommends a model, harness or API and effort level, using one Jev analysis request and then deterministic local selection of the cheapest profile meeting the quality floor. | Agents & orchestrationModel & task routing | |
| gregb100/gavelgithub.com | 1 | Stop burning LLM calls on classification. Route bugs, triage failures, and gate PRs in 200ms for $0.00002. OpenClaw plugin for TypeSafe Jev structured decisions. | Agents & orchestrationModel & task routing | |
| Grok Bot Jev Routergithub.com | 74 | Reference router that puts Jev in front of Grok Bot to classify each request before costly research, browser, retry or subagent work, choosing actions like reusing a cached artifact, stopping a retry or asking a human. | Agents & orchestrationModel & task routing | |
| grok-jev-routergithub.com | 1 | Jev decides, Grok Bot executes, humans control irreversible actions. A decision router for Grok Bot built on TypeSafe AI's Jev. | Agents & orchestrationModel & task routing | |
| Helmgithub.com | 14 | Coding-agent router and supervisor, installable as a Claude Code plugin, Gemini extension, or agent skill, that probes which agents you have and asks Jev to pick one per task and judge if the worker finished. | Agents & orchestrationModel & task routing | |
| Helm API Jev classifiergithub.com | — | Request classifier in the Helm API self-hosted LLM routing gateway that sends Jev the last user message and gets task type and complexity choices, which pick the lane and provider for the request. | Agents & orchestrationModel & task routing | |
| Helm API Jev classifier — repogithub.com | — | Request classifier in the Helm API self-hosted LLM routing gateway that sends Jev the last user message and gets task type and complexity choices, which pick the lane and provider for the request. | Agents & orchestrationModel & task routing | |
| Hermes Jev Routergithub.com | 1 | Experimental Hermes plugin whose jev_route_plan tool uses Jev to assess task complexity, then policy picks an allowed model within cost, token, context, latency and capability limits; not yet run live. | Agents & orchestrationModel & task routing | |
| hermes-jev-skillsgithub.com | 957 | Hands an agent's small decisions to Jev: which model answers the turn, which skills to load, which passages matter, which turns survive compaction. | Agents & orchestrationModel & task routing | |
| HiRoute Jev decidergithub.com | — | Reference decision service for the HiRoute local model router that makes one Jev call per agent-turn boundary to pick the next model branch and score how competently the previous segment went. | Agents & orchestrationModel & task routing | |
| HiRoute Jev decider — repogithub.com | — | Reference decision service for the HiRoute local model router that makes one Jev call per agent-turn boundary to pick the next model branch and score how competently the previous segment went. | Agents & orchestrationModel & task routing | |
| jcm-routergithub.com | 8 | Local proxy that uses Jev to pick model and effort per Claude Code message; after main-chat switches caused $17.12 of a $19.53 loss in cache rewrites, it routes subagents freely and the main chat only when a switch is cheap. | Agents & orchestrationModel & task routing | |
| jcpsimmons/jev-model-router-demogithub.com | 2 | Throwaway Jev demo: route coding tasks to Grok Build or Codex Astra | Agents & orchestrationModel & task routing | |
| Jev Agent Skill Routergithub.com | 23 | Routing layer that picks which Hermes Agent skill to load using confidence-aware Jev decisions and declines weak matches; it routed 68 of 72 synthetic requests correctly versus 51 for a lexical baseline. | Agents & orchestrationModel & task routing | |
| Jev Auto Routergithub.com | 3 | Prototype local Responses proxy for Codex in which Jev picks the GPT model and reasoning effort for every model call, followed by an independent check that the task was actually completed. | Agents & orchestrationModel & task routing | |
| Jev Router for Pigithub.com | 16 | Pi coding-agent extension that has Jev choose a model and reasoning effort through Vercel AI Gateway, then keeps that pick pinned for the session while generation uses your existing Pi providers. | Agents & orchestrationModel & task routing | |
| Jev Router Playgroundgithub.com | 2 | Playground where you build a pool of OpenRouter models with routing descriptions, give Jev a task, and compare its pick and probability distribution with the answer you judge best. | Agents & orchestrationModel & task routing | |
| Jev Router Playground — appkvhx37ziab90c.space.minimax.io | — | Playground where you build a pool of OpenRouter models with routing descriptions, give Jev a task, and compare its pick and probability distribution with the answer you judge best. | Agents & orchestrationModel & task routing | |
| Jev vs LLM model routerwww.reddit.com | — | Smart model router that runs Jev and a structured-output LLM side by side on every message; Jev returns the routing decision in about 1 second vs 4-14 seconds for the LLM. | Agents & orchestrationModel & task routing | |
| jev-agent-hooksgithub.com | 2 | TypeSafe Jev hooks for Claude Code, Codex and pi: per-turn skill suggestion and subagent model routing | Agents & orchestrationModel & task routing | |
| jev-agent-routergithub.com | 0 | Code when exact. Jev when useful. Agent when uncertain. A lightweight routing layer for computer-use agents, with real workflow benchmarks. | Agents & orchestrationModel & task routing | |
| jev-architectgithub.com | 2 | Makes Claude Code and Codex ask whether a step needs a generative LLM at all — or whether it belongs on Jev, TypeSafe's System One model. Architecture arbitrage before the code is written. | Agents & orchestrationModel & task routing | |
| jev-clawgithub.com | 4 | Typed model routing for OpenClaw agents, powered by TypeSafe Jev | Agents & orchestrationModel & task routing | |
| jev-codex-bridgegithub.com | 6 | Model and reasoning routing for Codex Desktop and CLI, with a Windows service and validated updates | Agents & orchestrationModel & task routing | |
| jev-codex-model-and-effort-routergithub.com | 2 | Codex wrapper for macOS that picks the model and reasoning effort for the first task in a thread and pins it to keep the cache warm, with a menu bar app for routing settings. | Agents & orchestrationModel & task routing | |
| jev-codex-router-skillgithub.com | — | Decision Tools: Portable Codex Skill for Jev model and reasoning-effort routing, with safe installation and Chinese usage guides | Agents & orchestrationModel & task routing | |
| jev-control-planegithub.com | 0 | JEV powered model and reasoning routing for delegated Codex work | Agents & orchestrationModel & task routing | |
| jev-decision-gatewaygithub.com | 0 | Provider-agnostic gateway in front of expensive LLMs: Jev answers small continue, tool, relevance, and verification questions, and a policy layer calls a generative model only when generation is needed. | Agents & orchestrationModel & task routing | |
| jev-flash-routergithub.com | 3 | open-sourced jev-flash-router: an MCP server for TypeSafe's new Jev model. AI coding agents waste hundreds of reasoning tokens just deciding which file to edit, which route to pick, or whether a diff breaks tests. Jev evaluates state and outputs calibrated probabilities. Works with Cursor, Windsurf, & Claude Code | Agents & orchestrationModel & task routing | |
| jev-free-routergithub.com | 0 | OpenCode provider plugin where Jev estimates, on every new message, which model in the free Zen or Go catalogue is best for the request and routes to it, with OpenAI and Anthropic models excluded. | Agents & orchestrationModel & task routing | |
| jev-gategithub.com | 3 | Experimental Claude Code model routing where Sonnet coordinates, a read-only planner decomposes the job, and Jev picks a model tier for every delegated planner and worker task; no cost benefit is established yet. | Agents & orchestrationModel & task routing | |
| jev-healthcare-support-routergithub.com | 0 | TypeScript demo using Jev as a decision layer for healthcare customer-support routing, urgency detection, and human escalation. | Agents & orchestrationModel & task routing | |
| jev-in-jevgithub.com | 1 | JEV in JEV: an AI task router that uses Jev to decide whether a task needs Jev. | Agents & orchestrationModel & task routing | |
| jev-mistral-reflexgithub.com | 0 | Route bounded decisions through Jev and hand open-ended work to Mistral. | Agents & orchestrationModel & task routing | |
| jev-model-routergithub.com | 9 | TypeScript router that sends each query to the cheapest capable model in a two-tier or three-tier cascade, using Jev to classify how demanding the query is instead of an LLM call. | Agents & orchestrationModel & task routing | |
| jev-model-router — demowww.reddit.com | — | TypeScript router that sends each query to the cheapest capable model in a two-tier or three-tier cascade, using Jev to classify how demanding the query is instead of an LLM call. | Agents & orchestrationModel & task routing | |
| jev-model-router — discussionwww.reddit.com | — | TypeScript router that sends each query to the cheapest capable model in a two-tier or three-tier cascade, using Jev to classify how demanding the query is instead of an LLM call. | Agents & orchestrationModel & task routing | |
| jev-model-router — writeupaz9713.github.io | — | Web chat where Jev picks which LLM should answer each message and the chosen model replies, both through the Vercel AI Gateway, with a documented development journey and reliability evaluation. | Agents & orchestrationModel & task routing | |
| jev-pi-model-routergithub.com | 0 | pi extension: TypeSafe Jev picks the model for each turn, only from models pi can use | Agents & orchestrationModel & task routing | |
| jev-routergithub.com | 8 | Pass-through model router for Claude Code and Codex CLI that picks a model tier per human turn with Jev, TypeSafe AI's decision model | Agents & orchestrationModel & task routing | |
| jev-routergithub.com | 14 | OpenAI-compatible LiteLLM proxy where clients request one model id and Jev picks which provider model serves each request, after filters for vision, output length and tool support. | Agents & orchestrationModel & task routing | |
| jev-routergithub.com | 5 | automatic model and effort switching for Claude Code | Agents & orchestrationModel & task routing | |
| jev-skill-routergithub.com | 6 | Claude Code plugin: asks TypeSafe Jev which installed skill fits each prompt and logs the answer (shadow-first). A working reference for the skill-suggestion cookbook on Claude Code — the README records why it is unlikely to help a strong model as a router. | Agents & orchestrationModel & task routing | |
| jev-smart-routergithub.com | 4 | JEV Smart Router — a Databricks App that uses TypeSafe JEV to pick which model answers each message, then runs inference on the chosen Databricks Foundation Model API endpoint. | Agents & orchestrationModel & task routing | |
| jev-subagent-routergithub.com | 0 | Use TypeSafe Jev to select subagent models and reasoning effort in Codex and Claude Code. | Agents & orchestrationModel & task routing | |
| jev-telegram-channel-routergithub.com | 0 | Jev as a channel message router | Jev频道消息路由 | Agents & orchestrationModel & task routing | |
| jev-usage-routerx.com | — | Usage router for Grok Bot: before browsing, research, retries or spawning extra bots, a Jev Choice picks the route, with a shadow mode, logs and a kill switch before it goes active. | Agents & orchestrationModel & task routing | |
| jev_routergithub.com | 0 | Standalone TypeSafe Jev-based auto-routing for Agent Zero | Agents & orchestrationModel & task routing | |
| jevgaugegithub.com | 1 | One-shot model and reasoning routing for Hermes Desktop. Manual control wins. Explicit upstream integration preview. | Agents & orchestrationModel & task routing | |
| JevRouter — appjevrouter.co | — | Treats models, subagents, skills, MCP servers, and CLIs as one candidate pool and routes with a two-stage Choice when the pool is large. | Agents & orchestrationModel & task routing | |
| LiteLLM auto routerdocs.litellm.ai | — | Picks a model tier for each request with a Jev Choice. | Agents & orchestrationModel & task routing | |
| LiteLLM Jev complexity routergithub.com | 59,200 | LiteLLM's complexity-based router can use Jev to classify each request into configured tiers that pick the backend model; a separate guardrail blanks tool results Jev judges no longer needed. | Agents & orchestrationModel & task routing | |
| LiteLLM Jev complexity router — guardrailgithub.com | 59,200 | LiteLLM's complexity-based router can use Jev to classify each request into configured tiers that pick the backend model; a separate guardrail blanks tool results Jev judges no longer needed. | Agents & orchestrationModel & task routing | |
| Loki Jev Autogithub.com | 23 | Jev Auto router in the Loki self-improving agent that picks a model for each new session from a bounded catalog on the already-selected gateway, plus optional Jev typed-judgment tools. | Agents & orchestrationModel & task routing | |
| Loki Jev Auto — discussionwww.reddit.com | — | Jev Auto router in the Loki self-improving agent that picks a model for each new session from a bounded catalog on the already-selected gateway, plus optional Jev typed-judgment tools. | Agents & orchestrationModel & task routing | |
| Mandrilsquad1441/jev-model-routergithub.com | 1 | Pick the best AI model and reasoning effort for any task in ~1s. Plugin for Claude Code, Claude Desktop and Codex, powered by TypeSafe's Jev decision model and live OpenRouter pricing. Balance intelligence, speed and cost, or choose your priority. | Agents & orchestrationModel & task routing | |
| Misergithub.com | 5 | Rust AI gateway that has Jev classify every OpenAI-compatible prompt into a complexity tier from trivial to reasoning and routes it to the cheapest capable OpenRouter model; 90.5% exact accuracy on 116 held-out cases. | Agents & orchestrationModel & task routing | |
| model_idgithub.com | 3 | Ravinder82/jev-flash-router — open-sourced jev-flash-router: an MCP server for TypeSafe's new Jev model. AI coding agents waste hundreds of reasoning tokens just deciding which file to edit, which route to pick, or whether a diff breaks tests. Jev evaluates state and outputs calibrated probabilities. Works with Cursor, Windsurf, & Claude Code · model_id · TypeScript | Agents & orchestrationModel & task routing | |
| model_idgithub.com | 1 | Z761293629/pi-jev-helm — Pi extension that uses Jev task classification (via OpenRouter) to route each run to explicitly configured models with fail-open policy. Public Preview. · model_id · TypeScript | Agents & orchestrationModel & task routing | |
| model_idgithub.com | 1 | gregb100/gavel — Stop burning LLM calls on classification. Route bugs, triage failures, and gate PRs in 200ms for $0.00002. OpenClaw plugin for TypeSafe Jev structured decisions. · model_id · Python | Agents & orchestrationModel & task routing | |
| model_idgithub.com | 9 | matthewp/flue-jev-demo — Flue agent routing with TypeSafe Jev through Cloudflare AI Gateway · model_id · TypeScript | Agents & orchestrationModel & task routing | |
| model_idgithub.com | 1 | trajectoire-ai/hermes-structured-aux-models — A Hermes Agent model-provider plugin that routes selected auxiliary tasks through bounded Jev decision calls on OpenRouter instead of free-form chat prompts. · model_id · Python | Agents & orchestrationModel & task routing | |
| Nebius Physical AI Jev model routergithub.com | — | Advisory model router in Nebius's Physical AI Workbench agent backend that asks Jev whether a request needs the fast Nemotron model or MiniMax-M3 for multi-step reasoning, selecting no model on abstention or failure. | Agents & orchestrationModel & task routing | |
| Nebius Physical AI Jev model router — repogithub.com | — | Advisory model router in Nebius's Physical AI Workbench agent backend that asks Jev whether a request needs the fast Nemotron model or MiniMax-M3 for multi-step reasoning, selecting no model on abstention or failure. | Agents & orchestrationModel & task routing | |
| opencode-jev-routergithub.com | 8 | Adaptive reasoning effort for OpenCode via Jev, with an in-process plugin and Responses API proxy | Agents & orchestrationModel & task routing | |
| OriginOS Jev perception decisionsgithub.com | — | Optional Jev decision mode for the perception rules of OriginOS, a local AI work system: when an event could go to several authorized projects, roles or skills, Jev picks the target and low-confidence cases go to the user. | Agents & orchestrationModel & task routing | |
| OriginOS Jev perception decisions — appwolfgaze.cn | — | Optional Jev decision mode for the perception rules of OriginOS, a local AI work system: when an event could go to several authorized projects, roles or skills, Jev picks the target and low-confidence cases go to the user. | Agents & orchestrationModel & task routing | |
| OriginOS Jev perception decisions — repogithub.com | — | Optional Jev decision mode for the perception rules of OriginOS, a local AI work system: when an event could go to several authorized projects, roles or skills, Jev picks the target and low-confidence cases go to the user. | Agents & orchestrationModel & task routing | |
| Pi Jev Helmgithub.com | 1 | Pi coding-agent extension that sends each new unit of work to Jev through OpenRouter or TypeSafe for task classification and routes the run to an explicitly configured model, failing open on errors. | Agents & orchestrationModel & task routing | |
| pi-jev-routergithub.com | 0 | Automatic model routing for Pi using TypeSafe's Jev through Vercel AI Gateway | Agents & orchestrationModel & task routing | |
| pi-jev-task-routergithub.com | 3 | Per-prompt model routing for the Pi coding agent: classify each prompt with Jev, pick a model tier, and switch before the turn runs. | Agents & orchestrationModel & task routing | |
| pi-model-routergithub.com | 1 | Pi extension that classifies each prompt with TypeSafe's Jev and routes it to the best model you're logged into. | Agents & orchestrationModel & task routing | |
| pi-typesafe-routergithub.com | — | Pi extension that has Jev classify each request and route it to one of three configured models, with a doctor command that tests the classifier and models on synthetic requests before routing is enabled. | Agents & orchestrationModel & task routing | |
| purplesmoke05/opencode-plugin-jev-auto-model-routergithub.com | 1 | Opt-in Auto (Jev) model routing for OpenCode with a configurable model allowlist | Agents & orchestrationModel & task routing | |
| reflex-controlgithub.com | 0 | Rust policy engine using TypeSafe Jev and deterministic checks to route AI agent decisions. | Agents & orchestrationModel & task routing | |
| RelayPlane Jev classifiergithub.com | — | Local LLM cost proxy with an optional Jev complexity classifier that rates each request to inform model routing, enabled by one env var and capped at a 500 ms timeout. | Agents & orchestrationModel & task routing | |
| RelayPlane Jev classifier — apprelayplane.com | — | Local LLM cost proxy with an optional Jev complexity classifier that rates each request to inform model routing, enabled by one env var and capped at a 500 ms timeout. | Agents & orchestrationModel & task routing | |
| Safeplane routing advisorgithub.com | — | Experimental routing advisor in Safeplane, a local-first control plane for bounded AI-agent workflows with CLI and Telegram connectors, that asks Jev via OpenRouter Decisions which workflow an operator request should go to. | Agents & orchestrationModel & task routing | |
| Safeplane routing advisor — repogithub.com | — | Experimental routing advisor in Safeplane, a local-first control plane for bounded AI-agent workflows with CLI and Telegram connectors, that asks Jev via OpenRouter Decisions which workflow an operator request should go to. | Agents & orchestrationModel & task routing | |
| sdkgithub.com | 513 | gargpratyush/jev-router — Route to the cheapest model in claude code for your task using jev-router · sdk · JavaScript | Agents & orchestrationModel & task routing | |
| sdkgithub.com | 2 | lomeshdutta/skill-router — Tell Claude Code which installed skill a session needs, using Jev (TypeSafe AI) for the decision and skills.sh for discovery. · sdk · Python | Agents & orchestrationModel & task routing | |
| sdkgithub.com | 100 | nidhi-singh02/agent-router — CLI that picks Cursor, Claude Code, Codex, or OpenCode + model/effort for a task, then launches it. Powered by Jev and Herdr · sdk · TypeScript | Agents & orchestrationModel & task routing | |
| sizeupgithub.com | 0 | Jev-powered model router for AI coding agents (omp + Claude Code): small jobs go to smaller, cheaper models. | Agents & orchestrationModel & task routing | |
| skill-routergithub.com | 2 | Command-line tool for Claude Code that uses Jev to decide which installed skill a session needs and checks skills.sh for ones you have not installed yet. | Agents & orchestrationModel & task routing | |
| SLO Router — resultsgithub.com | 8 | OpenAI-compatible proxy that routes each request to the cheapest backend meeting its latency SLO, with optional Jev 1.13 semantic features; a live run found Jev changed no routes and raised p95 from 77.93 ms to 490.38 ms. | Agents & orchestrationModel & task routing | |
| Starchild prompt routingx.com | — | LLM router in the Starchild app that classifies prompts in real time (avg 140ms), reported as a 20x cost decrease and 6x speedup for prompt classification. | Agents & orchestrationModel & task routing | |
| SwarmRoutergithub.com | 0 | Routes a task to a research, code, browser, support or writer agent with a confidence value, shown on a visual swarm map. | Agents & orchestrationModel & task routing | |
| SwarmRouter — appswarmrouter.vercel.app | — | Routes a task to a research, code, browser, support or writer agent with a confidence value, shown on a visual swarm map. | Agents & orchestrationModel & task routing | |
| Switchboardgithub.com | 16 | Local router for Claude Code and Codex that assesses a new conversation's task with Jev, applies deterministic confidence rules to pick model and reasoning effort, and pins that route through follow-ups. | Agents & orchestrationModel & task routing | |
| the-array Jev decision layergithub.com | — | Opt-in Jev layer in a self-hosted LiteLLM gateway reference build that picks model tiers and referees replies as answered, refused, or non-answer, moving to the next model after a confident decline. | Agents & orchestrationModel & task routing | |
| the-array Jev decision layer — codegithub.com | — | Opt-in Jev layer in a self-hosted LiteLLM gateway reference build that picks model tiers and referees replies as answered, refused, or non-answer, moving to the next model after a confident decline. | Agents & orchestrationModel & task routing | |
| the-array Jev decision layer — discussionwww.reddit.com | — | Opt-in Jev layer in a self-hosted LiteLLM gateway reference build that picks model tiers and referees replies as answered, refused, or non-answer, moving to the next model after a confident decline. | Agents & orchestrationModel & task routing | |
| the-array Jev decision layer — repogithub.com | — | Opt-in Jev layer in a self-hosted LiteLLM gateway reference build that picks model tiers and referees replies as answered, refused, or non-answer, moving to the next model after a confident decline. | Agents & orchestrationModel & task routing | |
| the-llm-dispatchergithub.com | 0 | An LLM router that uses Jev (TypeSafe System One) to dispatch each request to the cheapest model and lowest reasoning effort that will do the job. OpenAI-compatible proxy, Claude Code and Codex wrappers, TypeScript and C#. | Agents & orchestrationModel & task routing | |
| tiershiftgithub.com | 3 | Model router for TypeScript and Python that sends each LLM request to the cheapest capable tier and escalates on multi-step reasoning, difficulty, stakes, or safety, with policy in plain YAML. | Agents & orchestrationModel & task routing | |
| tink-routegithub.com | 4 | Agent Skills router that keeps skill libraries offline and loads only the needed ones on demand through a two-stage Jev gate (a Noul on whether a skill is needed, then a Choice of which), avoiding per-turn prompt bloat. | Agents & orchestrationModel & task routing | |
| todo-jevgithub.com | 3 | Task-routing experiment that classifies a request with Jev and recommends a local rule, a specialist skill or a foundation model, combining skill apply/exclude conditions with local environment checks. | Agents & orchestrationModel & task routing | |
| typesafe-jev-incident-routergithub.com | 8 | Confidence-gated incident routing with TypeSafe Jev | Agents & orchestrationModel & task routing | |
| typesafe-mcpgithub.com | 0 | TypeSafe System One dispatch pipeline as a ZCode plugin: typed subagent routing (Battery #1), deterministic thresholds, destructive-op guard, usage ledger | Agents & orchestrationModel & task routing | |
| typesafe-skill-routergithub.com | 14 | Opt-in Hermes Agent plugin that asks Jev which one installed skill fits a request before the model call and appends a single hint line, about $0.001 per routed turn, injecting nothing when none fits. | Agents & orchestrationModel & task routing | |
| unhardcoded-engine decision protocolgithub.com | — | Decision-model routing in unhardcoded-engine, a pure-Lua policy algebra for LLM provider selection, that partitions candidates by protocol so Jev-style decision requests only route and fail over among decision models. | Agents & orchestrationModel & task routing | |
| unhardcoded-engine decision protocol — repogithub.com | — | Decision-model routing in unhardcoded-engine, a pure-Lua policy algebra for LLM provider selection, that partitions candidates by protocol so Jev-style decision requests only route and fail over among decision models. | Agents & orchestrationModel & task routing |
Memory & context111 of 111 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| ai_sdkgithub.com | 1 | Dharundp6/jev-carryforward — What your last session knew, scored against what this one is doing. MCP server: a per-project ledger written as things happen, recalled per task with TypeSafe's Jev evaluation model via Vercel AI Gateway. · ai_sdk · TypeScript | Agents & orchestrationMemory & context | |
| ai_sdkgithub.com | 28 | compozy/yoshi — Context-pruning proxy for Claude Code and Codex: Jev judges which history is still needed, measured not claimed. POC here now, heading soon into https://github.com/compozy/compozy · ai_sdk · TypeScript | Agents & orchestrationMemory & context | |
| alphaoptimizergithub.com | — | Context GC: AlphaOptimizer is an open-source tool from AlphaTales that helps Codex work with large command and tool outputs. Instead of sending a huge log or search result straight into the context window, AlphaOptimizer keeps the useful parts visible, keeps the original output available for a limited time, and uses Jev to help rank what matters when an API key is configured. | Agents & orchestrationMemory & context | |
| Archive Center Jev selectorgithub.com | — | Optional Jev selector and reviewer in Archive Center, a local-first memory backend for RisuAI chats, that picks which stored memories and source evidence to pass into the next request without ever writing memories. | Agents & orchestrationMemory & context | |
| Archive Center Jev selector — apparchive-center.vercel.app | — | Optional Jev selector and reviewer in Archive Center, a local-first memory backend for RisuAI chats, that picks which stored memories and source evidence to pass into the next request without ever writing memories. | Agents & orchestrationMemory & context | |
| Archive Center Jev selector — repogithub.com | — | Optional Jev selector and reviewer in Archive Center, a local-first memory backend for RisuAI chats, that picks which stored memories and source evidence to pass into the next request without ever writing memories. | Agents & orchestrationMemory & context | |
| Beacon Jev memory evaluationsgithub.com | ≈1,600 | Evaluation command in Beacon, a cross-harness memory layer for coding agents, that has Jev judge redacted session-trace projections against a rubric for reusable lessons, which a person reviews before saving. | Agents & orchestrationMemory & context | |
| Beacon Jev memory evaluations — repogithub.com | ≈1,600 | Evaluation command in Beacon, a cross-harness memory layer for coding agents, that has Jev judge redacted session-trace projections against a rubric for reusable lessons, which a person reviews before saving. | Agents & orchestrationMemory & context | |
| Beacon Jev memory evaluations — sitebeacon.sh | — | Evaluation command in Beacon, a cross-harness memory layer for coding agents, that has Jev judge redacted session-trace projections against a rubric for reusable lessons, which a person reviews before saving. | Agents & orchestrationMemory & context | |
| carryforwardgithub.com | 1 | MCP server that records facts during agent sessions in a per-project ledger and, at the start of a task, uses Jev via Vercel AI Gateway to recall only the facts that matter for it. | Agents & orchestrationMemory & context | |
| carryforward — npmwww.npmjs.com | — | MCP server that records facts during agent sessions in a per-project ledger and, at the start of a task, uses Jev via Vercel AI Gateway to recall only the facts that matter for it. | Agents & orchestrationMemory & context | |
| Character memory that judgesx.com | — | Character memory system from Mutuals where one Jev call per message decides whether to keep it, which exact words, where it fits, what it makes outdated and what it recalls. | Agents & orchestrationMemory & context | |
| Character memory that judges — companymutuals.inc | — | Character memory system from Mutuals where one Jev call per message decides whether to keep it, which exact words, where it fits, what it makes outdated and what it recalls. | Agents & orchestrationMemory & context | |
| christian-taillon/opencode-jev-compactorgithub.com | 2 | Jev powered OpenCode compaction | Agents & orchestrationMemory & context | |
| Citadel promotion decisionsgithub.com | — | Promotion step in Citadel, self-hosted memory for engineering teams and their agents, where Jev decides whether a collected item is relevant and safe to promote into shared knowledge at a 0.7 threshold. | Agents & orchestrationMemory & context | |
| Citadel promotion decisions — appgithub.com | 24 | Promotion step in Citadel, self-hosted memory for engineering teams and their agents, where Jev decides whether a collected item is relevant and safe to promote into shared knowledge at a 0.7 threshold. | Agents & orchestrationMemory & context | |
| Citadel promotion decisions — repogithub.com | — | Promotion step in Citadel, self-hosted memory for engineering teams and their agents, where Jev decides whether a collected item is relevant and safe to promote into shared knowledge at a 0.7 threshold. | Agents & orchestrationMemory & context | |
| codex-jev-compactiongithub.com | 4 | Codex plugin with a skill, MCP tool and CLI that builds compact, traceable handoff packets by having Jev pick which old read-only tool records to keep verbatim for the next task or session. | Agents & orchestrationMemory & context | |
| CompleteTech-LLC-AI-Research/jev-prune-kitgithub.com | 1 | Capability-aware context-pruning installer, bounded Jev assessment engine, and experimental harness request-projection adapters (experimental, not live-host tested) | Agents & orchestrationMemory & context | |
| Context Dietgithub.com | 3 | Codex plugin that asks Jev which bulky tool results a session still needs, such as test logs, builds and large file reads, and replaces the rest with a bounded head and a one-line note. | Agents & orchestrationMemory & context | |
| ctxjevgithub.com | 0 | Score AI agent context for relevance with Jev, and prune what's no longer useful. | Agents & orchestrationMemory & context | |
| deepseek-harness-jev-pre-compactiongithub.com | 5 | A pre-compaction advisor for DeepSeek Harness. Runs before the standard compaction-basic backend, using TypeSafe JEV to safely prune low-value tool results from model context. Original session events stay in the append-only log; only the model-visible view is replaced with compact markers or archive pointers to reduce context bloat. | Agents & orchestrationMemory & context | |
| DGUI-HyperMemgithub.com | 2 | Self-hosted memory MCP server on Cloudflare Workers that fuses vector and full-text recall, then uses Jev to rerank memories, assign type and salience, and drop non-durable chatter. | Agents & orchestrationMemory & context | |
| DGUI-HyperMem — datasethuggingface.co | — | Self-hosted memory MCP server on Cloudflare Workers that fuses vector and full-text recall, then uses Jev to rerank memories, assign type and salience, and drop non-durable chatter. | Agents & orchestrationMemory & context | |
| Drift developmental memoryx.com | — | AI companion given a Jev-based developmental memory that records what it tried, how the user reacted and what it learned, keeping behavior consistent across the different models that run it. | Agents & orchestrationMemory & context | |
| Drift developmental memory — specdrift-riverbed.com | — | AI companion given a Jev-based developmental memory that records what it tried, how the user reacted and what it learned, keeping behavior consistent across the different models that run it. | Agents & orchestrationMemory & context | |
| edwardyen724-g/jev-compactorgithub.com | 1 | Context compaction and safety gating for AI agents via TypeSafe Jev: keeps messages verbatim, no summarization. OpenAI, Anthropic, LangChain, CLI, MCP. | Agents & orchestrationMemory & context | |
| endpointgithub.com | 1 | CompleteTech-LLC-AI-Research/jev-prune-kit — Capability-aware context-pruning installer, bounded Jev assessment engine, and experimental harness request-projection adapters (experimental, not live-host tested) · endpoint · Python | Agents & orchestrationMemory & context | |
| endpointgithub.com | 79 | IAmUnbounded/save-token-jev-clean — Portable, Jev-guided context compaction for coding agents. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 6 | Nyarlathoteppppp/pi-jev-context — Cache-neutral context trimming for the pi coding agent, powered by TypeSafe Jev: long tool output cut to verbatim key lines before it enters context, with lossless recall. Measured, with pre-registered benchmarks. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 3 | QuentinDanblon/pi-fast-jev-compaction — Verbatim context pruning for the pi coding agent, scored by TypeSafe Jev: stale tool calls and results are dropped or truncated, everything kept stays verbatim. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 5 | Shashank-H/pi-jev-context-curator — A Jev based context curator for pi · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 4 | Wang-auspicious/codex-jev-compaction — Jev-powered context curation for Codex. Build compact, traceable handoff context through native plugins and skills. · endpoint · JavaScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 1 | Wang-auspicious/pi-jev-compaction — Jev-powered context compaction for Pi. Keep critical instructions and tool history, prune the noise, and fall back gracefully. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 1 | designnotdrum/fast-jev-compaction-codex — Jev-guided verbatim supplemental compaction for Codex sessions · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 6 | fatelei/jev-compact — Jev-scored context compaction for OpenAI Codex CLI — scores every tool call before compaction and restores critical tool outputs verbatim after it · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 9 | jerryfane/omp-jev-compaction — Verbatim Jev-scored context reduction for omp, over TypeSafe or OpenRouter · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 13 | joelhooks/pi-fast-jev-compaction — Pi extension: verbatim context compaction with TypeSafe Jev decisions · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 2 | kevinpita/pi-jev-context — Reversible context pruning for Pi, powered by TypeSafe Jev. Keep useful context without deleting session history. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 1 | kleosr/cursor-clijev-compaction — TypeSafe Jev-scored context recovery for Cursor CLI (agent). Capture tool I/O, score keep/drop, re-inject after native compact. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 3 | leonaaardob/fast-dev-compaction — Codex plugin: verbatim Jev-guided context restoration around session compaction. Port of tamaratran/fast-jev-compaction to Codex lifecycle hooks. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 4 | nrdz-labs/fast-jev-opencode — Jev-scored context pruning for OpenCode: drops stale tool calls and truncates bulky results on the outgoing request — fail-open, cache-backed, configurable live. Port of fast-jev-compaction to the V2 context hook. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 31 | samdotmak/jev-recall — Retrieve by relevance, not resemblance: filter an AI assistant's memories with TypeSafe's Jev · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 7,282 | tamaratran/fast-jev-compaction — Claude Code plugin that replaces the compaction summary with Jev decisions: every tool call and result is scored in one fast request, stale ones are dropped or truncated, everything kept stays verbatim. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 2 | tvdavies/pi-fast-compaction — Fast compaction for Pi: Jev-scored verbatim eviction of stale tool history, with LLM summarisation of the evicted transcript as the fallback. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| endpointgithub.com | 1 | willfish/pi-observational-memory-jev — Jev decides what to keep. Compaction never rewrites the transcript. · endpoint · TypeScript | Agents & orchestrationMemory & context | |
| fast-dev-compactiongithub.com | 3 | Codex port of fast-jev-compaction: before Codex compacts, Jev scores every tool call and result, and the verbatim history it keeps is re-injected afterwards; the author calls it an experimental proof of concept. | Agents & orchestrationMemory & context | |
| Fast-Jev-Agentsgithub.com | 2 | High-performance verbatim context compaction for coding agents (Claude, Codex, Antigravity, Gemini, OpenCode) | Agents & orchestrationMemory & context | |
| fast-jev-compaction-codexgithub.com | 1 | Jev-guided verbatim supplemental compaction for Codex sessions | Agents & orchestrationMemory & context | |
| fast-jev-compaction-pigithub.com | — | Pi extension that runs before context compaction and keeps, truncates, or drops each completed tool call's evidence verbatim, preserving failing test logs, command output, and exact errors. | Agents & orchestrationMemory & context | |
| fast-jev-opencodegithub.com | 0 | OpenCode v1 + v2 plugin: prune stale tool calls and truncate bulky tool results from the outgoing request using TypeSafe Jev decisions (verbatim compaction, no summarization). | Agents & orchestrationMemory & context | |
| Granitegithub.com | 32 | Local-first markdown knowledge compiler and MCP server that serves as agent memory; its only model is Jev, which picks which candidate from a deterministically bounded set answers a question and judges every capture. | Agents & orchestrationMemory & context | |
| Hafiza OSgithub.com | 12 | Turkish-language second-brain system that keeps agent memory in a local Markdown vault for Claude Code, Codex and Antigravity, with Jev as an optional retrieval and review advisor. | Agents & orchestrationMemory & context | |
| Hafiza OS — docsgithub.com | 12 | Turkish-language second-brain system that keeps agent memory in a local Markdown vault for Claude Code, Codex and Antigravity, with Jev as an optional retrieval and review advisor. | Agents & orchestrationMemory & context | |
| Hippo Jev rerankergithub.com | — | Opt-in reranker in the Hippo agent-memory system that batches Jev Noul judgments over the top 40 recalled memories and falls back to a local cross-encoder; the author's study found better ranking but no answer-quality gain. | Agents & orchestrationMemory & context | |
| Hippo Jev reranker — repogithub.com | — | Opt-in reranker in the Hippo agent-memory system that batches Jev Noul judgments over the top 40 recalled memories and falls back to a local cross-encoder; the author's study found better ranking but no answer-quality gain. | Agents & orchestrationMemory & context | |
| Hippo Jev reranker — websitehippo-memory.com | — | Opt-in reranker in the Hippo agent-memory system that batches Jev Noul judgments over the top 40 recalled memories and falls back to a local cross-encoder; the author's study found better ranking but no answer-quality gain. | Agents & orchestrationMemory & context | |
| ingebyd/fast-jev-compaction-openroutergithub.com | 2 | Fork of tamaratran/fast-jev-compaction: Jev via OpenRouter with zero data retention (zdr, data_collection: deny) | Agents & orchestrationMemory & context | |
| invalidategithub.com | 22 | Invalidation layer for agent memory that checks every stored fact against each new event with a Jev yes/no question (about 150 ms and $0.00006 per check), marking superseded memories and queuing unsure ones for a human. | Agents & orchestrationMemory & context | |
| invalidate — appinvalidate-playground.vercel.app | — | Invalidation layer for agent memory that checks every stored fact against each new event with a Jev yes/no question (about 150 ms and $0.00006 per check), marking superseded memories and queuing unsure ones for a human. | Agents & orchestrationMemory & context | |
| invalidate — discussionwww.reddit.com | — | Invalidation layer for agent memory that checks every stored fact against each new event with a Jev yes/no question (about 150 ms and $0.00006 per check), marking superseded memories and queuing unsure ones for a human. | Agents & orchestrationMemory & context | |
| JayDoubleu/cc-mod-jevgithub.com | 1 | Claude Code mod: Jev-scored context pruning through OpenRouter. Verbatim compaction, an optional gate on oversized tool outputs, tests and evals. | Agents & orchestrationMemory & context | |
| Jev Recallgithub.com | 31 | Memory retrieval for AI assistants that returns the memories actually relevant to a query rather than the ones that look similar, aiming for reranker quality at semantic-search cost and speed. | Agents & orchestrationMemory & context | |
| Jev Recall — appjev-recall.vercel.app | — | Memory retrieval for AI assistants that returns the memories actually relevant to a query rather than the ones that look similar, aiming for reranker quality at semantic-search cost and speed. | Agents & orchestrationMemory & context | |
| Jev Recall — demowww.reddit.com | — | Memory retrieval for AI assistants that returns the memories actually relevant to a query rather than the ones that look similar, aiming for reranker quality at semantic-search cost and speed. | Agents & orchestrationMemory & context | |
| Jev Recall — discussionwww.reddit.com | — | Memory retrieval for AI assistants that returns the memories actually relevant to a query rather than the ones that look similar, aiming for reranker quality at semantic-search cost and speed. | Agents & orchestrationMemory & context | |
| jev-compactgithub.com | 6 | Codex CLI plugin that has Jev score every tool call and result before compaction, then re-injects verbatim any output Jev marked as still needed that Codex's built-in summary dropped. | Agents & orchestrationMemory & context | |
| jev-compactiongithub.com | 4 | A context compactor that can only score, never write — so an agent's memory can't hold a fact the transcript never contained. Working demo, runs offline. | Agents & orchestrationMemory & context | |
| jev-compaction (picaye)github.com | 3 | Context compaction for Hermes Agent sessions that never summarizes: Jev scores every tool call, stale calls and results are removed or truncated, and everything kept stays verbatim. | Agents & orchestrationMemory & context | |
| Jev-Memgithub.com | 9 | Agent memory where Jev, as a System One controller, handles admission, multi-relational linking and retrieval while a language model answers; on LoCoMo with GPT-4o-mini the paper reports 11.0% higher answer quality. | Agents & orchestrationMemory & context | |
| Jev-Mem — apphuggingface.co | — | Agent memory where Jev, as a System One controller, handles admission, multi-relational linking and retrieval while a language model answers; on LoCoMo with GPT-4o-mini the paper reports 11.0% higher answer quality. | Agents & orchestrationMemory & context | |
| jev-memorygithub.com | 8 | Long-term memory layer for the Vercel AI SDK where Jev gates what gets written, retrieved and evicted, storing exact strings so memories are never summarized or rewritten. | Agents & orchestrationMemory & context | |
| jev.tggithub.com | — | Context GC: Local Telegram filter stores channel messages locally and sends them in batches to Jev or a local model to keep only messages matching natural-language conditions. | Agents & orchestrationMemory & context | |
| jevcache (semantic cache) — demowww.reddit.com | — | Local OpenAI-compatible proxy that reuses a cached chat completion when Jev judges a new request to have the same intent as an earlier one, and fails open on Jev errors; one test went from 3244ms to 394ms. | Agents & orchestrationMemory & context | |
| jevcache (semantic cache) — npmwww.npmjs.com | — | Local OpenAI-compatible proxy that reuses a cached chat completion when Jev judges a new request to have the same intent as an earlier one, and fails open on Jev errors; one test went from 3244ms to 394ms. | Agents & orchestrationMemory & context | |
| jeverifiergithub.com | — | Context GC: JeVerifier: cheap Jev (TypeSafe) checks that keep code maintainable and docs consistent, plus context retrieval — modest token savings | Agents & orchestrationMemory & context | |
| jevmorygithub.com | 11 | Local zero-dependency memory for coding agents where every stored fact is a verbatim quote graded by Jev confidence, and MEMORY.md is audited against session evidence with SQLite receipts. | Agents & orchestrationMemory & context | |
| jevprunegithub.com | 2 | Claude Code plugin and CLI that filters long command output down to the lines relevant to a task description before the agent reads them; one recorded run cut a 2,979-line test log to 54 lines. | Agents & orchestrationMemory & context | |
| jevskillgithub.com | 2 | Teach your coding agent to stop burning context. Jev (System One) via OpenRouter or TypeSafe: 325ms, 0.000013 USD per decision. A/B tested 99.3% fewer input tokens with accuracy up. Ships a reversible reduce and a ledger that learns when Jev pays off. | Agents & orchestrationMemory & context | |
| mem-jevgithub.com | 2 | Deterministic procedural memory for agents | Agents & orchestrationMemory & context | |
| memory-routergithub.com | 0 | Routing + write-lease service for multi-agent memory systems (Jev-backed, MCP) | Agents & orchestrationMemory & context | |
| mnemon-memory-agentgithub.com | 3 | Long-term memory for AI agents on Jev. Keep raw records, judge them with a fast System 1 model and answer from under 4k tokens of context. | Agents & orchestrationMemory & context | |
| model_idgithub.com | 1 | JayDoubleu/cc-mod-jev — Claude Code mod: Jev-scored context pruning through OpenRouter. Verbatim compaction, an optional gate on oversized tool outputs, tests and evals. · model_id · TypeScript | Agents & orchestrationMemory & context | |
| model_idgithub.com | 2 | ingebyd/fast-jev-compaction-openrouter — Fork of tamaratran/fast-jev-compaction: Jev via OpenRouter with zero data retention (zdr, data\_collection: deny) · model_id · TypeScript | Agents & orchestrationMemory & context | |
| omp-jev-compactiongithub.com | 9 | Context reduction extension for the Oh My Pi coding agent that scores every tool call and result with Jev and truncates those no longer needed to a head plus a recovery note, never rewriting text. | Agents & orchestrationMemory & context | |
| openclaw-jev-compactiongithub.com | 1 | Context engine for OpenClaw that asks Jev which tool calls and results a session still needs, drops the rest, and never summarizes, so what remains is the original text. | Agents & orchestrationMemory & context | |
| PerfectRecallgithub.com | 1 | Agent memory for Hermes that has Jev check every stored memory against short criteria from the calling agent, with no embeddings; an earlier version cut LongMemEval-S answer errors from 62 to 17 of 120. | Agents & orchestrationMemory & context | |
| pi-fast-jev-compactiongithub.com | 2 | Pi package port of fast-jev-compaction that uses Pi's non-destructive context hook to drop old tool calls and results no longer needed, without rewriting prose or touching the saved transcript. | Agents & orchestrationMemory & context | |
| pi-fast-jev-compactiongithub.com | 13 | Pi extension that keeps conversation text verbatim and prunes stale tool history by asking Jev two Nouls per tool call, keep the call and keep its result, falling back to Pi's summary compaction when needed. | Agents & orchestrationMemory & context | |
| pi-fast-jev-compaction (QuentinDanblon)github.com | 3 | Port of fast-jev-compaction to the Pi coding agent that asks Jev whether each old tool call and result is still needed and drops or truncates stale ones, leaving everything else verbatim. | Agents & orchestrationMemory & context | |
| pi-jev-compactgithub.com | — | Context compaction extension for the Pi coding agent that drops tool calls and results that are no longer needed and keeps the rest verbatim instead of summarizing. | Agents & orchestrationMemory & context | |
| pi-jev-compactiongithub.com | 1 | Extractive context compaction for the Pi coding agent where Jev judges complete read-only tool pairs and code drops the unnecessary ones, keeping original text instead of a generated summary. | Agents & orchestrationMemory & context | |
| pi-jev-contextgithub.com | 6 | Pi context extension that folds repeated file reads, has Jev filter long command logs under hard keep rules, and saves originals as searchable verbatim chunks, without rewriting old messages or the cache prefix. | Agents & orchestrationMemory & context | |
| pi-jev-contextgithub.com | 2 | Opt-in Pi extension that hides older low-value messages from future model requests while keeping the full session history, with a branch-aware cache. | Agents & orchestrationMemory & context | |
| RememberStack TypeSafe adaptergithub.com | — | Open memory infrastructure for AI agents that tracks claims, current beliefs and their sources, with a System One port whose TypeSafe adapter runs typed evaluations on Jev. | Agents & orchestrationMemory & context | |
| RememberStack TypeSafe adapter — appremember.dev | — | Open memory infrastructure for AI agents that tracks claims, current beliefs and their sources, with a System One port whose TypeSafe adapter runs typed evaluations on Jev. | Agents & orchestrationMemory & context | |
| RememberStack TypeSafe adapter — repogithub.com | — | Open memory infrastructure for AI agents that tracks claims, current beliefs and their sources, with a System One port whose TypeSafe adapter runs typed evaluations on Jev. | Agents & orchestrationMemory & context | |
| sdkgithub.com | 22 | chopratejas/invalidate — The invalidation layer for AI memory. Every fact gets a lease; new evidence ends it. Built on TypeSafe Jev. · sdk · Python | Agents & orchestrationMemory & context | |
| sdkgithub.com | 1 | edwardyen724-g/jev-compactor — Context compaction and safety gating for AI agents via TypeSafe Jev: keeps messages verbatim, no summarization. OpenAI, Anthropic, LangChain, CLI, MCP. · sdk · TypeScript | Agents & orchestrationMemory & context | |
| Smriti with Jevx.com | — | Case study of adding Jev to Smriti, a local SQLite memory layer for agents: Jev pulled more evidence into context but some answers got worse, improving only modestly after relabeling. | Agents & orchestrationMemory & context | |
| Smriti with Jev — projectgithub.com | — | Case study of adding Jev to Smriti, a local SQLite memory layer for agents: Jev pulled more evidence into context but some answers got worse, improving only modestly after relabeling. | Agents & orchestrationMemory & context | |
| Smriti with Jev — repogithub.com | — | Case study of adding Jev to Smriti, a local SQLite memory layer for agents: Jev pulled more evidence into context but some answers got worse, improving only modestly after relabeling. | Agents & orchestrationMemory & context | |
| total-agent-memorygithub.com | 68 | Local persistent memory for coding agents that can check retrieved facts for contradictions with Jev; median pass time fell from 3.1 s to 1.9 s with accuracy unchanged, costing $0.038 for 78 questions. | Agents & orchestrationMemory & context | |
| total-agent-memory — apptotalmemory.dev | — | Local persistent memory for coding agents that can check retrieved facts for contradictions with Jev; median pass time fell from 3.1 s to 1.9 s with accuracy unchanged, costing $0.038 for 78 questions. | Agents & orchestrationMemory & context | |
| TrashCompactgithub.com | 0 | Deterministic transcript cleanup and compaction support using TypeSafe's Jev. | Agents & orchestrationMemory & context | |
| tvdavies/pi-fast-compactiongithub.com | 2 | Fast compaction for Pi: Jev-scored verbatim eviction of stale tool history, with LLM summarisation of the evicted transcript as the fallback. | Agents & orchestrationMemory & context | |
| virtual-context Jev judgment seamsgithub.com | — | Virtual context memory layer for LLM agents that adds per-seam Jev judgment modes for reranking, intent and temporal checks; on 140 LongMemEval questions Jev cut the mean first-gold rank from 2.33 to 1.14. | Agents & orchestrationMemory & context | |
| virtual-context Jev judgment seams — repogithub.com | — | Virtual context memory layer for LLM agents that adds per-seam Jev judgment modes for reranking, intent and temporal checks; on 140 LongMemEval questions Jev cut the mean first-gold rank from 2.33 to 1.14. | Agents & orchestrationMemory & context | |
| willfish/pi-observational-memory-jevgithub.com | 1 | Jev decides what to keep. Compaction never rewrites the transcript. | Agents & orchestrationMemory & context | |
| winnowgithub.com | 100 | Context sieve for Claude Code that judges every large Read, Bash or Grep result and swaps blocks it is confident you do not need for a three-line stub that restores the full text on demand. | Agents & orchestrationMemory & context | |
| Yoshigithub.com | 28 | Proof-of-concept local proxy for Claude Code and Codex in which Jev judges which conversation history is still needed above a size gate, and Yoshi drops the rest while keeping the tool protocol intact. | Agents & orchestrationMemory & context |
Harnesses & orchestration47 of 47 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| agent-harness-with-jev-llmgithub.com | 1 | A small multi-agent harness built around the pattern from LangChain's post: use a fast, non-generative "System One" classifier (TypeSafe AI's Jev) for structured in-loop decisions -- routing and risky-tool-call gating -- and reserve a real chat LLM for the genuinely open-ended reasoning. | Agents & orchestrationHarnesses & orchestration | |
| agentic-harness-cligithub.com | — | Rust CLI for governing agent-native repositories whose decisions commands build Jev payloads for boolean, choice, and ordinal decisions, store answers as review-required receipts, and gate calibration regressions in CI. | Agents & orchestrationHarnesses & orchestration | |
| AgenticOrg Jev integrationgithub.com | 11 | Enterprise agent orchestration platform with a Jev provider adapter, an offline evaluator and a shadow-mode hook that compares Jev tool-routing decisions with existing runtime outcomes. | Agents & orchestrationHarnesses & orchestration | |
| AgenticOrg Jev integration — docsgithub.com | 11 | Enterprise agent orchestration platform with a Jev provider adapter, an offline evaluator and a shadow-mode hook that compares Jev tool-routing decisions with existing runtime outcomes. | Agents & orchestrationHarnesses & orchestration | |
| AgentRunx.com | — | Agent harness built with pi and Jev in which an agent learns how to do a job, writes itself a general reusable solution, and then gets out of the way. | Agents & orchestrationHarnesses & orchestration | |
| Astragithub.com | 32 | Self-hosted runtime for long-running enterprise agents with EXPLAIN ANALYZE for context and native Jev judgments at four decision points: memory filtering, lesson dismissal, request classification and skill selection. | Agents & orchestrationHarnesses & orchestration | |
| astra-jev-harnessgithub.com | 3 | Agent Skills for Codex Desktop and Claude Code, plus a Codex CLI harness. Jev-assisted context selection. / Codex Desktop・Claude Code向けAgent SkillsとCodex CLI用ハーネス。Jevでコンテキストを選別。 | Agents & orchestrationHarnesses & orchestration | |
| Bicameralgithub.com | 8 | Hybrid coding harness for Pi where any LLM writes code while Jev scores tool calls and edits, and deterministic policy turns the scores into allow, confirm, block, warn or steer. | Agents & orchestrationHarnesses & orchestration | |
| choolagithub.com | 0 | Python framework for building automations with AI coding agents — single-file nodes, JSON payloads, built-in LLM cost guardrails. | Agents & orchestrationHarnesses & orchestration | |
| Clonk agent canvasx.com | — | Agent orchestrator on a canvas where Jev instantly decides actions such as launching agents, instead of waiting on a slow LLM loop. | Agents & orchestrationHarnesses & orchestration | |
| Clonk agent canvas — appclonk.ai | — | Agent orchestrator on a canvas where Jev instantly decides actions such as launching agents, instead of waiting on a slow LLM loop. | Agents & orchestrationHarnesses & orchestration | |
| coding-orchestratorgithub.com | 0 | Orchestrator and subagent setups for Claude Code and Codex: routing, verification, and safe global installation. | Agents & orchestrationHarnesses & orchestration | |
| dsh-jevgithub.com | 9 | Plugin suite for DeepSeek Harness with tool-result pruning, a loop guard, a safety guard, and headless ask handling; its 20-task DeepSWE A/B pilot reports no advantage and no stable harm. | Agents & orchestrationHarnesses & orchestration | |
| endpointgithub.com | 13 | Kevthetech143/super-jev — A small, extensible decision-to-action harness for TypeSafe Jev · endpoint · Python | Agents & orchestrationHarnesses & orchestration | |
| Eutryagithub.com | 11 | Standalone CLI-first agent harness for authorized security research, multi-agent swarms and long-running workflows, with Jev built into its decision loop for attention, candidate rubrics and research micro-steps. | Agents & orchestrationHarnesses & orchestration | |
| Goal-completion verifierx.com | — | Custom verifier for the /goal feature in an agent harness that uses Jev to check after every turn whether the goal is actually complete, replacing an expensive reasoning model. | Agents & orchestrationHarnesses & orchestration | |
| HarnessRouterx.com | — | Open-source layer that runs Codex, Claude Code, Hermes, DeepSeek Harness, a Jev-powered System One harness and 9 more behind one interface, with a Unified Harness Protocol and an OpenAI Responses-compatible API. | Agents & orchestrationHarnesses & orchestration | |
| HarnessRouter — repogithub.com | — | Open-source layer that runs Codex, Claude Code, Hermes, DeepSeek Harness, a Jev-powered System One harness and 9 more behind one interface, with a Unified Harness Protocol and an OpenAI Responses-compatible API. | Agents & orchestrationHarnesses & orchestration | |
| hermes-nervegithub.com | 34 | Nerve is a supervisory nervous system for Hermes agents, adding typed System One decisions, ranking, verification, token-aware oversight, and an opt-in tool gate powered by TypeSafe Jev or Open Source Laya | Agents & orchestrationHarnesses & orchestration | |
| Heym decision modelsgithub.com | — | Heym, a self-hosted agent orchestration runtime, adds a decision-model service built on TypeSafe's /v1/systemone contract so workflows can answer typed questions with tracing and SSRF guards. | Agents & orchestrationHarnesses & orchestration | |
| Heym decision models — appheym.run | — | Heym, a self-hosted agent orchestration runtime, adds a decision-model service built on TypeSafe's /v1/systemone contract so workflows can answer typed questions with tracing and SSRF guards. | Agents & orchestrationHarnesses & orchestration | |
| hraness/algalgithub.com | 2 | ALGAL is a new take on the agent graph: a language and runtime for agentic program evolution. | Agents & orchestrationHarnesses & orchestration | |
| jev-claude-orchestratorgithub.com | 0 | Jev-supervised madmax conductor for Claude Code: tiny slices, parallel subagent workers in worktrees, TypeSafe Jev makes routing/review/QA/supervision decisions. | Agents & orchestrationHarnesses & orchestration | |
| jev-fabricgithub.com | 0 | Native process orchestration with typed, explicit Jev decisions. One Bend executable owns your processes, keeps bounded logs, and asks Jev only when told. Agent skill + curl install. | Agents & orchestrationHarnesses & orchestration | |
| jev-harnessgithub.com | 26 | A custom coding harness for TypeSafe AI's Jev: an LLM proposes, Jev answers narrow questions, code decides, every step leaves a receipt. | Agents & orchestrationHarnesses & orchestration | |
| jev-harnessgithub.com | 0 | Jev (TypeSafe) as the decision layer for Claude Code: permission gate, skill routing, conditional instructions, effort-tiered subagents, context packs. Tested PoC with evals. | Agents & orchestrationHarnesses & orchestration | |
| jev-harnessgithub.com | 10 | Decision harness for coding agents that uses Jev Choice, Score and Noul gates to triage test failures such as missing packages or network flakiness and to stop circular doom loops before a frontier model spends tokens. | Agents & orchestrationHarnesses & orchestration | |
| jevcoregithub.com | 1 | JevCore Agent — local coding harness (Jev decisions, Guard hard policy, MCP + CLI) | Agents & orchestrationHarnesses & orchestration | |
| JevHarness — sitejev-harness.tianyuchen99.chatgpt.site | — | Framework where a strong LLM writes a task-specific harness that turns observations into Jev questions and actions, then freezes it and optionally improves it from rewards and full execution traces with GEPA. | Agents & orchestrationHarnesses & orchestration | |
| jevloopgithub.com | 6 | A Jev-driven general-purpose agent harness for faster, lower-cost execution, with built-in side-by-side experiments against LLM-only agents. | Agents & orchestrationHarnesses & orchestration | |
| JevLoopgithub.com | 21 | Runnable agent harness that routes every fork in the loop, such as which tool, is it safe, am I done, to a decision model like Jev or Laya and keeps the LLM for writing, with a web UI and no dependencies. | Agents & orchestrationHarnesses & orchestration | |
| jevonsgithub.com | 0 | A bounded Pi execution supervisor powered by TypeSafe Jev: decisions, recovery, review and verification. | Agents & orchestrationHarnesses & orchestration | |
| keeltrace/hermes-jevgithub.com | 34 | Nerve is a supervisory nervous system for Hermes agents, adding typed System One decisions, ranking, verification, token-aware oversight, and an opt-in tool gate powered by TypeSafe Jev or Open Source Laya | Agents & orchestrationHarnesses & orchestration | |
| model_idgithub.com | 2 | yunusemrejr/yunuspi — Yunus Pi is a heavily customized harness experience built on top of a forked Pi harness (pi.dev). It uses patches, extensions, skills, and custom provider configs to customize the overall experience. It uses advanced machine learning, Jev, Cactus Needle 3 and similar technologies. · model_id · JavaScript | Agents & orchestrationHarnesses & orchestration | |
| Puppetmaster Jevgithub.com | — | Puppetmaster, a control plane for durable-state agent swarms, uses Jev for conflict, already-answered, admission and stop decisions across its worker stitching. | Agents & orchestrationHarnesses & orchestration | |
| Robo Harnessgithub.com | 2 | SO-101 robot-arm agent workbench with a Bun/Effect coordinator, a React and Rerun workspace and a Python LeRobot motor owner, in which Jev picks bounded next arm actions from camera and sensor state. | Agents & orchestrationHarnesses & orchestration | |
| s18github.com | 7 | Open-source agent runtime and orchestration API where query routing, skill matching and planner fast paths can use Jev decisions, falling back to built-in regex guards when Jev is unkeyed, unreachable or unsure. | Agents & orchestrationHarnesses & orchestration | |
| sdkgithub.com | 8 | AbdelStark/bicameral — Hybrid coding harness: System 2 writes, System 1 (Jev) runs reflexes. · sdk · TypeScript | Agents & orchestrationHarnesses & orchestration | |
| sdkgithub.com | 3 | aaronshaf/opencode-jev-orchestrator — Keeps OpenCode on a cheap sticky model for warm cache; Jev escalates hard turns to stronger subagents. · sdk · TypeScript | Agents & orchestrationHarnesses & orchestration | |
| Soothe TypeSafe decisionsgithub.com | — | Goal-driven orchestration framework for long-running agents that classifies intake intent and audits goal completion with typed decisions, plus a benchmark against hosted Jev, Laya or NanoJev servers. | Agents & orchestrationHarnesses & orchestration | |
| Soothe TypeSafe decisions — appmirasoth.github.io | — | Goal-driven orchestration framework for long-running agents that classifies intake intent and audits goal completion with typed decisions, plus a benchmark against hosted Jev, Laya or NanoJev servers. | Agents & orchestrationHarnesses & orchestration | |
| Soothe TypeSafe decisions — repogithub.com | — | Goal-driven orchestration framework for long-running agents that classifies intake intent and audits goal completion with typed decisions, plus a benchmark against hosted Jev, Laya or NanoJev servers. | Agents & orchestrationHarnesses & orchestration | |
| super-jevgithub.com | 13 | Domain-independent TypeScript harness that links evidence, Jev judgments, permitted tool actions and verified outcomes, with pluggable data sources and tools, argument checks and local JSONL traces. | Agents & orchestrationHarnesses & orchestration | |
| System One Harnessgithub.com | 95 | Python controller that turns a System One model into an agent loop, compiling an environment's finite action space into typed Jev questions, gating each step by confidence, and recording the full trace. | Agents & orchestrationHarnesses & orchestration | |
| System One Harness — discussionnews.ycombinator.com | — | Python controller that turns a System One model into an agent loop, compiling an environment's finite action space into typed Jev questions, gating each step by confidence, and recording the full trace. | Agents & orchestrationHarnesses & orchestration | |
| Tidepoolgithub.com | 12 | Agent harness built on a live Haskell notebook where agents write procedures that mix commands, delegation and Jev judgments, then install them as tools or hooks; 322 questions returned in 351 ms. | Agents & orchestrationHarnesses & orchestration | |
| yunusemrejr/yunuspigithub.com | 2 | Yunus Pi is a heavily customized harness experience built on top of a forked Pi harness (pi.dev). It uses patches, extensions, skills, and custom provider configs to customize the overall experience. It uses advanced machine learning, Jev, Cactus Needle 3 and similar technologies. | Agents & orchestrationHarnesses & orchestration |
Tool selection & execution83 of 83 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| @maximem/jev-mcpwww.npmjs.com | — | MCP server giving Claude Code, Codex and Cowork Jev-backed classify, score, check and rank tools, plus plan, action and verify gates over the agent's own work. | Agents & orchestrationTool selection & execution | |
| actionreflexgithub.com | 0 | A pre-execution gate for AI agent actions, powered by TypeSafe's Jev (System One) model. | Agents & orchestrationTool selection & execution | |
| Agent-native tool prefetchgithub.com | — | Ranks up to 128 tools and skills with a single Choice and preloads the top three within 750 ms. | Agents & orchestrationTool selection & execution | |
| Agent-native tool prefetch — appwww.agent-native.com | — | Ranks up to 128 tools and skills with a single Choice and preloads the top three within 750 ms. | Agents & orchestrationTool selection & execution | |
| ai_sdkgithub.com | 4 | NicolasMontone/jev-tool-permissions — Jev-backed tool approval gate and tool-list pruning for the Vercel AI SDK · ai_sdk · TypeScript | Agents & orchestrationTool selection & execution | |
| alexsatch/omp-auto-modegithub.com | 2 | Plugin for oh-my-pi that uses Typesafe Jev API to classify tool calls as safe/unsafe/ask | Agents & orchestrationTool selection & execution | |
| antigravity-approvalgithub.com | 0 | Fail-closed Antigravity PreToolUse guardrail using deterministic rules and TypeSafe Jev to screen tool calls, block risky commands, and escalate ambiguous actions to manual approval. | Agents & orchestrationTool selection & execution | |
| clownware/bouncergithub.com | 1 | Jev-powered Judgment layer for Claude Code. Stops paying reasoning prices for if-statements: a PreToolUse hook scores every tool call against a YAML policy you own — allow / deny in ~100 ms, no LLM in the loop. TypeSafe Jev now, local models next. Dry-run by default, calibration table published. | Agents & orchestrationTool selection & execution | |
| decisions-judge-mcpgithub.com | 4 | Typed decisions for AI agents as an MCP tool: yes/no probability (noul), choice, and score in one fast request. Backed by the TypeSafe System One model. | Agents & orchestrationTool selection & execution | |
| Dehydratorgithub.com | 10 | Client-side tool search for LLM APIs that lets a model use thousands of tools without sending every definition, matching tools with BM25, Jev, or both. | Agents & orchestrationTool selection & execution | |
| dsh-jev-interceptorgithub.com | 22 | ⚡ Millisecond System-1 judgement for every tool call in DeepSeek Harness — Jev-powered risk classification & evidence-gated auto-approval. Fail-closed by construction. dsh 生态第一个 System-1 决策插件 | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 2 | RiskAverseTech/toolgate — Open auto mode for AI agents — a calibrated tool-call firewall powered by TypeSafe Jev. Ships as a Claude Code hook · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 1 | clownware/bouncer — Jev-powered Judgment layer for Claude Code. Stops paying reasoning prices for if-statements: a PreToolUse hook scores every tool call against a YAML policy you own — allow / deny in ~100 ms, no LLM in the loop. TypeSafe Jev now, local models next. Dry-run by default, calibration table published. · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 1 | getexcited/stepwarden — Every tool call your agent makes, checked before it runs. A Claude Code plugin that uses TypeSafe AI's Jev to verify each pending tool call against the session plan, then allows it, asks you, or blocks it. Proof of concept · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 2 | masa-med-ai/typesafe-screening-mcp — MCP server: screen PubMed titles/abstracts against a query or clinical question with TypeSafe Jev · endpoint · Python | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 2 | ndolinschi/toolgate — Agent tool/MCP call gate — allow / ask\_human / deny via TypeSafe Jev · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 7 | rashedInt32/jev-mcp — MCP server exposing TypeSafe Jev as typed, calibrated judgment tools: classify, score, check, batched ask. Ships as a Claude Code plugin. · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 3 | teempai/jev-in-codex — Jev-powered tool and skill selection, context search, and output triage for Codex via MCP · endpoint · TypeScript | Agents & orchestrationTool selection & execution | |
| endpointgithub.com | 60 | yikangy873-gif/jev-desktop — TypeSafe Jev action selection inside Codex Computer Use · endpoint · JavaScript | Agents & orchestrationTool selection & execution | |
| FastMCP Jev tool searchgithub.com | — | Experimental FastMCP transform that ranks a server's tool catalog with Jev: a wide Choice over one-line summaries, then a close read of a shortlist with full descriptions. | Agents & orchestrationTool selection & execution | |
| FastMCP Jev tool search — docsgofastmcp.com | — | Experimental FastMCP transform that ranks a server's tool catalog with Jev: a wide Choice over one-line summaries, then a close read of a shortlist with full descriptions. | Agents & orchestrationTool selection & execution | |
| finch-jevgithub.com | — | Mini tool for the Finch desktop agent that registers a finch_jev_evaluate tool, letting the agent send text plus Choice, Score and Noul questions to Jev. | Agents & orchestrationTool selection & execution | |
| finch-jev — appwww.finchwork.app | — | Mini tool for the Finch desktop agent that registers a finch_jev_evaluate tool, letting the agent send text plus Choice, Score and Noul questions to Jev. | Agents & orchestrationTool selection & execution | |
| Hive Jev routing backendgithub.com | — | Jev backend for Hive, a CPU-side action router for AI agents: Jev picks the next tool and rates whether it is safe to run without the LLM. A calibration probe found the default 0.95 safety gate never fires. | Agents & orchestrationTool selection & execution | |
| Hive Jev routing backend — benchmarkgithub.com | — | Jev backend for Hive, a CPU-side action router for AI agents: Jev picks the next tool and rates whether it is safe to run without the LLM. A calibration probe found the default 0.95 safety gate never fires. | Agents & orchestrationTool selection & execution | |
| Hive Jev routing backend — repogithub.com | — | Jev backend for Hive, a CPU-side action router for AI agents: Jev picks the next tool and rates whether it is safe to run without the LLM. A calibration probe found the default 0.95 safety gate never fires. | Agents & orchestrationTool selection & execution | |
| JCR (Jev Capability Resolver)github.com | 19 | Resolver that gives an agent one tool to find documented deterministic commands for a task, using Jev to search a nested capability tree and return only the selected operations' context. | Agents & orchestrationTool selection & execution | |
| JCR (Jev Capability Resolver) — sitejcr.niazmorshed.dev | — | Resolver that gives an agent one tool to find documented deterministic commands for a task, using Jev to search a nested capability tree and return only the selected operations' context. | Agents & orchestrationTool selection & execution | |
| Jev MCP servergithub.com | — | Zero-dependency Python MCP server bundled with the Narwhal Cloud container watcher that gives agents such as Claude Code, Cursor or Antigravity jev_choice, jev_noul, jev_score, jev_gate and jev_batch tools. | Agents & orchestrationTool selection & execution | |
| Jev MCP server — repogithub.com | — | Zero-dependency Python MCP server bundled with the Narwhal Cloud container watcher that gives agents such as Claude Code, Cursor or Antigravity jev_choice, jev_noul, jev_score, jev_gate and jev_batch tools. | Agents & orchestrationTool selection & execution | |
| Jev x402 tool selectionx.com | — | Live demo from Tenjin of Jev choosing which paid x402 endpoint to call from a curated set of tools. | Agents & orchestrationTool selection & execution | |
| jev-agent-authorizationgithub.com | 5 | Jev agent authorization for MCP tool calls: Kinde identity and permissions plus Jev's typed, calibrated decisions, checked server-side before every call runs | Agents & orchestrationTool selection & execution | |
| jev-dshgithub.com | 0 | System One (TypeSafe Jev) decision tool for DeepSeek Harness — bounded choice / noul / score judgments as an agent tool + skill. | Agents & orchestrationTool selection & execution | |
| jev-mcpgithub.com | 25 | Python MCP server that gives Claude Code, Codex, Cursor or any MCP client typed Jev classify, score, check, match and screen tools, each answer returned with a confidence. | Agents & orchestrationTool selection & execution | |
| jev-mcpgithub.com | 7 | MCP server and Claude Code plugin exposing Jev as classify, score, check and batched ask tools that return the typed answer with its full probability distribution. | Agents & orchestrationTool selection & execution | |
| jev-mcp — demowww.reddit.com | — | MCP server that gives agents such as Claude Code and OpenCode typed Jev tools: jev_evaluate for a map of parallel questions, plus jev_noul, jev_choice, jev_score, and jev_models. | Agents & orchestrationTool selection & execution | |
| jev-mcp — demox.com | — | Ten MCP judgment tools for agents, including claim verification, injection screening, reranking, and gating. | Agents & orchestrationTool selection & execution | |
| jev-mcp — discussionwww.reddit.com | — | MCP server that gives agents such as Claude Code and OpenCode typed Jev tools: jev_evaluate for a map of parallel questions, plus jev_noul, jev_choice, jev_score, and jev_models. | Agents & orchestrationTool selection & execution | |
| jev-mcp — npmwww.npmjs.com | — | MCP server and Claude Code plugin exposing Jev as classify, score, check and batched ask tools that return the typed answer with its full probability distribution. | Agents & orchestrationTool selection & execution | |
| jev-mcp — write-upx.com | — | MCP server and Claude Code plugin exposing Jev as classify, score, check and batched ask tools that return the typed answer with its full probability distribution. | Agents & orchestrationTool selection & execution | |
| jev-mcp-dispatchergithub.com | 3 | Natural-language MCP tool dispatcher powered entirely by TypeSafe's Jev — no general-purpose LLM. Discovers a simple MCP server's tool signatures at runtime and uses Jev's typed primitives (Choice/Noul) to pick the right tool and extract its arguments straight out of the sentence. | Agents & orchestrationTool selection & execution | |
| jev-mcp-middlewaregithub.com | 0 | Semantic MCP middleware that intercepts AI-agent tool calls and combines deterministic policy with TypeSafe Jev risk judgments before actions reach upstream MCP servers. | Agents & orchestrationTool selection & execution | |
| jev-tool-permissionsgithub.com | 4 | Tool-permission layer for the Vercel AI SDK that uses Jev to auto-approve, ask a human or block each tool call and to prune irrelevant tool definitions before a turn. | Agents & orchestrationTool selection & execution | |
| jev-tool-routergithub.com | 7 | MCP router for Codex that exposes a small routing surface instead of hundreds of tool schemas, with Jev picking one external tool from names and descriptions or returning a shortlist when unsure. | Agents & orchestrationTool selection & execution | |
| jev-toolspacegithub.com | — | Context GC: Jev answers yes/no (\noul\) questions about a shared state and returns an independent probability for each one. One API call scores every tool in a menu with one question per tool, and a tool's score does not compete with the others. | Agents & orchestrationTool selection & execution | |
| jev2mcpgithub.com | 0 | A small intelligence layer for context-aware MCP, plugin, and tool selection. Built with Jev. | Agents & orchestrationTool selection & execution | |
| jevselectorgithub.com | 3 | OpenCode v2 plugin that uses Jev (TypeSafe AI) to dynamically select relevant skills and tools per turn, reducing context window usage. | Agents & orchestrationTool selection & execution | |
| legacybridge-tech/pi-typesafe-jevgithub.com | 3 | A pi extension that exposes TypeSafe (Jev, System One) judgments as five pi tools, so a model can make narrow semantic judgments while your code and your users keep control of thresholds, weights, and actions. | Agents & orchestrationTool selection & execution | |
| loom-skill-selectorgithub.com | 0 | Proactive Jev-powered skill selection for Loom, running as a jailed extension | Agents & orchestrationTool selection & execution | |
| masa-med-ai/typesafe-screening-mcpgithub.com | 2 | MCP server: screen PubMed titles/abstracts against a query or clinical question with TypeSafe Jev | Agents & orchestrationTool selection & execution | |
| Mimir TypeSafe skill and tool selectiongithub.com | — | TypeSafe integration in Mimir, a Rust recursive-language-model agent runtime, that selects the applicable skill and shortlists needed tools before a model call, rolled out through shadow, canary and acceptance benchmarks. | Agents & orchestrationTool selection & execution | |
| Mimir TypeSafe skill and tool selection — repogithub.com | — | TypeSafe integration in Mimir, a Rust recursive-language-model agent runtime, that selects the applicable skill and shortlists needed tools before a model call, rolled out through shadow, canary and acceptance benchmarks. | Agents & orchestrationTool selection & execution | |
| opencode-langsearchgithub.com | 0 | Unofficial OpenCode v2 plugin: LangSearch web search provider, optionally gated by Jev (TypeSafe AI) to cut tokens and filter noisy/injected results before they reach the model. | Agents & orchestrationTool selection & execution | |
| OpenHuman Jev tool rankergithub.com | — | Tool-search ranker for the OpenHuman agent harness: embedding search shortlists 20 tools and one Jev request picks one or none and checks whether a tool is needed; without an embedding provider it skips Jev for BM25. | Agents & orchestrationTool selection & execution | |
| OpenHuman Jev tool ranker — apptinyhumans.ai | — | Tool-search ranker for the OpenHuman agent harness: embedding search shortlists 20 tools and one Jev request picks one or none and checks whether a tool is needed; without an embedding provider it skips Jev for BM25. | Agents & orchestrationTool selection & execution | |
| Paprwork jev_decide toolgithub.com | — | Mastra tool in the Paprwork local-first desktop agent app that lets agents, sub-agents and scheduled jobs ask Jev typed noul, choice or score questions, with input guardrails and Papr-proxy or bring-your-own-key auth. | Agents & orchestrationTool selection & execution | |
| Paprwork jev_decide tool — appwww.papr.ai | — | Mastra tool in the Paprwork local-first desktop agent app that lets agents, sub-agents and scheduled jobs ask Jev typed noul, choice or score questions, with input guardrails and Papr-proxy or bring-your-own-key auth. | Agents & orchestrationTool selection & execution | |
| Paprwork jev_decide tool — repogithub.com | — | Mastra tool in the Paprwork local-first desktop agent app that lets agents, sub-agents and scheduled jobs ask Jev typed noul, choice or score questions, with input guardrails and Papr-proxy or bring-your-own-key auth. | Agents & orchestrationTool selection & execution | |
| permit-mcpgithub.com | 1 | Stop AI agents from executing MCP tools unchecked — policy, approval, JEV decisions, and one-use execution permits. | Agents & orchestrationTool selection & execution | |
| pi-jev-skill-pickergithub.com | 36 | Pi extension that removes the roughly 19,000-token skill catalog from the system prompt and adds a skill_search tool in which Jev rates every enabled skill against the task and returns the matching SKILL.md files. | Agents & orchestrationTool selection & execution | |
| Ratify Jev tool-selection adaptergithub.com | — | Reference LangChain gate for the Ratify delegated-authority protocol where Jev proposes which tool an agent should call and Ratify then verifies the signed authority on the receiver side. | Agents & orchestrationTool selection & execution | |
| Ratify Jev tool-selection adapter — appratifyprotocol.com | — | Reference LangChain gate for the Ratify delegated-authority protocol where Jev proposes which tool an agent should call and Ratify then verifies the signed authority on the receiver side. | Agents & orchestrationTool selection & execution | |
| Ratify Jev tool-selection adapter — repogithub.com | — | Reference LangChain gate for the Ratify delegated-authority protocol where Jev proposes which tool an agent should call and Ratify then verifies the signed authority on the receiver side. | Agents & orchestrationTool selection & execution | |
| sdkgithub.com | 3 | abhishekashokvkumar/jev-mcp-dispatcher — Natural-language MCP tool dispatcher powered entirely by TypeSafe's Jev — no general-purpose LLM. Discovers a simple MCP server's tool signatures at runtime and uses Jev's typed primitives (Choice/Noul) to pick the right tool and extract its arguments straight out of the sentence. · sdk · Python | Agents & orchestrationTool selection & execution | |
| sdkgithub.com | 2 | alexsatch/omp-auto-mode — Plugin for oh-my-pi that uses Typesafe Jev API to classify tool calls as safe/unsafe/ask · sdk · TypeScript | Agents & orchestrationTool selection & execution | |
| SilkChat Jev skill selectiongithub.com | — | Jev integration in SilkChat, a TanStack Start and Convex AI chat app, that picks which skills to enable on a conversation's opening turn: web search, code execution, memory, math or image generation. | Agents & orchestrationTool selection & execution | |
| SilkChat Jev skill selection — appsilkchat.dev | — | Jev integration in SilkChat, a TanStack Start and Convex AI chat app, that picks which skills to enable on a conversation's opening turn: web search, code execution, memory, math or image generation. | Agents & orchestrationTool selection & execution | |
| SilkChat Jev skill selection — repogithub.com | — | Jev integration in SilkChat, a TanStack Start and Convex AI chat app, that picks which skills to enable on a conversation's opening turn: web search, code execution, memory, math or image generation. | Agents & orchestrationTool selection & execution | |
| SREGym Jev decision toolgithub.com | ≈305 | Opt-in Codex MCP tools for an SRE agent benchmark where Jev picks the most informative read-only diagnostic test and reviews Kubernetes evidence before a diagnosis or fix is submitted. | Agents & orchestrationTool selection & execution | |
| SREGym Jev decision tool — appsregym.com | — | Opt-in Codex MCP tools for an SRE agent benchmark where Jev picks the most informative read-only diagnostic test and reviews Kubernetes evidence before a diagnosis or fix is submitted. | Agents & orchestrationTool selection & execution | |
| stepwardengithub.com | — | Every tool call your agent makes, checked before it runs. A Claude Code plugin that uses TypeSafe AI's Jev to verify each pending tool call against the session plan, then allows it, asks you, or blocks it. Proof of concept | Agents & orchestrationTool selection & execution | |
| stepwardengithub.com | 1 | Every tool call your agent makes, checked before it runs. A Claude Code plugin that uses TypeSafe AI's Jev to verify each pending tool call against the session plan, then allows it, asks you, or blocks it. Proof of concept | Agents & orchestrationTool selection & execution | |
| Synkora TypeSafe toolsgithub.com | — | Built-in Jev evaluation tools for agents on the self-hosted Synkora platform: generic state-and-question calls, named profiles such as recruiting, a report formatter and a playground. | Agents & orchestrationTool selection & execution | |
| Synkora TypeSafe tools — appsynkora.ai | — | Built-in Jev evaluation tools for agents on the self-hosted Synkora platform: generic state-and-question calls, named profiles such as recruiting, a report formatter and a playground. | Agents & orchestrationTool selection & execution | |
| tool-prunegithub.com | 5 | Zero-dependency tool selection and schema pruning for agents that cuts MCP tool schemas to the relevant few before an LLM call, running offline via TurboQuant or on Jev. | Agents & orchestrationTool selection & execution | |
| toolgategithub.com | 2 | Tool-call firewall that runs as a Claude Code PreToolUse hook or an MCP proxy and asks Jev seven questions, such as destructive, exfiltration or off-task, before risky actions, with YAML policy and a local audit log. | Agents & orchestrationTool selection & execution | |
| ToolGategithub.com | 2 | Gate for planned agent tool or MCP calls that returns allow, ask a human or deny, with a five-level risk score and checks for data exfiltration, irreversibility and policy violations. | Agents & orchestrationTool selection & execution | |
| ToolGate — apptoolgate.vercel.app | — | Gate for planned agent tool or MCP calls that returns allow, ask a human or deny, with a five-level risk score and checks for data exfiltration, irreversibility and policy violations. | Agents & orchestrationTool selection & execution | |
| VisionClaw Jev validationgithub.com | — | Bounded, advisory-only Jev client in the VisionClaw self-hosted multi-tenant agent workspace, exposed as an agent tool and used to score deliverables for evidence support, requirement coverage and need for human review. | Agents & orchestrationTool selection & execution | |
| VisionClaw Jev validation — appagenticcorporation.net | — | Bounded, advisory-only Jev client in the VisionClaw self-hosted multi-tenant agent workspace, exposed as an agent tool and used to score deliverables for evidence support, requirement coverage and need for human review. | Agents & orchestrationTool selection & execution | |
| VisionClaw Jev validation — repogithub.com | — | Bounded, advisory-only Jev client in the VisionClaw self-hosted multi-tenant agent workspace, exposed as an agent tool and used to score deliverables for evidence support, requirement coverage and need for human review. | Agents & orchestrationTool selection & execution | |
| vit-cerny/jev-browser-kitgithub.com | 1 | Give any MCP-capable LLM harness an on-demand real-browser search tool (TypeSafe Jev) with per-run timing and cost tracking. | Agents & orchestrationTool selection & execution | |
| wahahaorg/jev-mcpgithub.com | 1 | jev-mcp turns browser-use/jev-ultrafast into one local, session-aware MCP server for fast read-only product research. Jev selects an action and observed target from a DOM snapshot; | Agents & orchestrationTool selection & execution |
Autonomous assistants62 of 62 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| agent-stardewgithub.com | 9 | Stardew Valley agent where Jev picks and executes the next farm action from dynamic candidates toward a user goal, via an SMAPI mod, a standalone CLI and a DeepSeek Harness plugin with a run panel. | Agents & orchestrationAutonomous assistants | |
| Annagithub.com | 149 | Governed local-first enterprise agent whose Crew feature suggests a person or worker for an unassigned task with Jev, which can abstain; it matched a DeepSeek judge on 22/22 cases at 85.23% lower p50 latency. | Agents & orchestrationAutonomous assistants | |
| Autonomous PS2 AI Agentgithub.com | 0 | PlayStation 2 emulator agent that captures frames, extracts scene telemetry and asks Jev a Choice (accelerate, steer, brake, attack) that is injected as virtual controller input, with a live telemetry HUD. | Agents & orchestrationAutonomous assistants | |
| browserjevgithub.com | 0 | Low-memory autonomous web classification with HTTPX, Lightpanda and Jev | Agents & orchestrationAutonomous assistants | |
| demo-video-javascript-jev-startergithub.com | 0 | This application is intended to showcase to developers a possible use case of Vonage Video API and TypeSafe AI to create an AI support agent that quickly detects if a user is getting frustrated and start a video call with a Human agent. | Agents & orchestrationAutonomous assistants | |
| Diana TypeSafe clientgithub.com | — | Self-hosted group-chat agent for QQ, Telegram, DingTalk, Feishu and WeCom that can use Jev to decide whether to chime in and whether a message is addressed to the bot. | Agents & orchestrationAutonomous assistants | |
| Diana TypeSafe client — appsuink.github.io | — | Self-hosted group-chat agent for QQ, Telegram, DingTalk, Feishu and WeCom that can use Jev to decide whether to chime in and whether a message is addressed to the bot. | Agents & orchestrationAutonomous assistants | |
| Diana TypeSafe client — repogithub.com | — | Self-hosted group-chat agent for QQ, Telegram, DingTalk, Feishu and WeCom that can use Jev to decide whether to chime in and whether a message is addressed to the bot. | Agents & orchestrationAutonomous assistants | |
| Discriminative MCTS Agentgithub.com | 7 | Autonomous agent that runs Monte Carlo Tree Search over actions proposed by harnesses like AGY or Pi, with a D3.js visualizer for replaying rollouts and value backpropagation. | Agents & orchestrationAutonomous assistants | |
| doormangithub.com | 0 | Bot and pitch detection for risky forms with one typed question to Jev. Zero dependencies, mock mode without a key, recorded live run included. | Agents & orchestrationAutonomous assistants | |
| duomindgithub.com | 16 | When Jev meets LLM -- Pair a small local LLM (System 2) with Jev model (System 1) to make small model more faster and accurate — OpenAI-compatible, private, and almost free to run in Kilo code and Cline. | Agents & orchestrationAutonomous assistants | |
| endpointgithub.com | 2 | luxus/ha-conversation-jev — Home Assistant custom component: Conversation agent with Jev fast-path + Grok fallback · endpoint · Python | Agents & orchestrationAutonomous assistants | |
| endpointgithub.com | 1 | rahulpalivela18/jiffy — general Chrome agent — Jev + OpenRouter · endpoint · Python | Agents & orchestrationAutonomous assistants | |
| endpointgithub.com | 4 | tonyzdev/pijev — PiJev: a terminal coding agent with Jev in the loop — Jev ranks the repository's files before the first call, picks skills and triages failures; your coding model writes the code. Built on Pi. · endpoint · TypeScript | Agents & orchestrationAutonomous assistants | |
| fable-jevgithub.com | 9 | ⚡ Sub-100ms cognitive reflexes for autonomous coding agents. Powered by TypeSafe AI's Jev & get-fable. | Agents & orchestrationAutonomous assistants | |
| Footworkgithub.com | 2 | A verified browser agent: a cheap Jev guard (evidence-checked completions, a destructive gate) in front of any LLM browser driver, with Jev taking the mechanical steps in dual mode. Built on browser-use; every number pre-registered and measured. | Agents & orchestrationAutonomous assistants | |
| ha-jev-conversationgithub.com | 0 | Czech Home Assistant Assist agent on TypeSafe Jev: typed decisions in ~300 ms, confirmations for locks and garage | Agents & orchestrationAutonomous assistants | |
| homeconsole.typesafe.ai | — | jev-decision-maker-at-meteora - Decision Maker at Meteora - a decision agent for Meteora DLMM liquidity on Solana. Named after Jev, TypeSafe AI's System One model \(https://console.typesafe.ai/home\), which answers its typed Noul and Score questions. Ranks the live pool universe, runs live rug cards, and holds 3-5 pools under explicit hard gates and book rules. · Python | Agents & orchestrationAutonomous assistants | |
| jeffreygithub.com | 2 | A coding agent CLI where Jev (TypeSafe System One) or Laya decide what to do next and a configurable LLM does the work. | Agents & orchestrationAutonomous assistants | |
| Jev BTC 5m Enginegithub.com | 2 | Trading agent for Polymarket's 5-minute Bitcoin Up or Down markets where code prices each window from the Chainlink feed and Jev acts as a sanity gate and veto, with maker orders and paper mode by default. | Agents & orchestrationAutonomous assistants | |
| Jev Market Reflexgithub.com | 3 | Feeds live BTC, ETH and SOL market data from Kraken to Jev as compact state, turns it into typed BUY, SELL or HOLD decisions and applies them to a simulated paper portfolio. | Agents & orchestrationAutonomous assistants | |
| jev-botgithub.com | 9 | JEV-powered market decision bot for stocks, crypto and memes. State in, BUY/SELL/HOLD/AVOID out, paper by default | Agents & orchestrationAutonomous assistants | |
| jev-botgithub.com | 1 | Computer Use Agent developed with Jev | Agents & orchestrationAutonomous assistants | |
| jev-brotatogithub.com | 0 | Let TypeSafe Jev play Brotato: local game bridge, autonomous menus, and recorded runs with probabilities, latency, tokens and cost. | Agents & orchestrationAutonomous assistants | |
| jev-canvasgithub.com | 2 | Browser demo where you say commands like 'make a yellow circle here' while pointing at the webcam, and Jev decides action, shape, colour, target and place on a tldraw canvas in 300 to 550 ms per decision. | Agents & orchestrationAutonomous assistants | |
| jev-claude-demogithub.com | 0 | This is a small Spring Boot service built around one idea: Jev decides, Claude writes | Agents & orchestrationAutonomous assistants | |
| jev-decision-maker-at-meteoragithub.com | 0 | Decision Maker at Meteora - a decision agent for Meteora DLMM liquidity on Solana. Named after Jev, TypeSafe AI's System One model (https://console.typesafe.ai/home), which answers its typed Noul and Score questions. Ranks the live pool universe, runs live rug cards, and holds 3-5 pools under explicit hard gates and book rules. | Agents & orchestrationAutonomous assistants | |
| jev-factorio-agentgithub.com | 1 | Jev picks what, code owns how - a System One Factorio agent driven by TypeSafe's Jev on FLE | Agents & orchestrationAutonomous assistants | |
| jev-on-airgithub.com | 0 | JevOnAir monitors livestreams in real time, transcribes speech, detects market-moving statements with Jev, and turns them into structured simulated trade signals. | Agents & orchestrationAutonomous assistants | |
| jev-playsgithub.com | 0 | Autonomous game agent powered by TypeSafe AI's Jev (System One decision engine) | Agents & orchestrationAutonomous assistants | |
| jev-plays-pokemongithub.com | 0 | Pokemon-playing agent in the style of Claude Plays Pokemon where Jev chooses the actions, with a local viewer that streams the game video beside a live decision sidebar. | Agents & orchestrationAutonomous assistants | |
| jev-plays-super-mario-brosgithub.com | 0 | AI agent that plays Super Mario Bros with TypeSafe Jev - a typed decision model reading NES RAM instead of pixels, verified by emulator rollback search. Clears 1-1 to 1-4 and 2-1 for ~$0.02. | Agents & orchestrationAutonomous assistants | |
| jev-terrarium-dinosaursgithub.com | 0 | Autonomous dinosaur ecosystem where code runs the physics and TypeSafe's Jev decides each creature's intent from prose laws. Pixel-art web UI with a live Jev monitor (JSON in/out per tick), interchangeable brains (hybrid / full Jev / offline). Experimental Jev integration. | Agents & orchestrationAutonomous assistants | |
| JEV-Trading-BOTgithub.com | 1 | An order-flow trading bot whose decisions are made by Jev | Agents & orchestrationAutonomous assistants | |
| jev-uipath-coded-agentgithub.com | 2 | FINS demo: a UiPath coded agent for AML alert triage where every decision is made by TypeSafe's Jev model (Noul/Score/Choice) instead of an LLM | Agents & orchestrationAutonomous assistants | |
| jev-voice-agentgithub.com | 0 | A realtime voice agent with no LLM in the loop: Deepgram Flux ends the turn, TypeSafe's Jev picks the reply, Deepgram Flux TTS speaks it. | Agents & orchestrationAutonomous assistants | |
| jevmegithub.com | 5 | Talk to your Mac and it does it — in any app, while you're still talking. Open-source voice agent that gets faster the more you use it. | Agents & orchestrationAutonomous assistants | |
| kmua bot Jev follow-up checkgithub.com | — | Telegram bot agent that experimentally uses Jev to decide whether a follow-up group message is relevant enough for the bot to keep replying, with a configurable threshold. | Agents & orchestrationAutonomous assistants | |
| laya-browser-agentgithub.com | 16 | Local, open-source Jev alternative: browser agent decisions with Laya (System One model) on your own machine. No cloud, no API key. Playwright/CDP, MCP-friendly. | Agents & orchestrationAutonomous assistants | |
| lightjevgithub.com | 0 | Read pages with Lightpanda, decide where to go with Jev — a self-steering crawler for tiny budgets | Agents & orchestrationAutonomous assistants | |
| LogiPulseAI_JEVgithub.com | 0 | Real-time last-mile delivery exception triage: Jev (TypeSafe AI System One) evaluates each event, deterministic guardrails decide. FastAPI + Streamlit. | Agents & orchestrationAutonomous assistants | |
| model_idgithub.com | 9 | imMamdouhaboammar/fable-jev — ⚡ Sub-100ms cognitive reflexes for autonomous coding agents. Powered by TypeSafe AI's Jev & get-fable. · model_id · TypeScript | Agents & orchestrationAutonomous assistants | |
| model_idgithub.com | 9 | rottenpen/agent-stardew — 星露谷农场小助手：Jev 自主游玩、dsh 插件、独立 CLI 与 SMAPI Mod · model_id · TypeScript | Agents & orchestrationAutonomous assistants | |
| NyatBot Jev guardsgithub.com | — | Telegram group-chat agent that uses Jev to catch paraphrased repeats of its own messages and to block hallucinated facts that nobody in the chat mentioned or asked about. | Agents & orchestrationAutonomous assistants | |
| NyatBot Jev guards — repogithub.com | — | Telegram group-chat agent that uses Jev to catch paraphrased repeats of its own messages and to block hallucinated facts that nobody in the chat mentioned or asked about. | Agents & orchestrationAutonomous assistants | |
| PetGPT Jev social signalsgithub.com | — | AI desktop pet with a social agent in QQ, Telegram and WhatsApp groups that uses Jev to tag message mood and read the group's atmosphere before replying. | Agents & orchestrationAutonomous assistants | |
| PetGPT Jev social signals — sandboxgithub.com | — | AI desktop pet with a social agent in QQ, Telegram and WhatsApp groups that uses Jev to tag message mood and read the group's atmosphere before replying. | Agents & orchestrationAutonomous assistants | |
| Phantombotgithub.com | 16 | Persistent-identity AI assistant around terminal harnesses with long-term memory across PhantomChat, Telegram and editors; an optional Jev screener acts as a threat judge and routes each turn to the primary or coder model in ~300 ms. | Agents & orchestrationAutonomous assistants | |
| Phantombot — appphantombot.bot | — | Persistent-identity AI assistant around terminal harnesses with long-term memory across PhantomChat, Telegram and editors; an optional Jev screener acts as a threat judge and routes each turn to the primary or coder model in ~300 ms. | Agents & orchestrationAutonomous assistants | |
| pi-ask-jeffgithub.com | 0 | An advisor called Jeff for pi: calibrated decisions via ask_jeff, backed by the Jev-compatible System One API (OpenRouter / TypeSafe / jev-agent) | Agents & orchestrationAutonomous assistants | |
| pi-jevgithub.com | 11 | Pi coding-agent extension where Jev picks the file excerpts to send and a local model such as Qwen or DeepSeek writes the code, with exact-result caching to cut end-to-end latency. | Agents & orchestrationAutonomous assistants | |
| pokemon-showdown-jev-agentgithub.com | 0 | Autonomous Pokemon Showdown battle agent powered by Jev AI (System One Model by TypeSafe AI), with a validated poke-env harness, safe fallback decisions, and a live dashboard for transparent battle state, legal moves, probabilities, telemetry, and Jev decision history. | Agents & orchestrationAutonomous assistants | |
| rahulpalivela18/jiffygithub.com | 1 | general Chrome agent — Jev + OpenRouter | Agents & orchestrationAutonomous assistants | |
| relay-macos-assistantgithub.com | 0 | Experimental local-first macOS voice automation using whisper.cpp, TypeSafe Jev routing, typed actions, and optional Ollama generation. | Agents & orchestrationAutonomous assistants | |
| risk-analysis-servergithub.com | 2 | An autonomous risk-monitoring MCP server. Define a risk profile — a topic, the geographic locations to watch, and the policy triggers that matter — and the server runs an agentic sweep: live web search scoped by a judgment model (Jev), source scoring, and a Markdown briefing you can read from any MCP client. | Agents & orchestrationAutonomous assistants | |
| rockygithub.com | 0 | Say it, and your computer does it. Voice-controlled computer agent on TypeSafe Jev for macOS and Windows. | Agents & orchestrationAutonomous assistants | |
| smart-home-assistantgithub.com | 0 | Demo app to showcase the System One models functionality. Say what you want in natural language, System decides what to do on the home.. | Agents & orchestrationAutonomous assistants | |
| Starlight dialogue continuationgithub.com | — | Telegram bot that asks Jev, for each group message, whether to reply with text, react with an emoji, or stay silent, and which emoji fits the message. | Agents & orchestrationAutonomous assistants | |
| Starlight dialogue continuation — appstarlight.day | — | Telegram bot that asks Jev, for each group message, whether to reply with text, react with an emoji, or stay silent, and which emoji fits the message. | Agents & orchestrationAutonomous assistants | |
| Starlight dialogue continuation — repogithub.com | — | Telegram bot that asks Jev, for each group message, whether to reply with text, react with an emoji, or stay silent, and which emoji fits the message. | Agents & orchestrationAutonomous assistants | |
| system1-agentsgithub.com | 136 | System 1 decision models (Jev, Laya, Cua-S1) as brain for agents: Browser use, computer use, games and robotics | Agents & orchestrationAutonomous assistants | |
| typesafe-minesweepergithub.com | 0 | Autonomous Minesweeper powered by TypeSafe System One and local constraint inference | Agents & orchestrationAutonomous assistants |
More projects & source code959 of 959 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| 2023Anita/codex-jev-assistantgithub.com | 1 | Codex Plugin with typed Jev judgments for risk review, evidence checks, context screening, and reranking. | Agents & orchestrationMore projects & source code | |
| 24601/Augustusgithub.com | 10 | Agent skills for designing, training, evaluating and improving application-specific decision systems. Primitive/model selection, data assembly, export/reload and bounded hill climbing. TypeSafe Jev is the default hosted exemplar; independent of TypeSafe. | Agents & orchestrationMore projects & source code | |
| a0-typesafe-ai — READMEgithub.com | 5 | A community Agent Zero plugin adds Jev probability-card judgments through the official TypeSafe Python SDK and bundles the official agent skill; it is not an official product. | Agents & orchestrationMore projects & source code | |
| abmind System One judgmentsgithub.com | — | Pluggable System One judgment provider in abmind, a persistent memory and RAG engine for AI agents, that sends recall and diagnostics questions to Jev or a locally served Laya model as advice to its combining code. | Agents & orchestrationMore projects & source code | |
| abmind System One judgments — repogithub.com | — | Pluggable System One judgment provider in abmind, a persistent memory and RAG engine for AI agents, that sends recall and diagnostics questions to Jev or a locally served Laya model as advice to its combining code. | Agents & orchestrationMore projects & source code | |
| actiongate-jev — READMEgithub.com | 2 | ActionGate combines deterministic policy with TypeSafe Jev to authorize exact tool actions; its README says no external security review has occurred. | Agents & orchestrationMore projects & source code | |
| ActionJevgithub.com | 1 | Structured code review for GitHub and Gitea Actions, powered by TypeSafe Jev and written in Rust. | Agents & orchestrationMore projects & source code | |
| Agent Handoff Gategithub.com | 0 | Experimental protocol for delegated coding work that checks the evidence behind a worker agent's PASS or BLOCKED report before it reaches the lead agent, with schemas, a Jev integration contract and a benchmark. | Agents & orchestrationMore projects & source code | |
| Agent! Jev advisorgithub.com | — | Jev advisor in Agent!, a native macOS agent app, that gives a second opinion on shell commands already passed by pattern rules and refuses those whose data-destruction probability exceeds a user-set threshold. | Agents & orchestrationMore projects & source code | |
| Agent! Jev advisor — repogithub.com | — | Jev advisor in Agent!, a native macOS agent app, that gives a second opinion on shell commands already passed by pattern rules and refuses those whose data-destruction probability exceeds a user-set threshold. | Agents & orchestrationMore projects & source code | |
| Agent! Jev advisor — sdkgithub.com | — | Jev advisor in Agent!, a native macOS agent app, that gives a second opinion on shell commands already passed by pattern rules and refuses those whose data-destruction probability exceeds a user-set threshold. | Agents & orchestrationMore projects & source code | |
| agent-battle-royalegithub.com | 0 | Interactive agent-routing demo. The API key and Sites project id stay in local env files. | Agents & orchestrationMore projects & source code | |
| agent-battle-royale — READMEgithub.com | 0 | Agent Battle Royale routes a prompt to a specialist selected by Jev; without an API key, the page uses a backup referee instead of Jev. | Agents & orchestrationMore projects & source code | |
| agent-chaperonegithub.com | 21 | Firewall that screens an AI agent's tool calls before they run and tool results before the agent reads them, via an MCP proxy and a hooks adapter, with thresholds in a policy file and a non-blocking shadow mode. | Agents & orchestrationMore projects & source code | |
| agent-chaperone — READMEgithub.com | 21 | Agent tool-call and result screening combines deterministic policy rules with Jev typed judgments; it is not a sandbox or security guarantee. | Agents & orchestrationMore projects & source code | |
| Agent-Defensegithub.com | 0 | A Jev-powered security guard for AI agents: blocks dangerous tool calls, strips prompt injection, scans skills. Works with Claude Code, Codex, Copilot CLI, Gemini CLI, Cursor, and OpenCode. | Agents & orchestrationMore projects & source code | |
| Agent-Defense — READMEgithub.com | 0 | AgentDefense is documented as an AI-agent security guard that feeds narrow Jev judgments into explicit local rules; its README says Jev does not make the final decision. These are project claims, not a security audit. | Agents & orchestrationMore projects & source code | |
| agent-fastpathgithub.com | 3 | Jev MCP server: a decision layer for coding agents, built on TypeSafe Jev (System One model). Ship gates, risk checks, file triage that keeps files out of context, and a safe headless browser, with calibrated confidence. For Claude Code, Codex, Cursor. | Agents & orchestrationMore projects & source code | |
| agent-fastpath — READMEgithub.com | 3 | Agent Fastpath is a Jev-backed MCP decision layer for coding agents, documenting action gates, risk checks, file triage, and browser verification. | Agents & orchestrationMore projects & source code | |
| agent-handoff-gate — READMEgithub.com | 0 | This experimental offline specification documents a TypeSafe/Jev integration contract for checking whether a worker’s evidence supports a task before handoff; it is not a runtime adapter. | Agents & orchestrationMore projects & source code | |
| agent-harness-with-jev-llm — READMEgithub.com | 1 | This Jev-pattern harness demonstrates routing and risky-tool gating, but its local classifier is a TF-IDF/cosine-similarity substitute rather than TypeSafe’s trained Jev model. | Agents & orchestrationMore projects & source code | |
| agent-kitgithub.com | 1 | Agent make dozens of silent judgments a turn: is this done, is this correct, do I have enough to decide, which way now. Made inside your own text stream, those judgments are uncalibrated and biased toward finishing, you tend to certify your own work. judge hands one such judgment to Jev. | Agents & orchestrationMore projects & source code | |
| agent-kit — READMEgithub.com | 1 | An oh-my-pi skill and extension bundle delegates bounded completion and UI judgments to Jev; unauthenticated skills can instead use a small chat model. | Agents & orchestrationMore projects & source code | |
| agent-reflexgithub.com | 0 | A ~400ms System-One safety reflex for AI coding agents: auto-approves routine tool calls, escalates risky ones, blocks exfiltration & prompt injection. Powered by TypeSafe's Jev. | Agents & orchestrationMore projects & source code | |
| agent-reflex — READMEgithub.com | 0 | Coding-agent permission layer uses Jev risk and context signals with deterministic allow/ask/deny policy; unreachable Jev defers to the agent’s normal permission behavior. | Agents & orchestrationMore projects & source code | |
| agent-router — READMEgithub.com | 100 | A pre-release CLI for Herdr that filters coding-agent subscriptions by deterministic quota rules, then asks TypeSafe System One to rank eligible options and choose reasoning effort before launching an agent. Production code constructs the TypeSafe SDK client; task text is sent to TypeSafe, so “local-first” does not mean offline routing. | Agents & orchestrationMore projects & source code | |
| agent-router — typesafe-client.tsgithub.com | 100 | A pre-release CLI for Herdr that filters coding-agent subscriptions by deterministic quota rules, then asks TypeSafe System One to rank eligible options and choose reasoning effort before launching an agent. Production code constructs the TypeSafe SDK client; task text is sent to TypeSafe, so “local-first” does not mean offline routing. | Agents & orchestrationMore projects & source code | |
| agent-tool-routergithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| agent-tool-router — READMEgithub.com | 1 | A TypeScript agent-tool decision SDK lets Jev select a tool while deterministic policy code decides whether execution is permitted. | Agents & orchestrationMore projects & source code | |
| agent0github.com | 5 | Desktop AI agent runtime with MCP integration, task planning and persistent memory where optional Jev routing selects the relevant tools, falling back to the main model on low confidence, with each decision shown in an Activity panel. | Agents & orchestrationMore projects & source code | |
| AgentBeam Jev action judging — repogithub.com | 11 | Optional check in the AgentBeam local security CLI for coding agents that sends redacted proposed tool actions to Jev and allows, reviews or denies them in observe or enforce mode. | Agents & orchestrationMore projects & source code | |
| AgenticX Jev group routinggithub.com | — | Multi-agent platform whose group chat routes un-mentioned messages with Jev, dispatching by label when confident and falling back to an LLM, and shows a decision card each turn. | Agents & orchestrationMore projects & source code | |
| AgentX decision backendsgithub.com | — | Self-hosted AI agent mesh for small businesses whose decision layer can run Jev-style typed decisions on a local LLM or a simple-jev server, recomputing confidence as normalized entropy. | Agents & orchestrationMore projects & source code | |
| AgentX decision backends — repogithub.com | — | Self-hosted AI agent mesh for small businesses whose decision layer can run Jev-style typed decisions on a local LLM or a simple-jev server, recomputing confidence as normalized entropy. | Agents & orchestrationMore projects & source code | |
| ai-csuite Jev judgmentsgithub.com | — | AI C-Suite strategic debate skill that can use Jev to classify the decision topic and judge the executives' round-one positions, falling back to keyword logic without a key. | Agents & orchestrationMore projects & source code | |
| ai-csuite Jev judgments — repogithub.com | — | AI C-Suite strategic debate skill that can use Jev to classify the decision topic and judge the executives' round-one positions, falling back to keyword logic without a key. | Agents & orchestrationMore projects & source code | |
| ai_sdkgithub.com | 1 | paulobueno164/jev-mcts — Busca em arvore (MCTS/PUCT) com avaliacao tipada do TypeSafe Jev e portao humano obrigatorio. A arvore supoe, a sonda mede: marco so e concedido por codigo de saida verde. · ai_sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| ai_sdkgithub.com | 16 | reachjalil/jevlogs — Open-source Jev log triage for OpenTelemetry. Score the signal before expensive LLM analysis. · ai_sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| ailerix — READMEgithub.com | 2 | Ailerix routes requests by having Jev classify task families, then uses a cost-per-task frontier to select a provider; provider completions remain mocked without provider keys. | Agents & orchestrationMore projects & source code | |
| AIMEAT decidegithub.com | 8 | Federated, self-hosted AI operating system whose nodes carry Jev beside the text model, so apps and agents can ask closed triage, routing, scoring and pick-one questions via a JS library, MCP tools and a skill. | Agents & orchestrationMore projects & source code | |
| AIQSA Jev decision featuresgithub.com | — | Jev integration in AIQSA, a self-hosted multi-provider AI workspace, that uses decision calls via OpenRouter for memory and knowledge-base relevance, tool discovery and skill suggestions, each enabled after its own qualification. | Agents & orchestrationMore projects & source code | |
| AIQSA Jev decision features — repogithub.com | — | Jev integration in AIQSA, a self-hosted multi-provider AI workspace, that uses decision calls via OpenRouter for memory and knowledge-base relevance, tool discovery and skill suggestions, each enabled after its own qualification. | Agents & orchestrationMore projects & source code | |
| Aivory decision policiesgithub.com | — | Self-hosted AI chat platform that lets admins assign Jev to file routing, tool routing, memory dedup, memory conflict adjudication and content moderation, each with confidence fallbacks. | Agents & orchestrationMore projects & source code | |
| Aivory decision policies — docsgithub.com | — | Self-hosted AI chat platform that lets admins assign Jev to file routing, tool routing, memory dedup, memory conflict adjudication and content moderation, each with confidence fallbacks. | Agents & orchestrationMore projects & source code | |
| AIWG Jev decision adaptergithub.com | — | Decision framework in an AI-augmented dev toolkit where versioned rulesets and bindings let workflows switch between Jev and an LLM subagent, with a hardened Jev transport. | Agents & orchestrationMore projects & source code | |
| AIWG Jev decision adapter — docsgithub.com | — | Decision framework in an AI-augmented dev toolkit where versioned rulesets and bindings let workflows switch between Jev and an LLM subagent, with a hardened Jev transport. | Agents & orchestrationMore projects & source code | |
| AmberAgent Jev runtimegithub.com | — | Jev runtime in the AmberAgent Android AI workspace used for bounded phone-screen and web goals, model-council routing, memory reranking, semantic tool search and trimming long tool outputs before they reach the model. | Agents & orchestrationMore projects & source code | |
| AmberAgent Jev runtime — repogithub.com | — | Jev runtime in the AmberAgent Android AI workspace used for bounded phone-screen and web goals, model-council routing, memory reranking, semantic tool search and trimming long tool outputs before they reach the model. | Agents & orchestrationMore projects & source code | |
| android-jevgithub.com | 1 | MCP server and skill that let an agent drive an Android phone naturally over adb | Agents & orchestrationMore projects & source code | |
| android-jev — READMEgithub.com | 1 | An MCP Android-control tool reads device state over adb and uses Jev to choose subsequent actions; documented device verification is limited, not general autonomy assurance. | Agents & orchestrationMore projects & source code | |
| antigravity-approval — READMEgithub.com | 0 | Antigravity PreToolUse guard applies deterministic rules before Jev classification for ambiguous calls, failing to manual approval on uncertainty/errors; host permission prompts may still appear after allow decisions. | Agents & orchestrationMore projects & source code | |
| approval-judge-bridgegithub.com | 0 | OpenAI-compatible judge endpoint for agent approval gates: typed judgements (Jev), any OpenAI-compatible model, or a rule file — fail-closed, calibrated, with a replay battery | Agents & orchestrationMore projects & source code | |
| approval-judge-bridge — READMEgithub.com | 0 | Approval-guardian bridge adapts Jev typed judgments to APPROVE, DENY or ESCALATE for agent commands, escalating when it cannot judge safely. | Agents & orchestrationMore projects & source code | |
| arbitypegithub.com | 0 | Typed decision tools for AI agents, powered by TypeSafe Jev | Agents & orchestrationMore projects & source code | |
| arbitype — READMEgithub.com | 0 | An MCP-native typed decision layer for agents, powered by TypeSafe Jev, with tools for classifying, scoring, verifying, gating, routing, and reviewing. | Agents & orchestrationMore projects & source code | |
| Arbos Jev routergithub.com | — | Structured router in the Arbos file-system-native agent coordinator where Jev picks the next mechanical tool move from a situation card, while the LLM handles writing, planning and talking; act=llm hands the turn to the chat model. | Agents & orchestrationMore projects & source code | |
| Arbos Jev router — repogithub.com | — | Structured router in the Arbos file-system-native agent coordinator where Jev picks the next mechanical tool move from a situation card, while the LLM handles writing, planning and talking; act=llm hands the turn to the chat model. | Agents & orchestrationMore projects & source code | |
| ask-jev — READMEgithub.com | 1 | ask-jev is a standard-library CLI that delegates bounded semantic questions to TypeSafe System One while leaving reasoning and action authority with the calling agent. | Agents & orchestrationMore projects & source code | |
| ask-jev-skill — READMEgithub.com | 41 | The Hermes skill calls TypeSafe Jev as a typed tiebreaker and documents Choice, Score, and Noul answers for agent decisions or escalation. | Agents & orchestrationMore projects & source code | |
| askjevgithub.com | 1 | Unofficial MCP server, local over stdio or hosted on Cloudflare Workers, where an agent asks a plain question about material it holds and Jev picks yes/no, scale or choice and answers with calibrated probabilities. | Agents & orchestrationMore projects & source code | |
| askjevgithub.com | 41 | Hermes skill, usable by other agents, that calls Jev as a typed tiebreaker so the agent can act on a confident Choice, Score, or Noul answer or escalate. | Agents & orchestrationMore projects & source code | |
| astra-jev-harness — READMEgithub.com | 3 | Astra provides a Codex/Claude Code context-selection skill and harness: it narrows file candidates locally, Jev selects what the coding model should read, and that separate model writes code. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_intentiontriggergithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_intentiontrigger — READMEgithub.com | 0 | AstrBot plugin uses Jev intent classification to route group-chat messages to the main LLM only when they seek interaction with the bot. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_jev_gategithub.com | 0 | AstrBot 主动回复门控插件：用 TypeSafe Jev 模型给群聊消息打分，只在真值得接话时才触发一次主动回复 | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_jev_gate — READMEgithub.com | 0 | AstrBot group-chat plugin uses Jev to score whether a message deserves an unsolicited reply; AstrBot's main pipeline generates the reply. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_jev_radargithub.com | 0 | astrbot\_plugin\_jev\_radar - Jev 意图雷达 — 用 TypeSafe SystemOne\(Jev\) 读懂消息背后的意图、情绪与风险 · Python | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_jev_radar — READMEgithub.com | 0 | This AstrBot plugin uses Jev to classify chat-message intent and risk; the README says enabled message text is sent to TypeSafe for judgment. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_reply_gategithub.com | 2 | Decides whether a message should reach the LLM at all, using TypeSafe Jev choice probabilities. | Agents & orchestrationMore projects & source code | |
| astrbot_plugin_reply_gate — READMEgithub.com | 2 | An AstrBot plugin uses Jev Choice probabilities to decide whether a message should reach the LLM; rejected messages are stopped before the LLM request. | Agents & orchestrationMore projects & source code | |
| Augustusgithub.com | 10 | Agent skill: design judgment-assisted systems with TypeSafe Jev (System One). Maps Choice/Score/Noul onto decision theory, reranking, and routing. Composition algebra, question design, validation gates. MIT. | Agents & orchestrationMore projects & source code | |
| Augustus — READMEgithub.com | 10 | Augustus is an independent decision-model design and evaluation toolkit that lists TypeSafe Jev as its default hosted exemplar alongside other model families. | Agents & orchestrationMore projects & source code | |
| AutoGPT TypeSafe blocksgithub.com | ≈187,500 | Seven no-code blocks, including a five-exit router, a yes, no, or unsure split, and a score filter. | Agents & orchestrationMore projects & source code | |
| axiom-agent-runtimegithub.com | 1 | Reference prototype exploring agentic commerce: TypeSafe/Jev → Axiom → Argent → Silverscript on Kaspa. | Agents & orchestrationMore projects & source code | |
| azdajagithub.com | 13 | Recursive language model layer for Claude Code, Codex, Gemini, and OpenCode that keeps full sources in a local evaluator; Jev is an optional leaf for reranking, verification, classification, and semantic joins, with budgeted, checkpointed batches. | Agents & orchestrationMore projects & source code | |
| backoffice-agentsgithub.com | 1 | Piloto de agentes de backoffice: LLM gera e raciocina, Jev (TypeSafe AI) decide, roteia e verifica | Agents & orchestrationMore projects & source code | |
| backoffice-agents — READMEgithub.com | 1 | This back-office agent pilot separates roles: an LLM generates and reasons, while Jev makes routing and verification decisions, including safety gates. | Agents & orchestrationMore projects & source code | |
| beadsortgithub.com | 6 | The intelligence layer for beads: typed, calibrated labels for your backlog | Agents & orchestrationMore projects & source code | |
| beadsort — READMEgithub.com | 6 | beadsort uses System One judgments to label Beads work items and surface agent-ready tasks separately from decisions that still need the user. | Agents & orchestrationMore projects & source code | |
| BerriAI/litellmgithub.com | 59,200 | The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (or native) format with cost tracking, guardrails, load balancing, and logging [Bedrock, Azure, OpenAI, Anthropic, OpenAI, VertexAI, vLLM, Nvidia NIM] | Agents & orchestrationMore projects & source code | |
| bicameral — READMEgithub.com | 8 | A Pi coding harness separates code generation by a configurable LLM from Jev's non-generative System 1 scoring; the README explicitly says it is not a sandbox. | Agents & orchestrationMore projects & source code | |
| BlueNoise Jev second pass — repogithub.com | — | Experimental Jev second pass in BlueNoise, a privacy-first extension that blurs or hides noisy X replies with local rules, judging replies the rules missed with your own TypeSafe key. | Agents & orchestrationMore projects & source code | |
| bright-sightgithub.com | 0 | macOS 命令行工具：把一句中文指令变成受控的应用操作，每一步执行完都由代码回读验证 | Agents & orchestrationMore projects & source code | |
| bright-sight — READMEgithub.com | 0 | macOS CLI turns Chinese requests into allowlisted Chrome-tab and Notes operations, using typed System One choices while code controls and verifies execution. | Agents & orchestrationMore projects & source code | |
| BrowserClawgithub.com | 5 | Chrome MV3 extension and native-messaging MCP server that lets AI agents drive your everyday logged-in Chrome, including a chrome_act_toward_goal tool where a Jev micro-loop picks actions over a pruned DOM tree. | Agents & orchestrationMore projects & source code | |
| browserjev — READMEgithub.com | 0 | A non-generative browser agent accepts a domain and classification schema, explores the site, gathers evidence, and uses Jev typed questions to return structured decisions. | Agents & orchestrationMore projects & source code | |
| building-with-jev-skillgithub.com | 147 | A skill for writing and improving programs that call Jev, TypeSafe's System One model. | Agents & orchestrationMore projects & source code | |
| Calleegithub.com | 74 | Markdown- and YAML-defined agent workflows for ACP runtimes that mix coding models, shell checks and human steps with a native TypeSafeJev evaluator for typed decisions and routing. | Agents & orchestrationMore projects & source code | |
| Callee — docsgithub.com | 74 | Markdown- and YAML-defined agent workflows for ACP runtimes that mix coding models, shell checks and human steps with a native TypeSafeJev evaluator for typed decisions and routing. | Agents & orchestrationMore projects & source code | |
| Cannygithub.com | 106 | Claude Code and Codex CLI hooks that keep an append-only ledger of edits and checks and refuse a "done" claim when no check has passed since the last code edit. | Agents & orchestrationMore projects & source code | |
| Cascade-Agentgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Cascade-Agent — READMEgithub.com | 0 | Small agent loop uses Jev to route execution tiers and answer typed steps, while documenting that its output-verification gate poorly separates good and bad results. | Agents & orchestrationMore projects & source code | |
| ChainlessChain skill decision layergithub.com | — | Skill decision layer in a personal AI management CLI that can ask Jev which skill should handle a request, with an 800 ms timeout, metered usage and a benchmark harness. | Agents & orchestrationMore projects & source code | |
| ChainlessChain skill decision layer — repogithub.com | — | Skill decision layer in a personal AI management CLI that can ask Jev which skill should handle a request, with an 800 ms timeout, metered usage and a benchmark harness. | Agents & orchestrationMore projects & source code | |
| ChipJevgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| ChipJev — READMEgithub.com | 0 | Circuit-design agent applies Jev's System One approach using ChipLaya, a Laya fine-tune, to seed topology/sizing search and open-source EDA checks; it is not a hosted Jev client. | Agents & orchestrationMore projects & source code | |
| choola — READMEgithub.com | 0 | Choola adopts Jev-style typed Choice/Score/Noul questions, but its README says local Laya answers them first and uncertain cases go to a Claude teacher. | Agents & orchestrationMore projects & source code | |
| chrome_jev_agentgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| chrome_jev_agent — READMEgithub.com | 0 | The Chrome agent builds a finite option table in code for each round and asks Jev to choose from it before browser actions. | Agents & orchestrationMore projects & source code | |
| claude-code-jevgithub.com | 4 | Experimental PreToolUse hook that runs Claude Code's allow/block/ask permission gate on Jev via OpenRouter, measured at 264 ms and $0.0000227 per call with 0 dangerous actions allowed on an 18-case fixture. | Agents & orchestrationMore projects & source code | |
| claude-jevgithub.com | 18 | Claude Code plugin: Jev for rule checks, verbatim compaction, and prompt routing | Agents & orchestrationMore projects & source code | |
| claude-jevgithub.com | 7 | Claude Code plugin that runs Claude's review findings, debugging hypotheses and design options past Jev as an independent judge, dropping false positives below a probability threshold. | Agents & orchestrationMore projects & source code | |
| claude-jevgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| claude-jev — READMEgithub.com | 0 | A Claude Code proxy uses Jev to route each prompt to a suitable model tier; the README warns it launches Claude with permission prompts skipped. | Agents & orchestrationMore projects & source code | |
| claude-router — READMEgithub.com | 0 | Claude Code routing gateway uses Jev to advise a model tier and effort for each incoming prompt. | Agents & orchestrationMore projects & source code | |
| clay-jev-people-rankergithub.com | 2 | Agent skill and Python workflow that pulls people from Clay CLI with deterministic filters, then has Jev judge whether each person really fits the target role, such as current operating founders, before enrichment. | Agents & orchestrationMore projects & source code | |
| clay-jev-people-ranker — READMEgithub.com | 2 | The Clay workflow retrieves leads with deterministic filters and uses TypeSafe JEV for semantic founder-fit qualification, leaving retrieval to Clay. | Agents & orchestrationMore projects & source code | |
| clear-headgithub.com | 2 | Claude Code Stop hook that checks the factual claims in an assistant's answer against the tool output it actually read that session, and blocks the turn when claims are contradicted or unsupported. | Agents & orchestrationMore projects & source code | |
| codearia-sieve — relatedgithub.com | 472 | MCP server and TypeScript library that parses web pages into the state a decision model like Jev needs: dates as dates, numbers with units, chunks sized to fit; in its test Jev verified 11 of 11 extracted facts. | Agents & orchestrationMore projects & source code | |
| codex-jevgithub.com | 1 | Jev-powered selective context handoffs for Codex, with a GitHub Copilot portability plan. | Agents & orchestrationMore projects & source code | |
| codex-jev — READMEgithub.com | 0 | This Codex router sends each prompt through Jev to choose a model-effort tier; README says uncertain or unsupported inputs take a conservative fallback. | Agents & orchestrationMore projects & source code | |
| codex-jev — READMEgithub.com | 1 | An MCP plugin uses Jev to select verbatim conversation blocks for a next-topic handoff, preserving active history and retaining blocks when provider calls fail. | Agents & orchestrationMore projects & source code | |
| codex-jev-imagegengithub.com | 0 | Jev-powered decision workflows for Codex image generation. Explicit constraints, bounded repairs, and honest evaluation. | Agents & orchestrationMore projects & source code | |
| codex-jev-imagegen — READMEgithub.com | 0 | Jev ImageGen is a Codex skill that uses Jev to classify image requests while Codex writes prompts and inspects generated images. | Agents & orchestrationMore projects & source code | |
| codex-jev-router — READMEgithub.com | 4 | This repository documents a Codex subagent setup using JevRouter or another typed decision engine; it is setup guidance, not itself the router engine. | Agents & orchestrationMore projects & source code | |
| codex-subagent-router — READMEgithub.com | 0 | Codex spawn-agent hook lets Jev choose a configured task profile only when model, effort and agent type are not explicitly specified. | Agents & orchestrationMore projects & source code | |
| coding-orchestrator — READMEgithub.com | 0 | Claude Code/Codex orchestration bundle optionally uses Jev hooks for subagent routing, task-change detection, commit checks, report checks and handoff grading. | Agents & orchestrationMore projects & source code | |
| composio — READMEgithub.com | 30,386 | Composio's TypeSafe adapter turns tools into Jev questions, then passes Jev's decision to Composio for execution. | Agents & orchestrationMore projects & source code | |
| computer-use-cachegithub.com | 42 | OpenAI-compatible caching proxy that replays repeated computer-use and coding-agent requests. On an exact miss, Jev compares recent cached requests and picks a reusable response or none; anything uncertain goes upstream. | Agents & orchestrationMore projects & source code | |
| connectedGraph/claude-jev-wardengithub.com | 1 | Real-time quality gate and Art Director Warden for Claude Code powered by TypeSafe Jev 1.13 non-autoregressive decision model | Agents & orchestrationMore projects & source code | |
| CoWork OS TypeSafe decisionsgithub.com | — | CoWork OS, a local-first agentic everything-app, adds a TypeSafe decision provider so its agents can make typed judgments over /v1/systemone. | Agents & orchestrationMore projects & source code | |
| cribrixgithub.com | 1 | A precision-first RAG orchestrator: filters before it generates, verifies before it answers. | Agents & orchestrationMore projects & source code | |
| cribrix — READMEgithub.com | 1 | A RAG orchestrator using Jev/System One routing and evidence judgments before generation and answer verification, while a separate LLM writes text. | Agents & orchestrationMore projects & source code | |
| ctxjev — READMEgithub.com | 0 | ctxjev applies Jev relevance judgments to entries in an agent’s running context and prunes entries deemed less useful. | Agents & orchestrationMore projects & source code | |
| cursor-jevgithub.com | 0 | Jev for Cursor. Jev decides. The agent writes. The router is off until you turn it on. | Agents & orchestrationMore projects & source code | |
| cursor-jev — READMEgithub.com | 0 | This Cursor plugin sends typed questions to TypeSafe Jev for message sizing and routing; a slow request or missing key is skipped and the agent continues. | Agents & orchestrationMore projects & source code | |
| de-niji/jev-hermesgithub.com | 1 | Jev for Hermes: cheap intent gates + verbatim tool compaction on OpenRouter | Agents & orchestrationMore projects & source code | |
| decidegithub.com | 1 | Bulk decisions for AI agents. Jev classifies files and logs; your agent reviews exceptions. Reproducible cost and quality benchmarks. | Agents & orchestrationMore projects & source code | |
| decide — READMEgithub.com | 1 | This MCP tool applies Jev decisions in bulk to files, logs, or JSONL and sorts the resulting items. | Agents & orchestrationMore projects & source code | |
| decidexgithub.com | 1 | Industrial-grade Decision Foundation Layer for System One Game-Playing Models (Jev / NanoJev) | Agents & orchestrationMore projects & source code | |
| decidex — typesafe.pygithub.com | 1 | A structured-decision framework includes a TypeSafe Jev HTTP provider and game-oriented decision/gating examples; performance, calibration and safety claims remain unverified. | Agents & orchestrationMore projects & source code | |
| Decision Graph Protocolgithub.com | — | Open protocol for decision-based agents: apps expose immutable evidence frames, typed decisions and guarded actions, and a Jev adapter produces assessments while application code keeps control of effects. | Agents & orchestrationMore projects & source code | |
| decision-circuits — READMEgithub.com | 1 | A Python library builds deterministic gates over calibrated model answers and includes a System One backend; policy code owns threshold and escalation actions. | Agents & orchestrationMore projects & source code | |
| decision-firstgithub.com | 2 | Agent skill that makes a coding agent notice bounded-judgment steps (classify, triage, rank, route, verify), try Jev before a regex or LLM prompt, and record each attempt in a reusable lab directory. | Agents & orchestrationMore projects & source code | |
| Decision-Theoretic-Mixture-of-Agentsgithub.com | 0 | A fully audited, reproducible decision‑theoretic mixture‑of‑agents framework comparing multiple selector strategies, including OpenJev, with transparent calibration, limitations, and verification artifacts. | Agents & orchestrationMore projects & source code | |
| decisions-judge-mcp — READMEgithub.com | 4 | decisions-judge-mcp exposes typed System One decisions as MCP tools and currently uses TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| Deco Studio Jev task dedupegithub.com | — | Agent control plane whose task board first asks Jev whether incoming tasks duplicate existing cards, falling back to a fast LLM when Jev is unavailable or unsure. | Agents & orchestrationMore projects & source code | |
| DeepChat Jev judgment modelgithub.com | — | DeepChat's desktop agent client adds Jev as a provider plus an opt-in judgment-model slot that reviews tool-permission requests through the System One protocol. | Agents & orchestrationMore projects & source code | |
| DeepChat Jev judgment model — docsgithub.com | — | DeepChat's desktop agent client adds Jev as a provider plus an opt-in judgment-model slot that reviews tool-permission requests through the System One protocol. | Agents & orchestrationMore projects & source code | |
| deepdocgithub.com | 307 | Deep research tool for local knowledge base. | Agents & orchestrationMore projects & source code | |
| deepdoc — READMEgithub.com | 307 | A local-document research workflow using TypeSafe Jev to check and rerank Qdrant retrieval results and assess whether evidence covers required subsections. A separate generative LLM handles planning and report writing, so Jev is the evidence-decision layer rather than the prose generator. | Agents & orchestrationMore projects & source code | |
| deepseek-harness-jev-pre-compaction — READMEgithub.com | 5 | A DeepSeek Harness advisor uses Jev to prune low-value tool results before normal compaction; it is a companion, not a compaction backend or summarizer. | Agents & orchestrationMore projects & source code | |
| demo-video-javascript-jev-starter — READMEgithub.com | 0 | Vonage video-support demo uses Jev sentiment and frustration judgments to choose between AI chat and escalation to a human video call. | Agents & orchestrationMore projects & source code | |
| deslopgithub.com | 0 | Score web pages for ads, slop, SEO and second-hand content. An agent skill built on TypeSafe Jev: four probabilities per page, no verdict, the caller sets the thresholds. | Agents & orchestrationMore projects & source code | |
| deslop — READMEgithub.com | 0 | A Jev agent skill scores full web-page text across four categories and returns probabilities without a verdict; the caller sets thresholds, and short snippets are refused. | Agents & orchestrationMore projects & source code | |
| determinategithub.com | 6 | TypeScript agent library that treats LLMs as next-action predictors; its Jev decider picks the next valid tool and fills closed-set parameters, handing only free-text fields or low-confidence cases to the LLM. | Agents & orchestrationMore projects & source code | |
| DevMortimer/pi-wardengithub.com | 153 | Guardrails for Pi that steer instead of interrupt: enforces your project rules on every write, holds only hard-to-undo actions (3 per 1,000 calls, measured on real sessions), catches unverified "done" claims and stuck loops, and trims tool output. Built on pi-typesafe. | Agents & orchestrationMore projects & source code | |
| dgui-hypermem — READMEgithub.com | 2 | A Cloudflare Workers memory MCP server combining vector/full-text retrieval with optional TypeSafe Jev reranking and memory curation; Workers AI is a separate fallback. | Agents & orchestrationMore projects & source code | |
| dhava-gautama/noulgategithub.com | 1 | A gate for your agent's expensive steps, powered by TypeSafe Jev (System One). Offline-first, OpenRouter or direct, MIT. | Agents & orchestrationMore projects & source code | |
| Dicklesworthstone/skillrankergithub.com | 125 | Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context. Includes Claude Code hooks, structured JSON, abstention, and local feedback. Requires a TypeSafe API key. | Agents & orchestrationMore projects & source code | |
| docsignalsgithub.com | 4 | Ask typed questions of your documents. Get structured signals for agentic workflows, with the page each answer came from. | Agents & orchestrationMore projects & source code | |
| docsignals — READMEgithub.com | 4 | docsignals asks Jev questions about each page; the README's displayed sample output comes from an offline --fake stand-in, not live Jev. | Agents & orchestrationMore projects & source code | |
| doorman — READMEgithub.com | 0 | Form-submission demo uses Jev to distinguish real customers from bots or sales pitches and route to deliver, drop or human review; without a key it uses a labelled mock. | Agents & orchestrationMore projects & source code | |
| droidjevgithub.com | 0 | A fast, screenshot-free Android emulator clicker powered by TypeSafe's jev | Agents & orchestrationMore projects & source code | |
| droidjev — READMEgithub.com | 0 | A screenshot-free Android emulator CLI asks Jev which indexed UI element advances a goal, taps it through adb, and checks the resulting screen. | Agents & orchestrationMore projects & source code | |
| dsh-jevgithub.com | — | Context GC: This DeepSeek Harness plugin sends each round's latest user message to the systemone (Jev) API for emotion and intent classification and injects the result as plugin-sourced runtime context, with API Key and all parameters configured in the GUI settings page. | Agents & orchestrationMore projects & source code | |
| dsh-jevgithub.com | 24 | Decision layer for the DeepSeek Harness agent loop where Jev narrows the relevant tools before each step, gates risky tool calls, and picks a model route, and can never widen a permission. | Agents & orchestrationMore projects & source code | |
| dsh-jev — READMEgithub.com | 24 | DSH Jev documents a Jev decision layer for DeepSeek Harness, with mock-provider/shadow defaults that keep the default run offline and behavior-neutral. | Agents & orchestrationMore projects & source code | |
| dsh-jev — READMEgithub.com | 9 | dsh-jev pairs TypeSafe Jev with DeepSeek Harness; its README says end-to-end outcome gains have not been shown. | Agents & orchestrationMore projects & source code | |
| dsh-jev-interceptor — READMEgithub.com | 22 | A DeepSeek Harness plugin adds Jev risk and task-fit judgments at tool-call and recalled-message boundaries; its advertised figures are the project’s report, not independent measurements. | Agents & orchestrationMore projects & source code | |
| dsh-jev-router — READMEgithub.com | 0 | DeepSeek Harness router uses Jev to select reasoning strength and by default leaves the model and prompt prefix unchanged. | Agents & orchestrationMore projects & source code | |
| dsh-plugin-jevgithub.com | 5 | Plugin for the DeepSeek Harness web GUI that gives the main agent and subagents a jev_ask tool for calibrated judgments under uncertainty, with per-chat usage tracking and a settings tab for the key and model. | Agents & orchestrationMore projects & source code | |
| dsh-skill-auto-load-typesafegithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| dsh-skill-auto-load-typesafe — READMEgithub.com | 0 | DeepSeek Harness plugin uses Jev Noul applicability judgments to preload selected skill instructions into the same main-model request without changing its provider. | Agents & orchestrationMore projects & source code | |
| duomind — READMEgithub.com | 16 | Local llama.cpp generation server integrates cloud Jev classification to guide requests, sending prompts and context to TypeSafe for those decisions. | Agents & orchestrationMore projects & source code | |
| eachann1024/pi-jev-replygithub.com | 1 | Pi extension: clearer replies via Jev review + optional rewrite/visuals | Agents & orchestrationMore projects & source code | |
| EchoFlow Jev routergithub.com | — | Per-turn Jev router in the EchoFlow privacy-first Android AI chat app that decides whether to recall or skip memory, force web search, or save a memory before the answering model runs, for cloud chats only. | Agents & orchestrationMore projects & source code | |
| EchoFlow Jev router — repogithub.com | — | Per-turn Jev router in the EchoFlow privacy-first Android AI chat app that decides whether to recall or skip memory, force web search, or save a memory before the answering model runs, for cloud chats only. | Agents & orchestrationMore projects & source code | |
| EdgeJevgithub.com | 15 | Offline CPU runtime that converts open Jev reimplementations such as laya, kev, NanoJev and PlayJev to ONNX with INT8 quantization behind a protocol-compatible local server; 15.6 ms per question on 4 CPU cores. | Agents & orchestrationMore projects & source code | |
| edgejev — READMEgithub.com | 15 | EdgeJev packages local CPU inference for open Jev/System One reproductions, rather than the TypeSafe-hosted model. | Agents & orchestrationMore projects & source code | |
| effort-router — READMEgithub.com | 0 | Claude Code mod uses Jev by default to select per-turn reasoning effort, with shadow logging or enforced effort settings. | Agents & orchestrationMore projects & source code | |
| ego-jevgithub.com | 11 | Agent skill that gives the ego lite browser a Jev inner loop: each DOM step numbers the interactive elements and one Jev call of about 0.4 s picks the operation and target; logins and payments go to the planner. | Agents & orchestrationMore projects & source code | |
| ego-jev — READMEgithub.com | 11 | An ego lite browser skill combines operation and target selection in a typed Jev decision per DOM step; ego executes, while sensitive tasks are escalated. | Agents & orchestrationMore projects & source code | |
| elons-jobgithub.com | 11 | Local-first Chrome extension that hides high-confidence sexual and solicitation replies on X post pages behind reversible placeholders, with customizable rules, caching, and cost limits. | Agents & orchestrationMore projects & source code | |
| emirbartu/opencode-system-onegithub.com | 1 | Opencode plugin using Jev (system one model) as part of software development process. Not affiliated with Opencode team. | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | 24601/rh-guard — Reward-hack radar for coding agents: structural denies + TypeSafe Jev System One sidecar for Claude Code & Cursor hooks · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 5 | 3clyp50/a0-typesafe-ai — TypeSafe AI Jev judgments for Agent Zero, with typed tools and probability cards. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 5 | AkashPriyadarshii/jev-scout — Zero-hallucination open-source repo and crate scout powered by TypeSafe AI Jev System One scoring · endpoint · Rust | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 91 | AkashPriyadarshii/jev-seo — 100% free ₹0 agent-first SEO & GEO CLI suite and MCP server in Rust replacing Semrush and OpenSEO via DuckDuckGo and TypeSafe Jev System One · endpoint · Rust | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | BYK/jev-mcp — An eval-first MCP server for TypeSafe's Jev, a System One model that returns typed judgments (noul, choice, score) with probabilities instead of generated text. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 13 | BorisLeMeec/jev — A claude code plugin for jev · endpoint · Go | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 21 | Brainwires/jevwire — Jev decision layer for agents: MCP server, embeddable DecisionModel library, and an escalate-only Claude Code plugin (TypeSafe AI's Jev) · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 338 | Devin-AXIS/jev-dsh-decision — Jev DSH 决策引擎｜面向 Agent Harness 的结构化决策插件。原生支持 DeepSeek Harness，通过 iPolloWork 支持 OpenCode、Codex Harness。 · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 125 | Dicklesworthstone/skillranker — Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context. Includes Claude Code hooks, structured JSON, abstention, and local feedback. Requires a TypeSafe API key. · endpoint · Rust | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 62 | EliaAlberti/jev-rules — Jev picks which of your rules apply to each prompt, so Claude only sees the ones that matter. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | FFatTiger/new-api-plugin-typesafe — TypeSafe AI System One (Jev) task plugin for QuantumNous/new-api — native /v1/systemone, synchronous evaluation, token billing · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 34 | FrancoisChastel/jev-code — Jev, TypeSafe's System One classifier, as a tool inside Claude Code, Codex, Pi, and OpenCode: typed classify, check, score, rank, and ask, plus one-command setup. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 6 | HyunjunJeon/jev-judgment — Agent Skill: send closed coding-agent judgments to TypeSafe Jev · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 12 | HyunjunJeon/pi-quiet-ask — TypeSafe Jev as the pi coding agent's quiet decision layer · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | KamilPostrozny/pi-fast-jev-compaction — Fast JEV compaction extension for pi · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 4 | Kelbie/hunch — Semantic code review with Jev, plain-English rules and Agent Skills. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 230 | NiazMorshed2007/jev-review — Local-first MCP plugin for continuous software-quality review by AI coding agents, powered by Jev. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | Obrais-cloud/typesafe-mcp — MCP server exposing TypeSafe (Jev/System One) to the fleet: judge, rerank, systemone · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 7 | ShivamPansuriya/jev-skill-gate — Cut Claude Code's skill manifest by ~75% with TypeSafe Jev. Scores every installed skill for relevance and hides the rest via skillOverrides — 12,750 → 3,185 tokens on a 217-skill install, for $0.0009 a session. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | VladyslavHontar/clear-head — Claude Code Stop hook that checks an AI assistant's claims against what it actually read this session, using TypeSafe's Jev as the judge · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | aegsrl7/jevmap — Map a codebase into units and let Jev (TypeSafe AI) hand an AI coding agent the ten files that matter for a task · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 8 | ajensenwaud/hermes-jev-plugin — TypeSafe Jev (System One) decision tools for Hermes Agent: jev\_check / jev\_route / jev\_score / jev\_evaluate · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 36 | altryne/jevify — An agent skill to discover TypeSafe Jev opportunities, design typed questions, and learn from recent community experiments. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | axiomarchitecture/axiom-agent-runtime — Reference prototype exploring agentic commerce: TypeSafe/Jev → Axiom → Argent → Silverscript on Kaspa. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | carlsonchik/judging-with-typesafe — Скилл для агентов Letta: суждения по критериям через TypeSafe System One (Jev) · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | chrishan17/claude-jev-mod — Typed decisions in Claude Code: adds $.jev over TypeSafe's Jev, through OpenRouter, Vercel AI Gateway, Cloudflare Workers AI, LiteLLM or the TypeSafe API. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | christian-taillon/opencode-jev-compactor — Jev powered OpenCode compaction · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 5 | cyberofficial/dsh-plugin-jev — Agent-facing TypeSafe Jev (System One) for the DSH Web GUI. The model itself · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | dhava-gautama/noulgate — A gate for your agent's expensive steps, powered by TypeSafe Jev (System One). Offline-first, OpenRouter or direct, MIT. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | eachann1024/pi-jev-reply — Pi extension: clearer replies via Jev review + optional rewrite/visuals · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | felpsdev/jev-classifier — Local tool-routing classifier for coding agents, with a gateway, MCP integrations, and decision logs. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 12 | fornhere/hafiza-os — Linux, macOS ve Windows için kaynaklı ikinci beyin. Claude Code, Codex ve Antigravity adaptörleri; yerel Markdown kasa, ayrı hafıza incelemesi, isteğe bağlı Mem0/Jev. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | greenyamao/Antigravity-mcp-semantic-search-with-TypeSafeAi — Fast semantic code search & diff sanity auditor for AI coding assistants (Antigravity, Cursor, Claude Code) powered by TypeSafe System One. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | harrymunro/decision-first — Agent skill that spots bounded-judgment steps, tries a typed decision model (TypeSafe's Jev) first, and documents every attempt · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 12 | harshwasan/pi-jev-sentinel — Pi coding-agent extension: TypeSafe Jev checks for tool calls, tool outputs and replies (prompt injection, approvals, secret scrubbing, task pinning) · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 9 | iefnaf/pi-jev — Pi extension suite powered by Jev: selective context compaction and model routing · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | integrate-your-mind/jev-codex-plugin — Open-source Codex plugin for TypeSafe Jev decision consultation, failure diagnosis, and evidence-based completion review · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 339 | itsmostafa/typesafe-mcp — mcp connector to give your AI agent direct access to typesafe ai's jev model · endpoint · Go | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 18 | jiawei686/jev-ultrafast-mcp — Hand the browser work off: an MCP server where a decision model drives the page for your agent, so a flow costs one tool call instead of a turn per click. Ref-based element tables, code-checked assertions, zero-model macro replay. Speaks CDP to your Chrome. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | jonathanavis96/jev-kit — Everything you need to run TypeSafe's Jev with Claude Code: a tool-call guard, tier guard, file search, browser agent, review, belay, compaction and installers. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 46 | kbhuw/jev-sift — Classify first. Read selectively. A portable agent plugin and MCP tool for batch text classification. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 34 | keeltrace/hermes-jev — Typed System One decisions, ranking, verification, and an opt-in Hermes tool gate using TypeSafe Jev. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 957 | kerpopule/hermes-jev-skills — Jev-powered model routing, memory, compaction, skill selection, computer and browser use for Hermes agents (also Claude Code and Codex) · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | kindintelligence/jev-rust-review — Rust-aware code review for Claude Code and coding agents, powered by TypeSafe Jev · endpoint · Rust | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 255 | kitze/skillbox — Self-hosted, versioned skills library for AI agents. MCP, scoped clients, and optional Jev recommendations. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 3 | konstantinosbotonakis/codex-context-diet — Codex plugin: Jev-guided dieting of bulky tool results · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 50 | leepokai/jev-guard — Auto mode for every coding agent, built on Jev: risk-scores every tool call with session context (deny / ask / allow), flags prompt injection in results, checks skills and plugins. Claude Code, Codex, Copilot, Gemini, Cursor, pi, OpenCode, ACP. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 11 | madeye/pi-jev — Jev-assisted file retrieval and request caching for faster Pi workflows · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | mja00/cmdc-auto-mode — Auto permission mode for Command Code: screens every tool call with TypeSafe Jev before it runs, and rejects anything out of scope. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | mozbz/jev-review-hook — Jev powered code review hook for agents · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | muse0509/jev-preflight — A bounded Jev risk check for Claude Code: eight risk axes, one request, one optional reinspection. · endpoint · Go | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | mwijanarko1/pi-jev-agent — pi-jev-agent is an experimental extension for the Pi coding agent. It uses TypeSafe Jev to choose the next tool before each language-model step. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 5 | noplan-inc/limpet — A Stop hook that stops your coding agent from stopping too early. Plain-language rules, judged by jev. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 7 | okooo5km/jev — Typed decisions from the shell: an unofficial stdlib-Python CLI and Agent Skill for TypeSafe's Jev model, via the TypeSafe API (default) or OpenRouter. Yes/no, choice and ordinal scores with calibrated probabilities, semantic grep and batch mode. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | poponline63/hermes-jev-north-star — Hermes Agent skill whose north-star gate is judged by Jev (TypeSafe System One): turn an intention into a checkable finish line, generate the run prompt, and let Jev rank what is still unproven. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 106 | qkal/Canny — Stops AI coding agents from claiming work is done without evidence. Deterministic hooks decide, TypeSafe's Jev advises. Append-only ledger, zero runtime dependencies. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 41 | shantanugoel/ask-jev-skill — Skill for Hermes, and other agents, to ask typesafe's jev · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 39 | shitianfang/jev-use — Claude Code / Codex / pi plugin that hands agent steps needing no text output to Jev (TypeSafe's judgment model) — measured p50 ~230 ms and ~$0.02 per 1,000 judgments, with typed escalation back to the LLM · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 4 | simota/tenbin — MCP server and agent skill for the TypeSafe AI System One API (Jev): decompose a judgment into Choice / Score / Noul questions, lint them, measure on labelled data, and put calibrated thresholds in code · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 45 | tacticocc/Jevbridge — ACP and MCP adapter that bridges TypeSafe Jev with any LLM — computer use and typed decisions alongside Codex, Claude, Grok, and OpenCode. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 154 | tamaratran/jev-pruner — Claude Code plugin: trim long Bash output with TypeSafe Jev before the model sees it · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | thesammykins/jev_ampcode — A small advisory tool for comparing supplied alternatives against supplied evidence. · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | typakon4/jev-layer — Portable System-1 decision layer for agent harnesses with host-owned routing, receipts, replay, and fail-open integrations. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 16 | utk2103/jev-studio — if you're experimenting with jev it will be easier from here · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 19 | valentynkit/jev-belay — Claude Code Stop hook that blocks an unverified done: reads the transcript for evidence, asks Jev once, fails open on everything else · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 30 | win4r/jev-skill-suggester — 用 TypeSafe Jev 推荐已安装 Skill / Bounded installed-skill recommendations with TypeSafe Jev. Python CLI, Codex skill, bilingual docs and live examples. · endpoint · Python | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 152 | y0usaf/pi-jev — TypeSafe Jev as a decision layer for the Pi coding agent: a measured tool-call gate plus jev\_ask for typed, calibrated answers · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 2 | y9Finsi/jev-mcp — Universal Model Context Protocol (MCP) Server for TypeSafe Jev (System One) semantic code search and validation. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | zbush/jev-context — Codex code-search plugin using Jev relevance filtering with auditable token metrics · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 20 | zszz3/Pi-Jev-Guide — 原版 Pi Coding Agent 插件：按时机配置规则，并自带风险检查、输出脱敏、重复失败和缺少验证提醒。 · endpoint · TypeScript | Agents & orchestrationMore projects & source code | |
| endpointgithub.com | 1 | zurfyx/jev-browser-skill-demo — Let Jev, TypeSafe's ~100ms decision model, drive your browser. A plug-and-play skill for Claude Code and Codex. · endpoint · JavaScript | Agents & orchestrationMore projects & source code | |
| evokegithub.com | 22 | Software, by reflex. A sentence becomes a call of a small program, chosen by Jev, TypeSafe AI's classifier, and run only when it is sure enough. Reflexes are recipes anyone can write, share and improve. A CLI you talk to, a package manager for reflexes from git, and a TypeScript SDK. | Agents & orchestrationMore projects & source code | |
| evoke — READMEgithub.com | 22 | A CLI, package manager and TypeScript SDK choose and run installed small-program recipes with Jev; a confidence gate can confirm or ask instead of running. | Agents & orchestrationMore projects & source code | |
| exemplo-hev-roteamento — READMEgithub.com | 0 | Express incident-routing demo uses Jev through OpenRouter Decisions to choose the receiving backend team, while routeIncident performs the operational forwarding. | Agents & orchestrationMore projects & source code | |
| ExploitHunter Jev decision specialistgithub.com | — | Offensive-security research harness that uses Jev as a bounded decision specialist to pick tools and render formats, with a defer option and a persisted receipt for every call. | Agents & orchestrationMore projects & source code | |
| ExploitHunter Jev decision specialist — repogithub.com | — | Offensive-security research harness that uses Jev as a bounded decision specialist to pick tools and render formats, with a defer option and a persisted receipt for every call. | Agents & orchestrationMore projects & source code | |
| fable-jev — READMEgithub.com | 9 | Fable-Jev describes pairing TypeSafe AI's Jev with get-fable's deterministic lifecycle engine. | Agents & orchestrationMore projects & source code | |
| fabricioctelles/skillsgithub.com | 96 | A collection of skills for AI agents (Kiro, Cursor, Windsurf, Claude Code, and others). Each skill is a reusable module that teaches the agent to perform complex tasks with context, structure, and best practices. | Agents & orchestrationMore projects & source code | |
| Fake / Realgithub.com | 1 | Checks claims against cited excerpts with Jev in an English and Romanian fact-checking site, with a public integration example and a closed-source full application. | Agents & orchestrationMore projects & source code | |
| fast-jev (OpenCode plugin)github.com | 4 | OpenCode V2 plugin that uses Jev scores to drop stale tool calls and truncate bulky tool results on each outgoing model request, leaving stored history and /compact untouched and failing open. | Agents & orchestrationMore projects & source code | |
| Fast-Jev-Agents — READMEgithub.com | 2 | Fast-Jev-Agents describes verbatim context compaction: Jev scores historical tool calls alongside local heuristics, while obsolete results may be truncated and irrelevant calls pruned. | Agents & orchestrationMore projects & source code | |
| fast-jev-opencode — READMEgithub.com | 0 | OpenCode context-pruning plugin uses Jev to drop stale tool calls and truncate bulky outputs only in outgoing requests, preserving remaining wording and original history with fail-open behavior. | Agents & orchestrationMore projects & source code | |
| fastgate-jevgithub.com | 0 | Jev (TypeSafe AI) as a System One decision layer for a multilingual EN/UZ/RU RAG helpdesk, with an independent benchmark | Agents & orchestrationMore projects & source code | |
| fastgate-jev — READMEgithub.com | 0 | Multilingual university-helpdesk demo uses Jev for triage and grounded routing, with optional generative answers; without a key it uses an offline keyword mock rather than Jev judgments. | Agents & orchestrationMore projects & source code | |
| flowaidgithub.com | 0 | Open-source, backend-first AI agent and workflow platform with typed decision intelligence. | Agents & orchestrationMore projects & source code | |
| flowaid — READMEgithub.com | 0 | Beta agent/workflow runtime separates generative nodes from Jev routing, classification, risk scoring and approval gates, keeping decisions as typed inspectable data. | Agents & orchestrationMore projects & source code | |
| Footwork — READMEgithub.com | 2 | A browser/macOS computer-use harness placing TypeSafe Jev judgments over observed state and claimed completion. Included for its guard integration, not certification of verified operation or safety. | Agents & orchestrationMore projects & source code | |
| foreman — READMEgithub.com | 621 | Foreman places Jev above coding workers as a supervisory decision layer for assessing work state; the README describes this design, not an independently verified run. | Agents & orchestrationMore projects & source code | |
| forgegithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| forge — READMEgithub.com | 0 | A Rust coding agent defaults to an on-device Needle 3 decision model and consults Jev only as a conditional fallback when that local model is unsure. | Agents & orchestrationMore projects & source code | |
| frost — READMEgithub.com | 0 | Frost uses Jev to assess a task and recommend what to run, but its README says it does not launch the recommended model. | Agents & orchestrationMore projects & source code | |
| gevgithub.com | 1 | Jev from DiffusionGemma | Agents & orchestrationMore projects & source code | |
| gev — READMEgithub.com | 1 | gev is a separate GCP/Gemma service implementing a Jev-style typed-decision API; its README says it approximates the interface with token probabilities. | Agents & orchestrationMore projects & source code | |
| git-jev-stagegithub.com | 2 | Git CLI that classifies each changed hunk against a one-sentence description of what to stage, shows the plan and stages the selected hunks after confirmation. | Agents & orchestrationMore projects & source code | |
| GitHub-Issue-Classification-Using-Jevgithub.com | 0 | This repository contains a GitHub issue classifier built on Jev, TypeSafe AI's System One model. It labels every new issue with typed values and calibrated confidence in milliseconds, labelling what it is sure about and escalating what it is not. Three guardrail layers guard every write, and a frozen eval suite gates each deploy. | Agents & orchestrationMore projects & source code | |
| GitHub-Issue-Classification-Using-Jev — READMEgithub.com | 0 | This GitHub-issue classifier combines Jev’s typed risk and severity questions with application-side routing and abstention when uncertain; README-reported evaluation claims are not independently verified here. | Agents & orchestrationMore projects & source code | |
| gmail-triagegithub.com | 0 | Gmail トリアージ自動化 - Claude Code CLI でメールを分類し、不要メール削除・重要メール Discord 通知 | Agents & orchestrationMore projects & source code | |
| gmail-triage — READMEgithub.com | 0 | A Gmail triage tool uses Jev to classify email automatically. | Agents & orchestrationMore projects & source code | |
| grok-jev-router — READMEgithub.com | 1 | A Grok Bot decision router asks Jev before expensive or risky actions; the README says to begin in shadow mode and keeps human rules as the boundary. | Agents & orchestrationMore projects & source code | |
| ha-gutcheckgithub.com | 2 | Home Assistant integration that makes small judgment calls about your install | Agents & orchestrationMore projects & source code | |
| ha-gutcheck — READMEgithub.com | 2 | A Home Assistant checkup integration sending entity/integration observations and fixed answer choices to TypeSafe Jev for issue judgments and confidence. | Agents & orchestrationMore projects & source code | |
| ha-jev-conversation — READMEgithub.com | 0 | Czech Home Assistant conversation integration uses Jev to select intents and devices while code validates and executes Home Assistant services. | Agents & orchestrationMore projects & source code | |
| HarnessJudgegithub.com | 0 | Paste an agent step trace and get an ok, retry, escalate or stop decision for agent builders. | Agents & orchestrationMore projects & source code | |
| harnessjudge — READMEgithub.com | 0 | HarnessJudge takes an agent-step trace and asks TypeSafe Jev for an ok, retry, escalate, or stop decision. | Agents & orchestrationMore projects & source code | |
| hass-hunchgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| hass-hunch — READMEgithub.com | 0 | hass-hunch turns Home Assistant voice or text requests into typed actions by asking Jev many narrow questions; the project describes application code as calculating while Jev judges. The workspace documentation distinguishes planned integration components. | Agents & orchestrationMore projects & source code | |
| Hearthgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Hearth — READMEgithub.com | 0 | Hearth is a house-agent runtime with a Jev tool-call gate; the README says its default shadow mode logs hypothetical decisions without taking them. | Agents & orchestrationMore projects & source code | |
| Helm — READMEgithub.com | 14 | Helm routes coding tasks among local or hosted agents; the README separates probing from Jev’s decision about agent choice, risk and completion. | Agents & orchestrationMore projects & source code | |
| hermes-jev — READMEgithub.com | 3 | Hermes plugins offer Jev-based pre-tool gating and optional per-turn skill selection; with no configured key, both plugins fail open without sending a request. | Agents & orchestrationMore projects & source code | |
| hermes-jev-approvalsgithub.com | 19 | Provider for Hermes Agent's smart command approvals where Jev returns APPROVE, DENY or ESCALATE from six typed questions; the author's corrected 156-command run shows 9.8x faster, 4.2x fewer prompts, a retest 1.24x. | Agents & orchestrationMore projects & source code | |
| hermes-jev-curatorgithub.com | 2 | Hermes Agent plugin that has Jev label each pair of agent-created skills as duplicate, subset, conflict, unrelated and so on, and turns the results into safe archive plans. | Agents & orchestrationMore projects & source code | |
| hermes-jev-north-stargithub.com | 1 | Hermes Agent skill that interviews the owner until a run has a checkable finish line, generates the run prompt, and gates completion with shell checks plus Jev for the requirements no script can decide. | Agents & orchestrationMore projects & source code | |
| hermes-jev-routergithub.com | 3 | Plugin for Hermes Agent that adds a Jev decision layer to compact tool results, suppress duplicate tool calls, and skip main-model calls that would only restate results. | Agents & orchestrationMore projects & source code | |
| hermes-jev-router — READMEgithub.com | 1 | Hermes Jev Router currently returns model-routing plans, not automatic dispatch; its README says live Jev calls are pending and fixtures are synthetic. | Agents & orchestrationMore projects & source code | |
| hermes-jev-skills — READMEgithub.com | 957 | A Hermes skill collection uses Jev for small typed decisions rather than text generation, according to the README. | Agents & orchestrationMore projects & source code | |
| hermes-nerve — READMEgithub.com | 34 | A supervisory layer for Hermes Agent using Jev for unresolved semantic judgments, verification and trajectory decisions, with optional Laya/OpenJev backends. The documented authority boundary keeps canonical task state and final stop/continue control in Hermes rather than Jev; inclusion is for that decision integration, not proven agent safety or savings. | Agents & orchestrationMore projects & source code | |
| hermes-plugin-jevgithub.com | 0 | Jev (TypeSafe AI) System One decision engine plugin for Hermes Agent, backed by Agnes AI Flash. | Agents & orchestrationMore projects & source code | |
| hermes-plugin-jev — READMEgithub.com | 0 | This Hermes plugin describes Jev-based structured judgments and routing primitives, including an optional guard; its README says failures may fall back to local rules. The documented features are author claims, not tested behavior. | Agents & orchestrationMore projects & source code | |
| hermes-switchyardgithub.com | 4 | Jev-powered decision plugin for Hermes Agent: automatic skill selection, policy-constrained model routing, typed assessments, and approval-gated computer use on Windows, macOS, and Linux. | Agents & orchestrationMore projects & source code | |
| hermes-switchyard — READMEgithub.com | 4 | Hermes Switchyard uses Jev for structured choices and local policy validation; its model-routing tool is advisory and does not switch Hermes’s active model. | Agents & orchestrationMore projects & source code | |
| Hersona decidegithub.com | — | Persona attribute library for AI agents whose hersona decide command uses TypeSafe Jev to recommend reply, ask, search, use a tool or hold, with confidence, persona alignment and risk. | Agents & orchestrationMore projects & source code | |
| home-assistant-typesafegithub.com | 2 | Home Assistant conversation agent that maps commands to intents and devices with Jev and hands low-confidence or compound requests to an LLM fallback. | Agents & orchestrationMore projects & source code | |
| home-assistant-typesafe — READMEgithub.com | 2 | The Home Assistant conversation integration uses TypeSafe System One Choice/Noul questions for intent routing, entity resolution, and device control. | Agents & orchestrationMore projects & source code | |
| Hunchgithub.com | 4 | CLI and Agent Skill that sends each in-scope code chunk to Jev to answer a plain-English question across a repository, returning scored source locations, and also runs recurring plain-English review rules on diffs. | Agents & orchestrationMore projects & source code | |
| iii-directory Jev searchgithub.com | — | Worker for the iii engine that searches registered functions, installed skills and triggers with Jev relevance judgments, falling back to hybrid search. | Agents & orchestrationMore projects & source code | |
| ilmari-plugin-jevgithub.com | 0 | ilmari plugin: the jev step, a quick decision on TypeSafe Jev via OpenRouter | Agents & orchestrationMore projects & source code | |
| ilmari-plugin-jev — READMEgithub.com | 0 | ilmari workflow plugin adds a Jev decision step that selects from supplied answers and lets later steps route on that choice, normally through OpenRouter rather than an agent loop. | Agents & orchestrationMore projects & source code | |
| Imajin TypeSafe connectorgithub.com | — | Service connector in the reference implementation of the MJN trust protocol that lets identities call Jev with a sealed, kernel-held API key under a scoped typesafe:decide grant. | Agents & orchestrationMore projects & source code | |
| Imajin TypeSafe connector — repogithub.com | — | Service connector in the reference implementation of the MJN trust protocol that lets identities call Jev with a sealed, kernel-held API key under a scoped typesafe:decide grant. | Agents & orchestrationMore projects & source code | |
| insurance-claimsgithub.com | 0 | Jev + Claude prototype: FNOL claim triage and cost-aware Claude model routing | Agents & orchestrationMore projects & source code | |
| insurance-claims — READMEgithub.com | 0 | A first-notice-of-loss triage prototype uses Jev for typed judgments; README says end-to-end tests use simulated responses and real-claim accuracy is unmeasured. | Agents & orchestrationMore projects & source code | |
| IPFS Accelerate TypeSafe advisorgithub.com | — | Model server and agent supervisor with advisory TypeSafe integrations that route task kinds, guard traces and calibrate decisions beside the project's own Intelligence Index. | Agents & orchestrationMore projects & source code | |
| IPFS Accelerate TypeSafe advisor — repogithub.com | — | Model server and agent supervisor with advisory TypeSafe integrations that route task kinds, guard traces and calibrate decisions beside the project's own Intelligence Index. | Agents & orchestrationMore projects & source code | |
| JarvisCore Jev decisionsgithub.com | — | Native Jev support in the JarvisCore multi-agent runtime: agents ask typed Choice, Score and Noul questions through a decision client separate from the text model, and the kernel picks specialist subagents by Choice. | Agents & orchestrationMore projects & source code | |
| JarvisCore Jev decisions — repogithub.com | — | Native Jev support in the JarvisCore multi-agent runtime: agents ask typed Choice, Score and Noul questions through a decision client separate from the text model, and the kernel picks specialist subagents by Choice. | Agents & orchestrationMore projects & source code | |
| jcm-router — READMEgithub.com | 8 | jcm-router is a local proxy that uses TypeSafe Jev to pick a model and effort level per Claude Code message. | Agents & orchestrationMore projects & source code | |
| jcr — READMEgithub.com | 19 | JCR uses Jev to search a nested capability tree and return context for selected operations; the resolver returns documentation rather than executing commands. | Agents & orchestrationMore projects & source code | |
| Jcybergithub.com | 2 | MCP toolkit for agent-driven bug bounty work with scope gates, an evidence graph and memory, using Jev to score finding severity and check for duplicates. | Agents & orchestrationMore projects & source code | |
| jeffrey — READMEgithub.com | 2 | A coding-agent CLI separates decisions from execution: Jev is the default decision model, while a configurable LLM supplies tool arguments and code. | Agents & orchestrationMore projects & source code | |
| jermesgithub.com | 0 | Jev decision layer for Hermes Agent | Agents & orchestrationMore projects & source code | |
| jermes — READMEgithub.com | 0 | Hermes Agent plugin uses Jev for bounded skill, tool and context judgments, starting in shadow mode that logs suggestions without changing Hermes behavior. | Agents & orchestrationMore projects & source code | |
| Jev Architectgithub.com | 7 | Agent skill that guides a coding agent through finding where Jev decision loops could help a new idea or existing codebase, designing them, and testing the premise before committing to it. | Agents & orchestrationMore projects & source code | |
| Jev Assistgithub.com | 2 | Home Assistant conversation agent where Jev classifies each utterance and either calls a light, climate or cover service directly or hands off to Grok. | Agents & orchestrationMore projects & source code | |
| Jev Contextgithub.com | 1 | Local Codex code-search plugin that runs ripgrep and asks Jev a yes/no relevance question per passage, returning only likely matches and logging paired payloads to measure token savings. | Agents & orchestrationMore projects & source code | |
| Jev Decisionsgithub.com | 2 | Plugin for Hermes and other agents that adds tools to review risky plans, recommend human approval, check answers against sources, assess whether a task is really finished, and keep a local decision journal. | Agents & orchestrationMore projects & source code | |
| Jev DSH Decision Enginegithub.com | 338 | Agent-harness plugin that recommends which available tool, skill, or agent to use and rates output quality, running natively in DeepSeek Harness and via iPolloWork in OpenCode and Codex Harness. | Agents & orchestrationMore projects & source code | |
| Jev in Codexgithub.com | 3 | Codex plugin for batch labeling that turns a JSONL file of text records into a fully labeled file with custom question and criteria policies, while Codex reviews uncertain decisions. | Agents & orchestrationMore projects & source code | |
| Jev judgment modality for agent-rdf-memorygithub.com | — | RDF HowTo in OpenLink's OPAL agent skills that adds an optional System One judgment mode (Jev, Laya or NanoJev) to its memory harness, gating HowTo candidates and ask-vs-act elicitation, run in shadow or active mode. | Agents & orchestrationMore projects & source code | |
| Jev judgment modality for agent-rdf-memory — repogithub.com | — | RDF HowTo in OpenLink's OPAL agent skills that adds an optional System One judgment mode (Jev, Laya or NanoJev) to its memory harness, gating HowTo candidates and ask-vs-act elicitation, run in shadow or active mode. | Agents & orchestrationMore projects & source code | |
| Jev Logsgithub.com | 16 | TypeScript library and OpenTelemetry exporter wrapper that scores each log record's diagnostic value, priority and routing with Jev before any expensive LLM analysis, keeping every record in the archive. | Agents & orchestrationMore projects & source code | |
| JEV MCPgithub.com | 7 | MCP server exposing Jev as a single evaluate tool for Claude Code, Claude Desktop and Codex, plus an evaluation harness that measures its accuracy against general-purpose LLMs on shared datasets. | Agents & orchestrationMore projects & source code | |
| Jev plugin for Hermes Agentgithub.com | 8 | Hermes Agent plugin exposing Jev as four tools, jev_check, jev_route, jev_score, and jev_evaluate, with a bundled skill on writing good questions for coding-agent workflows. | Agents & orchestrationMore projects & source code | |
| JEV pluginsgithub.com | 0 | Cursor and Hermes plugin marketplace for four JEV Labs selectors, for memories, agents, models, and MCP tools, that filter candidates by task and budget, locally by default with Jev as an optional judge. | Agents & orchestrationMore projects & source code | |
| Jev recipes on Treg — repogithub.com | — | Interactive guide to Jev with a live LLM-versus-Jev race and copyable GTM automation recipes built on Treg, such as signup fraud screening, buying-signal triage, and viral content monitoring. | Agents & orchestrationMore projects & source code | |
| Jev sites — repogithub.com | 1 | Eight small judgment tools on one site, covering support routing, lead scoring, marketplace listing trust, citation checks, prompt screening, agent skill routing, resume matching, and clause selection. | Agents & orchestrationMore projects & source code | |
| Jev Socialgithub.com | 132 | Social media research tool for Instagram, TikTok and LinkedIn in which Jev picks each next read-only operation, such as search, open a post or read comments, that socai runs in your real Chrome, ending in a cited report. | Agents & orchestrationMore projects & source code | |
| Jev System Architect Skillgithub.com | 2 | Agent skill for Cursor, Claude Code and compatible tools that looks for fuzzy semantic judgment in a system and turns it into small Choice, Score and Noul primitives, keeping control flow in code. | Agents & orchestrationMore projects & source code | |
| Jev use casesgithub.com | 18 | Python harnesses implementing most of the decision use cases TypeSafe lists for Jev, such as risk assessment, each with confidence-gated logic that thresholds Choice, Score and Noul probabilities in code. | Agents & orchestrationMore projects & source code | |
| jev-acceleratorgithub.com | 0 | A skill using jev to speed up LLM use. | Agents & orchestrationMore projects & source code | |
| jev-accelerator — READMEgithub.com | 0 | Jev Accelerator packages coding-agent tools such as candidate-file reranking, commit checks, and documentation audits around TypeSafe Jev judgments. | Agents & orchestrationMore projects & source code | |
| jev-agentgithub.com | 0 | agent for jev | Agents & orchestrationMore projects & source code | |
| jev-agent — READMEgithub.com | 0 | The agent loop assigns bounded decisions to Jev while an LLM generates tool arguments and natural-language replies. | Agents & orchestrationMore projects & source code | |
| jev-agent-authorization — READMEgithub.com | 5 | An agent-tool authorization starter kit combining Kinde identity/permissions with Jev semantic judgments. Application policy retains allow, user-approval or stop decisions. | Agents & orchestrationMore projects & source code | |
| jev-agent-hooks — READMEgithub.com | 2 | jev-agent-hooks uses TypeSafe Jev for coding-agent skill suggestions and subagent-model selection across Claude Code, Codex, and pi. | Agents & orchestrationMore projects & source code | |
| jev-agent-kit — READMEgithub.com | 2 | This command-line and MCP toolkit gives agents typed decisions from TypeSafe Jev; applications decide what to do with the answers. | Agents & orchestrationMore projects & source code | |
| jev-agent-router — READMEgithub.com | 0 | A computer-use router keeps exact matches in code, sends bounded semantic choices to Jev, and returns uncertainty to the main agent. | Agents & orchestrationMore projects & source code | |
| jev-agent-skillgithub.com | 0 | Decisões tipadas em milissegundos com o Jev (System One da TypeSafe) via OpenRouter — state + perguntas noul/choice/score entram, decisões calibradas saem. Validação pesada baseada nos conceitos do modelo, bandas auto/hitl/abstain, deteção de prompt injection e servidor MCP. Zero dependências (Node ≥20). | Agents & orchestrationMore projects & source code | |
| jev-agent-skillgithub.com | 0 | Claude Code and ZCode skill that offloads small classify, screen, score and verify judgments from the main model to Jev on OpenCode Zen's free /v1/systemone tier, with a zero-dependency caller and trigger rules. | Agents & orchestrationMore projects & source code | |
| jev-agent-skill — READMEgithub.com | 0 | Node-based agent skill calls Jev through OpenRouter with typed requests and probability-based action bands, including request validation, MCP mode and labeled-case evaluation commands. | Agents & orchestrationMore projects & source code | |
| jev-agent-skill — READMEgithub.com | 0 | The agent-skill package routes recurring classification, screening, scoring, and checking judgments from the main model to Jev. | Agents & orchestrationMore projects & source code | |
| jev-agent-skill-router — READMEgithub.com | 23 | An optional Jev routing layer chooses among skill names or returns no-skill/review; its README reports exploratory results on a synthetic benchmark, not production accuracy. | Agents & orchestrationMore projects & source code | |
| jev-agent-toolgithub.com | — | Bring-your-own-key CLI, Python API and local MCP server built on the official Python SDK, letting agents invent typed questions at runtime and get Jev decisions with probabilities. | Agents & orchestrationMore projects & source code | |
| jev-agent-toolkit — READMEgithub.com | 2 | A portable agent-skill toolkit with an optional MCP bridge exposes bounded, typed decisions from TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| Jev-Agenticgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Jev-Agentic — READMEgithub.com | 0 | macOS voice-agent implementation supplies Jev with bounded action, visible-element and completion choices rather than asking it to generate commands. | Agents & orchestrationMore projects & source code | |
| jev-ai-sdk-form-router — READMEgithub.com | 10 | A Vercel AI SDK demo routes sample form submissions with Jev, then falls back to a separate GPT model for low-confidence or failed judgments. | Agents & orchestrationMore projects & source code | |
| jev-android-supergithub.com | 3 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-android-super — READMEgithub.com | 3 | An Android automation agent asks Jev which action and element to choose; Android AccessibilityService deterministically performs the selected gestures. | Agents & orchestrationMore projects & source code | |
| jev-applygithub.com | 9 | Memory-backed job-application skill: Jev selects, an LLM writes only new text, Playwright fills. Greenhouse + Ashby. | Agents & orchestrationMore projects & source code | |
| jev-apply — READMEgithub.com | 9 | The job-application skill uses Jev to select saved personal facts for form fields; the README says Jev does not write the text. | Agents & orchestrationMore projects & source code | |
| jev-architect — READMEgithub.com | 2 | jev-architect describes a gate for deciding when to use Jev and says the skill itself does not call Jev. | Agents & orchestrationMore projects & source code | |
| jev-architect — READMEgithub.com | 7 | Jev Architect is an agent skill for exploring and testing possible Jev uses; its README calls suggested designs hypotheses, not measured results. | Agents & orchestrationMore projects & source code | |
| jev-askable-arm — READMEgithub.com | 11 | An askable-arm robotics demo maps a goal to hardcoded movement primitives, with Jev selecting one primitive per step and Python executing it. | Agents & orchestrationMore projects & source code | |
| jev-botgithub.com | — | Decision Tools: Self-hosted Jev decision workbench and Feishu bot: automatic choices, probabilities, and experimental word/character writing. | Agents & orchestrationMore projects & source code | |
| jev-bot — READMEgithub.com | 9 | A market-decision bot defaults to simulated data and a local offline decision engine; with a key it can use TypeSafe Jev, while execution remains paper-only. | Agents & orchestrationMore projects & source code | |
| jev-bot — READMEgithub.com | 1 | The README presents a macOS control app that lets TypeSafe Jev choose a UI control from its description, with a separate driver executing it. | Agents & orchestrationMore projects & source code | |
| jev-bouncergithub.com | 1 | Claude Code plugin: a 3-cent bouncer for your agent's shell. Jev typed probabilities auto-allow routine commands, deny destructive ones, and flag prompt injection in tool results. | Agents & orchestrationMore projects & source code | |
| jev-bouncer — READMEgithub.com | 1 | This Claude Code plugin asks Jev for typed-probability second opinions before commands, edits, and MCP calls. | Agents & orchestrationMore projects & source code | |
| jev-brotato — READMEgithub.com | 0 | The game harness reads structured Brotato state, asks TypeSafe Jev to choose a legal action, then applies and records that choice; its demo/tests can run offline. | Agents & orchestrationMore projects & source code | |
| jev-browser-bridgegithub.com | — | Browser agents: plugs any CDP browser into a Jev loop, where a Jev Choice picks the operation and its target element each step from candidates read off the DOM rather than the layout, so the same agent runs on Chrome and on engines that never draw a page (Moli, Lightpanda, Kitesurf), passing at least 90% of runs on each of fourteen browsers tested. | Agents & orchestrationMore projects & source code | |
| jev-builder-loopgithub.com | 3 | Grok skill that uses Jev as a judgment sensor at builder-agent loop boundaries, mixing its probabilities with priors to choose the next act, benchmarked on shipping a signed Android App Bundle. | Agents & orchestrationMore projects & source code | |
| jev-builder-loop — READMEgithub.com | 3 | This Grok skill returns typed Jev probabilities for operator-written questions and state; the operator combines them with priors and acts. | Agents & orchestrationMore projects & source code | |
| Jev-chat-assistantgithub.com | — | Decision Tools: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details. | Agents & orchestrationMore projects & source code | |
| jev-chat-windows-deepseek-jevgithub.com | — | Decision Tools: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details. | Agents & orchestrationMore projects & source code | |
| jev-chat-windows-layagithub.com | — | Decision Tools: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details. | Agents & orchestrationMore projects & source code | |
| jev-claude-demo — READMEgithub.com | 0 | A Spring Boot integration demo assigns structured decisions to Jev and prose to Claude, and adds Spring AI advisors for guardrails, RAG, tool selection, and evaluation. | Agents & orchestrationMore projects & source code | |
| jev-claude-orchestrator — READMEgithub.com | 0 | A Claude Code orchestration plugin separates frontier-model planning, Jev decisions, parallel subagent work, and verification. | Agents & orchestrationMore projects & source code | |
| jev-claw — READMEgithub.com | 4 | jev-claw adds a jev_route tool for TypeSafe Jev task routing, with automatic routing documented as opt-in. | Agents & orchestrationMore projects & source code | |
| jev-codegithub.com | 118 | Command-line toolkit that coding agents hand judgment-heavy work to, one typed Jev workflow per request. | Agents & orchestrationMore projects & source code | |
| jev-code — READMEgithub.com | 118 | A Jev-first coding agent routes bounded requests into built-in workflows and uses fixed-choice questions; code makes decisions, with Pi as a fallback for unsupported tasks. | Agents & orchestrationMore projects & source code | |
| jev-codex-bridge — READMEgithub.com | 6 | Jev Codex Bridge routes model and reasoning-effort selection for new Codex messages through TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| jev-codex-router — READMEgithub.com | 277 | This Codex integration uses Jev to select model and reasoning effort per model call; the README separately labels its savings figure as a historical simulation, not measured quota savings. | Agents & orchestrationMore projects & source code | |
| jev-compaction — READMEgithub.com | 4 | A coding-agent context compactor using Jev to select original text for retention, with removed content recoverable byte for byte. It is selective compaction, not Jev-written summaries. | Agents & orchestrationMore projects & source code | |
| jev-compaction — READMEgithub.com | 3 | A Hermes context-compaction tool asks Jev whether tool calls are stale and removes selected tool calls/results without rewriting kept context. | Agents & orchestrationMore projects & source code | |
| jev-control-plane — READMEgithub.com | 0 | JEV Control Plane recommends a runtime for coding agents from supplied task context; the README says verified local facts and fixed safety rules remain authoritative. | Agents & orchestrationMore projects & source code | |
| jev-corrective-raggithub.com | 0 | Corrective RAG where the retrieve-or-not, chunk-grading and groundedness gates are Jev calls instead of LLM judges, averaging 0.75 LLM calls per query with 7x fewer LLM calls and 4x lower p50 latency. | Agents & orchestrationMore projects & source code | |
| jev-corrective-rag — READMEgithub.com | 0 | A corrective-RAG design sends typed decision gates to Jev and reserves the language model for answer generation. | Agents & orchestrationMore projects & source code | |
| jev-craftgithub.com | 0 | 让 Agent 长出「操作杆」：用 System One 模型（Jev）承担高频判断、多模态模型当眼睛。给思路 + 可跑的参考实现（75 条离线断言，零依赖零 key） | Agents & orchestrationMore projects & source code | |
| jev-craft — READMEgithub.com | 0 | A Chinese-language agent-architecture reference assigns defined choices to Jev’s System One layer; README says its offline pipeline tests use mock contracts, not live inference. | Agents & orchestrationMore projects & source code | |
| jev-crawlersgithub.com | 1 | Unix-style bug-discovery tools (seed, expand, judge, verify, report, norms) that pipe JSON lines and have Jev judge every node as they follow leads through a repo, at about $0.00006 per node. | Agents & orchestrationMore projects & source code | |
| jev-cuagithub.com | 1 | Voice-controlled Mac computer use on TypeSafe Jev: speech to typed decisions to verified accessibility actions, no screenshots to a big model | Agents & orchestrationMore projects & source code | |
| jev-cua — READMEgithub.com | 1 | A voice-driven macOS computer-use application integrating TypeSafe Jev for action decisions. It is a control application, not proof of safe or successful real-world execution. | Agents & orchestrationMore projects & source code | |
| Jev-CustomerService-Demogithub.com | 0 | Ticket-Routing im Kundenservice mit dem Entscheidungsmodell Jev von typesafe.ai | Agents & orchestrationMore projects & source code | |
| Jev-CustomerService-Demo — READMEgithub.com | 0 | A German customer-service demo routes tickets into four queues using Jev answers to fixed typed questions and compares them with dispatcher-selected queues; reported metrics are not repeated as verified. | Agents & orchestrationMore projects & source code | |
| jev-decisiongithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-decision — READMEgithub.com | 0 | A local support-routing demo makes one Jev classification per message, then uses code for fixed replies or a local human-review ticket. | Agents & orchestrationMore projects & source code | |
| jev-decision-gateway — READMEgithub.com | 0 | The gateway sends small bounded questions from apps or agents to Jev and lets a deterministic policy layer handle typed answers before generative-model calls. | Agents & orchestrationMore projects & source code | |
| jev-decision-layergithub.com | 2 | Typed decision layer for coding agents — Claude Code, Codex, Antigravity, Hermes, VS Code. One MCP server, 20 tools, 36 recipes, per-provider calibrated gates. | Agents & orchestrationMore projects & source code | |
| jev-decision-layer — READMEgithub.com | 2 | An independent MCP/agent decision layer validating TypeSafe Jev or local Laya answers and recording receipts. The host, not the decision model, retains execution authority. | Agents & orchestrationMore projects & source code | |
| jev-decision-makergithub.com | 0 | omp plugin: JEV (TypeSafe System One) picks the next coding step from agent-supplied candidates | Agents & orchestrationMore projects & source code | |
| jev-decision-maker — READMEgithub.com | 0 | An omp branch-point tool submits 2–5 candidate actions to Jev for selection or rubric scoring, with the README describing OpenRouter as the route. | Agents & orchestrationMore projects & source code | |
| jev-decision-maker-at-meteora — READMEgithub.com | 0 | The README presents a Condor trading agent using Jev judgments to rank Meteora liquidity pools and size positions; this review does not validate trading performance. | Agents & orchestrationMore projects & source code | |
| Jev-decision-plugingithub.com | 0 | A Codex plugin that turns noisy input into small, typed, calibrated decisions using TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| Jev-decision-plugin — READMEgithub.com | 0 | This Codex plugin package offers patterns for composing TypeSafe System One typed judgments, including routing, ranking, extraction, and guardrails. | Agents & orchestrationMore projects & source code | |
| jev-decision-skillgithub.com | 0 | 让 AI Agent 直接调用 TypeSafe AI「Jev」决策模型的 Skill —— 免 API key，纯标准库 CLI。Agent Skill for TypeSafe AI Jev (System One). | Agents & orchestrationMore projects & source code | |
| jev-decision-skill — READMEgithub.com | 0 | The README describes a TypeSafe Jev agent skill with a direct official endpoint and a separate no-key third-party Playground route. | Agents & orchestrationMore projects & source code | |
| Jev-Defensegithub.com | 0 | A Jev-powered security guard for AI agents: blocks dangerous tool calls, strips prompt injection, scans skills. Works with Claude Code, Codex, Copilot CLI, Gemini CLI, Cursor, and OpenCode. | Agents & orchestrationMore projects & source code | |
| Jev-Defense — READMEgithub.com | 0 | Agent guardrail uses Jev’s narrow probability judgments with code-side policy to check actions, read content and skill files, but is not a sandbox or guaranteed defense. | Agents & orchestrationMore projects & source code | |
| jev-demogithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-demogithub.com | 0 | a demo of Typesafe's Jev a System One Model for structured decsion making | Agents & orchestrationMore projects & source code | |
| jev-demo — READMEgithub.com | 0 | This Jev demo culminates in an AI-agent tool-call guard that returns allow, confirm, review, or deny outcomes before the tool runs. | Agents & orchestrationMore projects & source code | |
| jev-demo — READMEgithub.com | 0 | This support-email demo sends typed questions to Jev, then application code uses the answers to choose a queue; canned samples are only inputs. | Agents & orchestrationMore projects & source code | |
| jev-demosgithub.com | — | Decision Tools: Every demo lives in its own folder with its own README, dependencies and instructions. Most run without an API key in a clearly labelled \SIMULATED\ mode; put \TYPESAFE\_API\_KEY=...\ in the demo folder's \.env\ for live results. | Agents & orchestrationMore projects & source code | |
| jev-dingtalkgithub.com | 1 | 钉钉信息 Jev 分拣器（jev-dingtalk）：把钉钉邮件与聊天分拣成 Now / Today / Queue / Ignore 清单——谁在等回复、谁需要人看一眼。dws 取数、Jev 分类、报告本地渲染，可直接作为 Agent 技能使用。| DingTalk mail & chat triage with Jev. | Agents & orchestrationMore projects & source code | |
| jev-dingtalk — READMEgithub.com | 1 | jev-dingtalk converts DingTalk emails and chats for Jev triage, categorizing reply needs and urgency while marking encrypted content rather than guessing. | Agents & orchestrationMore projects & source code | |
| jev-dsh — READMEgithub.com | 0 | DeepSeek Harness integration exposes Jev as a system_one tool and skill for bounded typed decisions, not as a generative chat model. | Agents & orchestrationMore projects & source code | |
| jev-dsh-decision — READMEgithub.com | 338 | Jev DSH connects TypeSafe AI's Jev decision capability to an agent; the README says the original agent retains planning and execution. | Agents & orchestrationMore projects & source code | |
| jev-editor-skillgithub.com | 0 | Editorial gate skill for Claude Code and other agents. Scores a draft on AI tells, your own voice, editorial quality, and SEO using TypeSafe's Jev. | Agents & orchestrationMore projects & source code | |
| jev-editor-skill — READMEgithub.com | 0 | Jev Editor is an agent skill for Markdown/MDX editorial review: Jev makes narrow qualitative judgments while code checks countable features; fact-checking stays with the user’s coding agent. | Agents & orchestrationMore projects & source code | |
| jev-effortgithub.com | 0 | Per-turn reasoning effort for coding agents, classified by TypeSafe's Jev via AI SDK (Vercel AI Gateway / OpenRouter / TypeSafe). Claude Code plugin; Codex, Cursor, Amp next. | Agents & orchestrationMore projects & source code | |
| jev-effort — READMEgithub.com | 0 | Coding-agent integration classifies each prompt with Jev and adds a per-turn reasoning/verification-effort directive; supported hooks steer the turn rather than proving actual effort control. | Agents & orchestrationMore projects & source code | |
| jev-eval-mcpgithub.com | 3 | Eval-first MCP server for Jev with tools to prototype a question, map a question set over many items, and measure question variants against labeled examples for accuracy, calibration and thresholds. | Agents & orchestrationMore projects & source code | |
| Jev-Evolvegithub.com | — | Agent framework where every branch is a typed Jev question and the policy text evolves from the agent's own mistakes, reporting how much of a gain is selection noise; one schema scored 0.188 or 0.542 depending on option order. | Agents & orchestrationMore projects & source code | |
| jev-eyes — READMEgithub.com | 1 | jev-eyes locally converts image content into text state for TypeSafe Jev; see() needs no key, while ask() requires TYPESAFE_API_KEY for the Jev call. | Agents & orchestrationMore projects & source code | |
| jev-fabric — READMEgithub.com | 0 | Native process supervisor owns processes and bounded logs, asking Jev typed decision questions only on an explicit request rather than autonomously. | Agents & orchestrationMore projects & source code | |
| jev-factorio-agent — READMEgithub.com | 1 | A Factorio agent uses Jev for high-level goals and next actions while deterministic code owns game rules; an offline mock example is not a live Jev run. | Agents & orchestrationMore projects & source code | |
| jev-fast-judgementsgithub.com | 0 | Agent Skill that offloads bulk yes/no, pick-one-option and rating judgments to TypeSafe's Jev so Claude spends fewer tokens during investigations. | Agents & orchestrationMore projects & source code | |
| jev-fast-judgements — READMEgithub.com | 0 | Agent skill teaches when to send bulk bounded investigation judgments to Jev, passing material from disk and returning short verdict tables while the agent handles retained and uncertain items. | Agents & orchestrationMore projects & source code | |
| jev-flash-router — READMEgithub.com | 3 | An MCP decision-router server powered by TypeSafe Jev evaluates agent context and returns probability-based routing judgments. | Agents & orchestrationMore projects & source code | |
| jev-free-router — READMEgithub.com | 0 | This OpenCode plugin offers Jev, jev-free, and jev-go as model choices and uses Jev to estimate which catalog model suits each new request. | Agents & orchestrationMore projects & source code | |
| jev-gate — READMEgithub.com | 3 | A Claude Code plugin uses Jev to choose between direct work and task decomposition into delegated jobs. | Agents & orchestrationMore projects & source code | |
| jev-gatesgithub.com | 0 | Three gates for any coding agent: a deterministic approval gate before irreversible actions, a completion gate that demands evidence instead of «done», and a rubric regression gate. Powered by TypeSafe Jev (System One). Works with Claude Code, Codex, DeepSeek Harness, Cursor, opencode, Hermes, plain shell and CI. | Agents & orchestrationMore projects & source code | |
| jev-gates — READMEgithub.com | 0 | The README describes TypeSafe Jev returning a probability for a typed question and explicitly labels the gate as advice, not enforcement. | Agents & orchestrationMore projects & source code | |
| jev-gateway — READMEgithub.com | 273 | An independent local gateway for coding agents that delegates tool-call selection to TypeSafe Jev while forwarding other work to the usual LLM. It is a client of the public API, not an official TypeSafe product; inclusion does not certify advertised speed or token savings. | Agents & orchestrationMore projects & source code | |
| jev-godfathergithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-godfather — READMEgithub.com | 0 | Web advisor decomposes a project workflow, uses Jev to screen bounded/observable/repeated decisions and lets code derive a fit verdict, rather than letting Jev design or execute products. | Agents & orchestrationMore projects & source code | |
| jev-guard — READMEgithub.com | 50 | The README describes jev-guard as a coding-agent hook powered by Jev. | Agents & orchestrationMore projects & source code | |
| jev-harness — READMEgithub.com | 15 | A TypeScript harness wraps Jev answers with action policies, confidence gates, shadow mode, reusable recipes and an offline evaluation CLI. | Agents & orchestrationMore projects & source code | |
| jev-harness — READMEgithub.com | 0 | Claude Code harness uses Jev for narrow permission, skill, dispatch and context-selection decisions while deterministic policy and Claude retain their separate roles. | Agents & orchestrationMore projects & source code | |
| jev-harness — READMEgithub.com | 10 | The README presents a coding-agent harness powered by TypeSafe Jev and documents a local heuristic fallback when offline or no API key is supplied. | Agents & orchestrationMore projects & source code | |
| jev-harness — READMEgithub.com | 26 | An independent community research-stage proposal-review contract: an LLM proposes an action, Jev answers narrow questions, and the host retains execution and authorization. Its offline benchmark uses scripted mock results, not measurements of Jev; it is not an official TypeSafe SDK or a production agent runtime. | Agents & orchestrationMore projects & source code | |
| jev-harness-routergithub.com | 4 | Per-turn router for agent harnesses where one Jev call of about 350 ms sets the model tier, effort, tools and skill behind a hard deadline; it picks the right skill 90.0% of the time versus 59.4% for keywords. | Agents & orchestrationMore projects & source code | |
| jev-harness-router — READMEgithub.com | 4 | This harness-router uses per-tool Jev Noul judgments and a Choice over candidate tools or skills to route an agent turn. | Agents & orchestrationMore projects & source code | |
| jev-healthcare-support-router — READMEgithub.com | 0 | A TypeScript demo uses Jev for structured intent and urgency judgments in an administrative healthcare support router; application code routes cases or escalates them to people, not to clinical decisions. | Agents & orchestrationMore projects & source code | |
| jev-helpergithub.com | 5 | A helper which use JEV to play ra2web(WannaFire Version)[王二火大] | Agents & orchestrationMore projects & source code | |
| jev-helper — READMEgithub.com | 5 | A Chrome/Edge game-control extension offering a Jev cloud Choice route and local-model configuration. Its documented cloud endpoint uses TypeSafe’s candidate-choice protocol, not chat completions. | Agents & orchestrationMore projects & source code | |
| jev-hermesgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-hermes — READMEgithub.com | 0 | jev-hermes is a Hermes Agent plugin that exposes TypeSafe’s hosted decision API for typed Choice, Score, and Noul questions. | Agents & orchestrationMore projects & source code | |
| jev-hft-model-routergithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-hft-model-router — READMEgithub.com | 1 | A decision router classifies trading requests and selects from compatible registered models; risk limits and other execution constraints remain code-controlled, and it does not execute orders. | Agents & orchestrationMore projects & source code | |
| jev-in-jev — READMEgithub.com | 1 | This Python reference router asks Jev whether a task suits typed decisions, then leaves the route choice to local policy. | Agents & orchestrationMore projects & source code | |
| Jev-in-the-Loopgithub.com | 4 | Codex plugin for fast browser work in local Chrome: Codex plans the task and prepares text, while Jev picks each next action and its target. | Agents & orchestrationMore projects & source code | |
| jev-inbox-queuegithub.com | — | Context GC: Turn an inbox into a short action queue with Jev (TypeSafe System One) | Agents & orchestrationMore projects & source code | |
| jev-inbox-triagegithub.com | 0 | Jev (TypeSafe) typed prefilter skill: two-phase inbox triage with confidence gating + pre-send guard | Agents & orchestrationMore projects & source code | |
| jev-inbox-triage — READMEgithub.com | 0 | This inbox-triage skill uses Jev only to choose processing depth and draft readiness; it does not do the work. | Agents & orchestrationMore projects & source code | |
| jev-judgmentgithub.com | 6 | Agent Skill that has a coding agent send closed questions to Jev before asking the user, before risky or out-of-scope actions and after failures to decide on retries, in about 250 ms, failing open without a key. | Agents & orchestrationMore projects & source code | |
| jev-judgment — READMEgithub.com | 6 | This agent skill sends bounded yes/no, choice, and score questions to TypeSafe Jev; it treats confidence as evidence, not permission to act. | Agents & orchestrationMore projects & source code | |
| jev-latam-lead-triagegithub.com | 0 | Triage de leads de WhatsApp/CRM con Jev (TypeSafe AI): ruteo por confianza, plantilla n8n y benchmark en español. Sin dependencias. No afiliado. | Agents & orchestrationMore projects & source code | |
| jev-latam-lead-triage — READMEgithub.com | 0 | A Spanish WhatsApp/CRM lead triage template routes messages by Jev confidence; README says it has never been run against the real Jev API. | Agents & orchestrationMore projects & source code | |
| jev-laya-benchmarkgithub.com | — | Decision Tools: Speed and accuracy benchmark: TypeSafe's Jev API vs the local Laya MLX typed-decision model on synthetic tasks | Agents & orchestrationMore projects & source code | |
| jev-layergithub.com | 2 | Portable decision layer for agent harnesses such as Hermes, OMP, Codex and generic MCP that routes bounded choices through Jev and records receipts for replay, while the host keeps execution, permissions and approvals. | Agents & orchestrationMore projects & source code | |
| jev-libero — READMEgithub.com | 79 | jev-libero describes a robot-control loop in which Jev selects intent, motion family and atomic input for LIBERO tasks; the README also documents multiple providers. | Agents & orchestrationMore projects & source code | |
| jev-loop-detectorgithub.com | 1 | Catch an AI agent stuck in a loop — Jev grades each step progress/repeating/stuck/escalate. Single-file, no build. Use free: jevtypesafeai.com/tools/agent-loop-detector | Agents & orchestrationMore projects & source code | |
| jev-loop-detector — READMEgithub.com | 1 | An agent-run step classifier labeling progress, repetition, stuck and escalation, documented for a community Jev proxy or configurable TypeSafe endpoint. A supervisor, not this display, must decide whether to halt. | Agents & orchestrationMore projects & source code | |
| jev-market-reflex — READMEgithub.com | 3 | Jev Market Reflex turns live market state into BUY/SELL/HOLD decisions and applies them to a simulated paper portfolio, not real-money trading. | Agents & orchestrationMore projects & source code | |
| jev-mcpgithub.com | 0 | Jev agent workstation. | Agents & orchestrationMore projects & source code | |
| jev-mcp (benballintyn)github.com | 1 | MCP server giving coding agents typed, calibrated judgments from TypeSafe's Jev model | Agents & orchestrationMore projects & source code | |
| jev-mcp — READMEgithub.com | 0 | This TypeScript MCP runtime creates and tests Jev-powered capabilities, with typed judgments, versioned definitions, and conditional batch workflows described in its guide. | Agents & orchestrationMore projects & source code | |
| jev-mcp-middleware — READMEgithub.com | 0 | The MCP middleware asks TypeSafe Jev for bounded risk signals, while deterministic policy code decides whether to allow, review, or deny; Jev is advisory, not an authorization guarantee. | Agents & orchestrationMore projects & source code | |
| jev-mctsgithub.com | 1 | Busca em arvore (MCTS/PUCT) com avaliacao tipada do TypeSafe Jev e portao humano obrigatorio. A arvore supoe, a sonda mede: marco so e concedido por codigo de saida verde. | Agents & orchestrationMore projects & source code | |
| jev-megithub.com | 13 | Agent skill that grills a plan as a design-tree interview, printing every unlocked decision each round with a recommended answer, and calling Jev ad hoc when a typed judgment helps. | Agents & orchestrationMore projects & source code | |
| jev-me — READMEgithub.com | 13 | A plan-review skill conducts a design-tree interview and uses Jev for typed judgments. | Agents & orchestrationMore projects & source code | |
| jev-mermaid — READMEgithub.com | 3 | Jev Flow Studio demonstrates typed Jev answers driving Mermaid branches; without a key, Preview mode uses fixed fixtures instead of a live call. | Agents & orchestrationMore projects & source code | |
| jev-mistral-reflex — READMEgithub.com | 0 | The README describes a Mistral-agent cascade where Jev selects declared tools or actions and uncertain or open-ended requests fall through to a Mistral-compatible responder; the demo uses synthetic fixtures. | Agents & orchestrationMore projects & source code | |
| jev-mobile — READMEgithub.com | 8 | A durable Android sub-agent uses Jev for typed action choices between local observation and execution; the README limits the validated scope and excludes sensitive operations. | Agents & orchestrationMore projects & source code | |
| jev-mobile — READMEgithub.com | 4 | jev-mobile uses TypeSafe Jev to choose an Android action and target from the accessibility tree. | Agents & orchestrationMore projects & source code | |
| jev-model-routergithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-model-router — READMEgithub.com | 0 | This local playground explores Jev routing choices; without a TypeSafe key it returns a setup error, not mock decisions. | Agents & orchestrationMore projects & source code | |
| jev-nethackgithub.com | 0 | TypeSafe's Jev model plays NetHack 5.0: code lists the legal moves, Jev picks one, no LLM in the loop | Agents & orchestrationMore projects & source code | |
| jev-nethack — READMEgithub.com | 0 | NetHack agent reads the terminal screen, presents a legal-action menu to Jev Choice, then executes the selected action with deterministic keystroke code. | Agents & orchestrationMore projects & source code | |
| jev-no-enemgithub.com | 0 | Reproducible benchmark evaluating TypeSafe AI's Jev (System One paradigm) on Brazil's ENEM 2025 standardized exam. Evaluates typed decision-making, domain-specific accuracy, and RLCD uncertainty calibration against open LLM baselines with an interactive GitHub Pages dashboard. | Agents & orchestrationMore projects & source code | |
| jev-on-air — READMEgithub.com | 0 | Livestream research tool transcribes speech and uses Jev judgments to drive simulated trading signals and a paper portfolio, never real orders. | Agents & orchestrationMore projects & source code | |
| jev-oncallgithub.com | 3 | Incident triage on TypeSafe Jev — the model judges, plain code decides. Routing on probability distributions with a human-review middle band and fail-open defaults. | Agents & orchestrationMore projects & source code | |
| jev-oncall — READMEgithub.com | 3 | An incident-triage service asks Jev four typed questions per alert; ordinary code routes the result, while Jev itself never pages anyone. | Agents & orchestrationMore projects & source code | |
| jev-one-systemgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-one-system — READMEgithub.com | 0 | A Next.js support-triage demo asks Jev five typed questions in one call and uses confidence thresholds in code to route tickets. | Agents & orchestrationMore projects & source code | |
| jev-organizegithub.com | 2 | Throw in a pile of company files and get them classified and organized by department, type, sensitivity, date, counterparty and PII, with an index for AI agents. Powered by TypeSafe's Jev on OpenRouter (17¢ per 1,000 files). Zero-dependency Node CLI + Claude skill + Codex agent. | Agents & orchestrationMore projects & source code | |
| jev-pi-model-router — READMEgithub.com | 0 | pi extension uses Jev difficulty/work-kind/error-cost judgments to switch to an available model matching the turn's tier and capabilities; uncertainty or failure leaves the current model unchanged. | Agents & orchestrationMore projects & source code | |
| jev-pi-skillsgithub.com | 0 | pi extension: TypeSafe Jev decisions as tools (search, rerank, supervise, pick) and hooks (skill select, compaction) | Agents & orchestrationMore projects & source code | |
| jev-pi-skills — READMEgithub.com | 0 | pi tools use Jev for search selection, reranking, worker supervision and bounded choices; automatic hooks select relevant skills and remove confidently disposable assistant/tool turns before normal compaction. | Agents & orchestrationMore projects & source code | |
| jev-playgroundgithub.com | 0 | A playground for TypeSafe AI's Jev (System One model), built around five real production workflows: support triage, RAG relevance gating, agent action firewall, inline moderation, and CI eval judging. | Agents & orchestrationMore projects & source code | |
| jev-playgroundgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-playground — READMEgithub.com | 0 | This local playground demonstrates six Jev workflows, including support triage, RAG relevance gating, agent-action checks, moderation, and CI evaluation; Jev returns judgments while application code enforces outcomes. | Agents & orchestrationMore projects & source code | |
| jev-playground — READMEgithub.com | 0 | Experimental game engine combines a visual decoder, structured-state Jev decisions and deterministic skills; its README explicitly limits current results to bounded probes. | Agents & orchestrationMore projects & source code | |
| jev-plays-pokemon — READMEgithub.com | 0 | A CLI game plays a Pokémon ROM with Jev and provides a combined video and live-decision viewer. | Agents & orchestrationMore projects & source code | |
| jev-plays-super-mario-bros — READMEgithub.com | 0 | A Mario demo asks Jev to choose moves from game state, then verifies them with emulator rollback search. | Agents & orchestrationMore projects & source code | |
| Jev-plugin-for-hermesgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Jev-plugin-for-hermes — READMEgithub.com | 0 | A standalone Hermes plugin exposes TypeSafe Jev as typed decision tools, a skill, and a command. Its README says the plugin evaluates supplied evidence and does not itself authorize or execute side effects. | Agents & orchestrationMore projects & source code | |
| jev-plugins — READMEgithub.com | 0 | This Cursor marketplace lists four Jev-related selection routers for memory, agents, models, and MCP; other concepts are marked design-only. | Agents & orchestrationMore projects & source code | |
| jev-pokemongithub.com | 116 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-pokemon — READMEgithub.com | 116 | A Pokémon Red agent experiment that presents game state and legal options to Jev for decisions. The documented gateway mode uses real Jev through Vercel AI Gateway, while the default mock mode is a stand-in; inclusion concerns the integration, not verification of the claimed game completion. | Agents & orchestrationMore projects & source code | |
| jev-predict-skillgithub.com | 3 | Host agent skill that predicts which closed verdict another skill would reach, such as SHIP or NEEDS_WORK, by handing its outcome set to Jev instead of running the skill. | Agents & orchestrationMore projects & source code | |
| jev-predict-skill — READMEgithub.com | 3 | jev-predict-skill asks TypeSafe Jev to predict a target skill’s closed outcome without running that skill. | Agents & orchestrationMore projects & source code | |
| jev-prefgithub.com | 10 | Linter that turns AGENTS.md preferences into semantic rules Jev checks against hunks, staged files or pull requests, then feeds the findings back to your coding agent, which also handles the setup. | Agents & orchestrationMore projects & source code | |
| jev-preflightgithub.com | 1 | Claude Code plugin that snapshots a Git baseline at turn start and, at Stop, scores the redacted turn diff on eight risk axes in one request, asking Claude to investigate once more when risk is high. | Agents & orchestrationMore projects & source code | |
| jev-previewgithub.com | 0 | TUI Sandbox for Typesafe Jev API | Agents & orchestrationMore projects & source code | |
| jev-recipesgithub.com | 13 | 200+ plug-and-play Jev recipes: small, calibrated AI decisions that route, grade, gate, compare, and label text for agents, RAG, support, code review, and music. Import from JavaScript or TypeScript, or call the CLI with JSON from any language. | Agents & orchestrationMore projects & source code | |
| jev-recipes — READMEgithub.com | 13 | Reusable Jev recipes for routing, reranking and gates in agents, RAG and support workflows. Recipes call TypeSafe System One or an injected compatible client and return structured decisions; the quick demos use saved responses rather than proving live accuracy. | Agents & orchestrationMore projects & source code | |
| jev-routegithub.com | 3 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-route — READMEgithub.com | 3 | A Python router starts with TypeSafe Jev cloud decisions and describes distilling accumulated traffic into a local routing model. | Agents & orchestrationMore projects & source code | |
| jev-router — READMEgithub.com | 8 | A local proxy for Claude Code and Codex that asks TypeSafe Jev which model tier each human turn needs, while forwarding the main conversation to its selected LLM. | Agents & orchestrationMore projects & source code | |
| jev-router — READMEgithub.com | 513 | Jev Router integrates with Claude Code and Codex to choose a model for each fresh user turn while leaving each upstream CLI interface intact. | Agents & orchestrationMore projects & source code | |
| jev-router — READMEgithub.com | 14 | Jev routing is optional: without a TypeSafe key a rules baseline routes requests; setting TYPESAFE_API_KEY enables Jev. | Agents & orchestrationMore projects & source code | |
| jev-router — READMEgithub.com | 9 | The router uses Jev to classify query difficulty before routing a request among configured LLM tiers. | Agents & orchestrationMore projects & source code | |
| jev-router — READMEgithub.com | 5 | A Claude Code router using TypeSafe Jev to choose model and reasoning effort per user message. Included for semantic routing, not verified automatic switching or savings. | Agents & orchestrationMore projects & source code | |
| jev-routing — READMEgithub.com | 1 | jev-routing asks Jev for tool-selection and completion decisions in its Go coding-agent proxy; without a key, hybrid/local modes use an on-device classifier. | Agents & orchestrationMore projects & source code | |
| jev-royal — READMEgithub.com | 3 | A Clash Royale macOS agent uses Jev to choose among code-validated tactics; without an API key it falls back to local tactics alone. | Agents & orchestrationMore projects & source code | |
| jev-scoutgithub.com | 1 | Agent Skill that finds where TypeSafe Jev decision models would pay off in a codebase. Works in Claude Code, Codex, Cursor and other Agent Skills hosts. | Agents & orchestrationMore projects & source code | |
| jev-scout — READMEgithub.com | 1 | An Agent Skill for identifying bounded decisions where TypeSafe Jev might fit an existing repository, grounded in code and documentation. It proposes opportunities rather than proving performance gains. | Agents & orchestrationMore projects & source code | |
| jev-searchgithub.com | 5 | Fast deep research for the pi coding agent: reads up to 100 pages in full per round, Jev keeps only the passages that answer your questions. | Agents & orchestrationMore projects & source code | |
| jev-search — READMEgithub.com | 5 | jev-search for the pi agent uses Jev to judge page parts and forwards selected passages to the answer model. | Agents & orchestrationMore projects & source code | |
| jev-semantic-cost-mapgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-semantic-cost-map — READMEgithub.com | 0 | The path-planning demo asks Jev four typed questions for each named zone and converts its answers into a cost layer. | Agents & orchestrationMore projects & source code | |
| jev-skillgithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-skill — READMEgithub.com | 1 | The agent skill includes a CLI example that asks Jev whether each tracked file belongs to a defined category such as authentication logic. | Agents & orchestrationMore projects & source code | |
| jev-skill-gategithub.com | 7 | Cut Claude Code's skill manifest by ~75% with TypeSafe Jev. Scores every installed skill for relevance and hides the rest via skillOverrides — 12,750 → 3,185 tokens on a 217-skill install, for $0.0009 a session. | Agents & orchestrationMore projects & source code | |
| jev-skill-gategithub.com | 7 | Cut Claude Code's skill manifest by ~75% with TypeSafe Jev. Scores every installed skill for relevance and hides the rest via skillOverrides — 12,750 → 3,185 tokens on a 217-skill install, for $0.0009 a session. | Agents & orchestrationMore projects & source code | |
| jev-skill-routergithub.com | 0 | Route each task to the one tool an agent should call, with Jev — with confidence. Single-file, no build. Use free: jevtypesafeai.com/tools/jev-skill-router | Agents & orchestrationMore projects & source code | |
| jev-skill-router — READMEgithub.com | 0 | Browser task-router demo uses Jev to select one agent tool per task and export the routing decisions as JSON. | Agents & orchestrationMore projects & source code | |
| jev-skill-router — READMEgithub.com | 3 | The local MCP skill router asks TypeSafe Jev which instructions fit a request; its README says live routing quality and client use are still being validated. | Agents & orchestrationMore projects & source code | |
| jev-skill-router — READMEgithub.com | 6 | A Claude Code hook/reference study using Jev to recommend installed skills. Default shadow mode logs decisions; a prompt hook cannot remove the existing skill listing, and the author reports discontinuing this setup. | Agents & orchestrationMore projects & source code | |
| jev-skill-selectiongithub.com | — | Context GC: Pre-message hook: use TypeSafe Jev to keep/drop skills and shrink agent context | Agents & orchestrationMore projects & source code | |
| jev-skillful — READMEgithub.com | 5 | Skillful routes coding-agent capability suggestions using TypeSafe System One with the jev-latest default, according to its README and client type definition. | Agents & orchestrationMore projects & source code | |
| jev-skillful — types.tsgithub.com | 5 | Skillful routes coding-agent capability suggestions using TypeSafe System One with the jev-latest default, according to its README and client type definition. | Agents & orchestrationMore projects & source code | |
| jev-skillsgithub.com | 1 | Claude Code & Codex skills powered by Jev, TypeSafe's System One model. 256 calibrated judgements for $0.0005 in 0.72s — 360x cheaper than Claude Opus 5. Includes the first published Jev calibration curve, measured on 4,995 real agent decisions. | Agents & orchestrationMore projects & source code | |
| jev-skillsgithub.com | 6 | Skills without the context tax. Claude Code and Codex plugin: TypeSafe's Jev decides on every turn which skills the model sees. Always-on context cost: 0 tokens. | Agents & orchestrationMore projects & source code | |
| jev-skillsgithub.com | 5 | Jev-powered decision skills for Claude Code and Codex. Opt-in, advisory, fail-open. | Agents & orchestrationMore projects & source code | |
| jev-skills — READMEgithub.com | 6 | A Claude Code/Codex plugin that asks TypeSafe Jev which skills should enter context each turn. It separates skill selection from the coding model rather than replacing that model. | Agents & orchestrationMore projects & source code | |
| jev-skills — READMEgithub.com | 5 | Installable Claude Code/Codex skills that use bounded Jev choices for selecting skills, models and investigation steps. Dry-run is offline; live decisions require explicit TypeSafe access and do not replace the coding model. | Agents & orchestrationMore projects & source code | |
| jev-skills — READMEgithub.com | 1 | jev-skills provides installable Claude Code and Codex skills that use TypeSafe Jev for coding-agent decisions; no benchmark claim is repeated here. | Agents & orchestrationMore projects & source code | |
| jev-smart-router — READMEgithub.com | 4 | A Databricks app uses TypeSafe Jev to choose which model should answer each message. | Agents & orchestrationMore projects & source code | |
| jev-social — READMEgithub.com | 132 | Jev selects social research operations such as searching posts and reading comments; socai executes the chosen CLI command in Chrome and returns results for another Jev decision. | Agents & orchestrationMore projects & source code | |
| jev-starter — READMEgithub.com | 2 | jev-starter describes an application layer around Jev and lists deterministic mock providers. | Agents & orchestrationMore projects & source code | |
| jev-stategithub.com | 6 | Build and regression-test conversational state machines powered by Jev. Inspect decisions, capture failing conversations as tests, and export runnable TypeScript for your app. | Agents & orchestrationMore projects & source code | |
| jev-state — READMEgithub.com | 6 | Jev State is a visual editor and regression runner for conversational state machines: Jev proposes the next typed state, while XState transitions and confidence checks control movement. Simulation values are fixtures, not accuracy estimates. | Agents & orchestrationMore projects & source code | |
| jev-subagent-router — READMEgithub.com | 0 | A Codex/Claude Code dispatcher asks Jev to choose a model and reasoning effort based on task complexity. | Agents & orchestrationMore projects & source code | |
| jev-superpowersgithub.com | 35 | Skills framework for coding agents such as Claude Code, Codex and Cursor that calls Jev at workflow checkpoints to vet dependencies before install, gate task completion and steer debugging decisions. | Agents & orchestrationMore projects & source code | |
| jev-superpowers — READMEgithub.com | 35 | jev-superpowers documents two decision backends: the TypeSafe Jev API or local Laya. | Agents & orchestrationMore projects & source code | |
| jev-support-agentsgithub.com | 0 | Multi-agent customer support API where the LLMs write the text and a typed decision model makes the decisions. Jev (TypeSafe System One) handles routing, response evaluation and retry/escalation; FastAPI + Ollama. Includes a 60-ticket benchmark against LLM free-text and structured-JSON baselines. | Agents & orchestrationMore projects & source code | |
| jev-support-agents — READMEgithub.com | 0 | A customer-support API uses LLM agents for text and Jev for routing; the README says its tests fake both services. | Agents & orchestrationMore projects & source code | |
| jev-switchboardgithub.com | 1 | A JEV-gated semantic communication layer for parallel coding agents. | Agents & orchestrationMore projects & source code | |
| jev-switchboard — READMEgithub.com | 1 | Claude Code and Codex communication gate uses Jev to select recipients, urgency and evidence for consequential agent events rather than broadcasting transcripts. | Agents & orchestrationMore projects & source code | |
| jev-synergy-screeninggithub.com | 0 | Jev (TypeSafe System One) × ASReview SYNERGY abstract screening demo — Choice/Noul vs gold labels. | Agents & orchestrationMore projects & source code | |
| jev-telegram-channel-router — READMEgithub.com | 0 | This Telegram bot filters public-channel posts with Jev judgments and routes matching posts to configured chats or channels. Its README describes natural-language template setup and one shared judgment per post; these are project claims, not a live test. | Agents & orchestrationMore projects & source code | |
| jev-terrarium-dinosaurs — READMEgithub.com | 0 | The dinosaur terrarium keeps physics and legality in code while Jev judges each creature’s intent from prose world laws. | Agents & orchestrationMore projects & source code | |
| jev-tetrisgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jev-tetris — READMEgithub.com | 0 | A Playwright loop asks Jev to choose a Tetris placement; no-key mock mode picks the top heuristic placement instead. | Agents & orchestrationMore projects & source code | |
| jev-tool-router — READMEgithub.com | 7 | An unofficial Codex MCP router uses Jev to narrow tool discovery on demand instead of exposing hundreds of external tool schemas at once. | Agents & orchestrationMore projects & source code | |
| jev-trader — READMEgithub.com | 2,721 | The trading demo distinguishes Jev-backed decisions from a default momentum mock and describes simulated fills in its no-private-key dry run. | Agents & orchestrationMore projects & source code | |
| JEV-Trading-BOT — READMEgithub.com | 1 | An experimental order-flow trading bot turns code-computed market features into typed Jev questions and decision gates; inclusion does not endorse profitability or safety. | Agents & orchestrationMore projects & source code | |
| jev-triagegithub.com | 1 | Near-free GitHub issue triage powered by TypeSafe AI's Jev - typed, confidence-gated labels that escalate only the uncertain cases. | Agents & orchestrationMore projects & source code | |
| jev-triagegithub.com | 1 | Millisecond-class test-failure triage for coding agents: RETRY / FIX_CODE / FIX_ENV, powered by TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| jev-triage — READMEgithub.com | 1 | A GitHub issue-labeling action uses Jev label/confidence judgments and escalation; this repository is now a Marketplace wrapper around code moved to jev-tools. | Agents & orchestrationMore projects & source code | |
| jev-triage — READMEgithub.com | 1 | A test-failure triage tool using TypeSafe Jev to choose RETRY, FIX_CODE or FIX_ENV, providing an action class rather than generating a code repair. | Agents & orchestrationMore projects & source code | |
| jev-ui-mapgithub.com | 0 | Map your app's UI so Jev can drive it. Demo: order food by voice. Built with AdaL. | Agents & orchestrationMore projects & source code | |
| jev-ui-map — READMEgithub.com | 0 | Food-ordering browser demo maps screens into typed choices for Jev and executes selected actions in code, with text or voice input; voice mode does not place the final order. | Agents & orchestrationMore projects & source code | |
| jev-uipath-coded-agent — READMEgithub.com | 2 | This UiPath AML-triage demo splits language work from decisions: the LLM reads/writes while TypeSafe Jev supplies typed risk and disposition judgments. | Agents & orchestrationMore projects & source code | |
| jev-un-squadrongithub.com | 0 | A TypeSafe Jev decision model flying U.N. Squadron (SNES, 1991) from raw WRAM via BizHawk + Lua + Python | Agents & orchestrationMore projects & source code | |
| jev-un-squadron — READMEgithub.com | 0 | A Python controller asks TypeSafe Jev for moves in a SNES U.N. Squadron emulator using raw game memory rather than screen pixels; README reports the experiment’s limits. | Agents & orchestrationMore projects & source code | |
| jev-usegithub.com | 39 | Plugin for Claude Code, Codex and pi that hands agent steps needing no text output to Jev, at about 230 ms p50 and $0.02 per 1,000 judgments, with typed escalation back to the LLM when writing is needed. | Agents & orchestrationMore projects & source code | |
| jev-use — READMEgithub.com | 39 | A Claude Code/Codex/pi helper routes no-text decisions to Jev while an LLM writes; README also offers a local mock backend. | Agents & orchestrationMore projects & source code | |
| jev-usecases — READMEgithub.com | 18 | A use-case guide describes typed software decisions and an initial implementation of many decisions listed for Jev; timing claims are README-reported. | Agents & orchestrationMore projects & source code | |
| jev-ushergithub.com | 1 | The doorman for Claude’s context window. JEV-powered model routing and recoverable context filtering for Claude Code. | Agents & orchestrationMore projects & source code | |
| jev-usher — READMEgithub.com | 1 | jev-usher uses TypeSafe Jev relevance judgments plus deterministic rules to select tool-output context; the README warns filtering can omit useful information. | Agents & orchestrationMore projects & source code | |
| jev-verifier-skillgithub.com | 0 | Fast 'System One' reflex for reasoning LLMs: typed probabilistic second opinions from Jev via OpenCode Zen, with PII-minimizing state redaction. | Agents & orchestrationMore projects & source code | |
| jev-verifier-skill — READMEgithub.com | 0 | An OpenCode skill uses Jev for bounded probabilistic second opinions; the main model keeps reasoning and tool work, and Jev’s output is advisory rather than proof. | Agents & orchestrationMore projects & source code | |
| jev-voice — READMEgithub.com | 107 | Jev Voice maps locally transcribed speech to a selected action: code supplies candidate values, Jev chooses, and code executes. | Agents & orchestrationMore projects & source code | |
| jev-voice-agent — READMEgithub.com | 0 | A realtime voice agent routes speech from Flux STT through Jev to Flux TTS; Jev selects among human-written responses instead of generating text. | Agents & orchestrationMore projects & source code | |
| jev-watchdoggithub.com | 0 | Event-driven trajectory watchdog for coding agents: per-event LLM judgments (Jev, Claude, GPT) accumulated into quarantine decisions | Agents & orchestrationMore projects & source code | |
| jev-watchdog — READMEgithub.com | 0 | A Claude Code thread watchdog uses Jev as its default judge; without enforcement, the README says it only reports would-be quarantines. | Agents & orchestrationMore projects & source code | |
| jev-workflowgithub.com | 0 | Decision contracts, adversarial testing, tracing, stability, and privacy controls for TypeSafe Jev | Agents & orchestrationMore projects & source code | |
| jev-workflow — READMEgithub.com | 0 | Jev workflow tooling compiles YAML decision contracts and adds adversarial tests, trace/replay, temporal-stability gates and declared data-egress controls. | Agents & orchestrationMore projects & source code | |
| jev2mcp — READMEgithub.com | 0 | This experimental ChatGPT/MCP companion uses Jev probabilities to select among enabled tools, leave a prompt unchanged, or hold it for review. | Agents & orchestrationMore projects & source code | |
| jev_router — READMEgithub.com | 0 | Agent Zero plugin batches Jev task/complexity/vision/delegation judgments to route incoming chats among configured model presets, optionally switching agent profiles and falling back to the active preset on failure. | Agents & orchestrationMore projects & source code | |
| Jev_steer_or_queuegithub.com | 2 | Let TypeSafe Jev decide whether a message you send mid-turn should steer, queue, or interrupt your coding agent. Claude Code plugin; Codex CLI in testing. | Agents & orchestrationMore projects & source code | |
| Jev_steer_or_queue — READMEgithub.com | 2 | A coding-agent hook asking TypeSafe Jev whether a new message should steer ongoing work, wait in a queue or stop it. | Agents & orchestrationMore projects & source code | |
| JevArenagithub.com | 0 | Spectator arena web app where two independent Jev agents play Snake against each other through Vercel AI Gateway, falling back to a labeled demo policy when no key is set. | Agents & orchestrationMore projects & source code | |
| jevarena — READMEgithub.com | 2 | A spectator demo has two TypeSafe Jev agents drive separate Snake browsers through indexed direction-button clicks. | Agents & orchestrationMore projects & source code | |
| JevArena — READMEgithub.com | 0 | JevArena runs two independent Jev Snake agents through Vercel AI Gateway when configured; leaving the key empty displays a labeled demo-policy match. | Agents & orchestrationMore projects & source code | |
| Jevatargithub.com | — | Decision Tools: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details. | Agents & orchestrationMore projects & source code | |
| Jevbridgegithub.com | 45 | ACP and MCP adapter plus CLI that puts Jev's typed decisions beside Codex, Claude, Grok and OpenCode for routing, gating, scoring and computer use, with an offline rule backend. | Agents & orchestrationMore projects & source code | |
| jevcgithub.com | 8 | Compiler that turns natural-language agent rules and JSON Schemas into Jev programs: a few narrow typed evidence questions plus a code reducer that computes the verdict, so a rule like never commit unless asked becomes testable. | Agents & orchestrationMore projects & source code | |
| jevchatgithub.com | 1 | ChatGPT-style chat UI where Jev never writes text but picks the answer from Yes/No, Maybe, Pirate, Tabloid or custom answer sets, and even picks the chat title from fragments of your question. | Agents & orchestrationMore projects & source code | |
| JevCodegithub.com | — | Decision Tools: JevCode - Jev can code. We want to dogfood JevCode | Agents & orchestrationMore projects & source code | |
| jevcore — READMEgithub.com | 1 | JevCore uses TypeSafe Jev for coding-task classification, risk, execution-mode, and security-review judgments, with a deterministic offline RuleProvider. | Agents & orchestrationMore projects & source code | |
| jevdevicegithub.com | 0 | MCP harness that turns a plain-language goal into exactly one verified action on an Android phone over adb or a local shell, with Jev (or local Laya) picking from live-discovered candidates or declining. | Agents & orchestrationMore projects & source code | |
| jevdevice — READMEgithub.com | 0 | Device-automation experiment reads current Android or shell state, uses hosted Jev or local Laya to narrow real action candidates, safety-gates execution and records verification/judgment data. | Agents & orchestrationMore projects & source code | |
| jevelrygithub.com | 7 | Use Jev everywhere to make & track decisions | Agents & orchestrationMore projects & source code | |
| jevelry — READMEgithub.com | 7 | Jevelry lets code and coding agents ask TypeSafe Jev typed questions; the README labels the project early development. | Agents & orchestrationMore projects & source code | |
| Jevexgithub.com | 3 | Agent loop experiment where Jev picks each next action and approves the concrete call, a LangChain chat model fills arguments and the final reply, and three local MCP servers expose 12 tools. | Agents & orchestrationMore projects & source code | |
| JevFlowgithub.com | 7 | A Claude Code plugin that keeps your AI agents honest. Claude plans tasks as phases with checks, and when it tries to stop, Jevflow runs them and asks Jev if the work is really done. Many agents, one plan, with a live viewer to watch it all. | Agents & orchestrationMore projects & source code | |
| JevFlow — READMEgithub.com | 7 | A Claude Code planning/completion plugin that runs phase checks and asks Jev to judge completion when the agent tries to stop. The plugin is not proof that every task is correctly completed. | Agents & orchestrationMore projects & source code | |
| jevgauge — READMEgithub.com | 1 | A preview Hermes Desktop plugin asks Jev to select a model and reasoning tier when starting a routed chat; automatic composer routing requires an optional host hook. | Agents & orchestrationMore projects & source code | |
| JevHarnessgithub.com | 1 | Pre-execution gate for a single agent tool call: one mixed Jev request returns an allow/ask/deny verdict, irreversibility score and yes/no checks, and Python code turns them into execute, confirm or reject. | Agents & orchestrationMore projects & source code | |
| Jevifygithub.com | 36 | Agent skill that teaches a coding agent to spot work Jev can handle, design Noul, Choice and Score question packs, and learn from recent community experiments, with a client script for API-backed test runs. | Agents & orchestrationMore projects & source code | |
| jevify — READMEgithub.com | 36 | Jevify teaches an agent which routine decisions a fast decision model can handle, then supplies tools for acting on them. | Agents & orchestrationMore projects & source code | |
| jevkitgithub.com | 2 | Zero-dependency CLI and MCP server with eleven Jev-backed tools such as route, triage, guard, grep, rank, compact and judge, plus six runnable use cases including a Claude Code guard hook and a PR ranker. | Agents & orchestrationMore projects & source code | |
| jevletgithub.com | 0 | Jev-powered System One operator for Kubernetes: observable, typed, policy-gated cluster decisions and automated remediation. | Agents & orchestrationMore projects & source code | |
| jevlet — READMEgithub.com | 0 | Go Kubernetes control-loop experiment combines cluster signals for Jev closed-schema decisions, then deterministic policy code authorizes and actuates; the model does not generate commands or YAML. | Agents & orchestrationMore projects & source code | |
| JevLoopgithub.com | 2 | The agent loop where decisions don't cost a large language model call. Zero deps, runs offline, no API key needed. | Agents & orchestrationMore projects & source code | |
| JevLoop — READMEgithub.com | 2 | JevLoop can use the hosted Jev API when TYPESAFE_API_KEY is set; the README says it otherwise uses local Laya. | Agents & orchestrationMore projects & source code | |
| jevloop — READMEgithub.com | 6 | An independent agent runtime assigning bounded tool decisions to Jev, authored language to an LLM and execution/state to deterministic code. | Agents & orchestrationMore projects & source code | |
| JevLoop — READMEgithub.com | 21 | JevLoop routes decisions to Jev or Laya; its README demo defaults to a Laya-to-rule-judge fallback path. | Agents & orchestrationMore projects & source code | |
| jevmapgithub.com | 1 | Map a codebase into units and let Jev (TypeSafe AI) hand an AI coding agent the ten files that matter for a task | Agents & orchestrationMore projects & source code | |
| jevme — READMEgithub.com | 5 | jevme uses TypeSafe Jev for routine voice-command decisions on macOS and reserves a separate LLM for commands needing deeper reasoning. | Agents & orchestrationMore projects & source code | |
| Jevmindgithub.com | 183 | Zero-dependency Python toolkit that pulls an agent's decisions out of prose into typed, confidence-scored answers behind a code gate, logged in a graded ledger, with a local rules brain or Jev behind one flag. | Agents & orchestrationMore projects & source code | |
| Jevmind — READMEgithub.com | 183 | Jevmind is an offline-first agent toolkit that can optionally swap in Jev for decisions that exceed its local rules; do not describe Jev as its only or default brain. | Agents & orchestrationMore projects & source code | |
| jevmory — READMEgithub.com | 11 | A coding-agent memory layer storing verbatim facts and receipts. Optional TypeSafe Jev grading assesses candidate facts while application code selects and composes memory; local storage does not imply that grading avoids remote model requests. | Agents & orchestrationMore projects & source code | |
| Jevoniangithub.com | 15 | Local OpenAI, Anthropic, and Responses-compatible proxy that serves one Jev call per turn to answer both the model route and the thinking level for its virtual model jevonian/auto, with code filtering candidates by protocol, context window, effort floor, and spent quota windows first, and pinned models or explicit routes skipping Jev entirely. | Agents & orchestrationMore projects & source code | |
| jevons — READMEgithub.com | 0 | Jevons is an explicitly experimental personal Pi extension that uses TypeSafe Jev for skill selection, verification checks, and recovery suggestions; the author gives no usability or correctness guarantee. | Agents & orchestrationMore projects & source code | |
| JevOpsgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| JevOps — READMEgithub.com | 0 | PlantPal checks plant-message relevance with Jev before invoking a domain-specific generative model; application code retains routing control. | Agents & orchestrationMore projects & source code | |
| JevPlanegithub.com | 0 | An audited decision control plane for Gemini agents. Use Jev to route models, preflight prompts and tool evaluate agent runs with Google ADC, structured reports, and hash-chained traces.calls, select bounded context, and | Agents & orchestrationMore projects & source code | |
| JevPlane — READMEgithub.com | 0 | JevPlane pairs Jev’s typed decisions with Gemini generation and reasoning; the README says it does not automatically execute tools or skills. | Agents & orchestrationMore projects & source code | |
| jevragithub.com | 0 | An open-source decision layer connecting coding agents with TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| JevRAGgithub.com | 0 | Option-ready retrieval for decision models on the CloudBTL landing layer | Agents & orchestrationMore projects & source code | |
| JevRAG — READMEgithub.com | 0 | Retrieval framework uses Jev to choose successive folders, documents, pages or stopping actions from small option sets, with descriptors/provenance and a keyless heuristic mode. | Agents & orchestrationMore projects & source code | |
| JevREgithub.com | 0 | Jev-powered decision layer for Android reverse engineering. Recommends the next investigation step from APK metadata or structured state, with an explicit offline Mock mode. | Agents & orchestrationMore projects & source code | |
| JevRE — READMEgithub.com | 0 | JevRE explores Jev as a decision layer in an Android reverse-engineering workflow; README says Jev chooses what to do next rather than reverse-engineering APKs itself. | Agents & orchestrationMore projects & source code | |
| JevRevgithub.com | 700 | Think 100. Run 5. Ship 1. | Agents & orchestrationMore projects & source code | |
| JevRev — READMEgithub.com | 700 | JevRev is a workflow CLI in which Jev sifts proposals and judges evidence while a separate coding agent implements alternatives. | Agents & orchestrationMore projects & source code | |
| jevriel — READMEgithub.com | 6 | A Codex/Claude skill and plugin teaches agents to add Jev decisions to existing workflows and compare outcomes; the README says frontier evaluation is still in progress. | Agents & orchestrationMore projects & source code | |
| JevRoutergithub.com | 330 | Models, subagents, skills, MCP tools, and CLIs as one candidate set; Jev picks, the router enforces permissions and risk; reports 44 percent first-five tool-call hits against 24 for DeepSeek on Toolathlon. | Agents & orchestrationMore projects & source code | |
| JevRouter — READMEgithub.com | 330 | JevRouter gives a candidate set to Jev for one typed Choice; deterministic checks then enforce availability, permissions, risk, and confirmation. | Agents & orchestrationMore projects & source code | |
| jevrouter — READMEgithub.com | 1 | JevRouter offers TypeSafe Jev decision routing; without the API key, its README says it uses a deterministic keyword mock. | Agents & orchestrationMore projects & source code | |
| jevscriptgithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jevscript — READMEgithub.com | 1 | Jevscript is a programming language that treats Jev’s typed semantic judgments as runtime operations; Hocket batches shared-state questions. | Agents & orchestrationMore projects & source code | |
| jevselector — READMEgithub.com | 3 | An OpenCode v2 plugin asks Jev which skills and tools fit each turn, then hides irrelevant items for that request without unregistering them globally. | Agents & orchestrationMore projects & source code | |
| jevshiftgithub.com | 0 | ⚡ Token Saver & Fast Decision Layer for OpenAI Codex — powered by Jev | Agents & orchestrationMore projects & source code | |
| jevskill — READMEgithub.com | 2 | A coding-agent skill asks Jev to shortlist relevant lines from large logs before the agent reads them; the README’s savings figures are project-reported. | Agents & orchestrationMore projects & source code | |
| JevTicktRouter — READMEgithub.com | 2 | A .NET/React support-ticket app asks Jev scoped questions and has a deterministic Mock mode; its code applies the final ticket rules. | Agents & orchestrationMore projects & source code | |
| Jevvygithub.com | 19 | Plugin collection for OpenCode and Claude Code whose permissions plugin auto-approves harmless shell requests and leaves anything uncertain to the agent's normal permission flow. | Agents & orchestrationMore projects & source code | |
| jevvy — READMEgithub.com | 19 | Jevvy Permissions is a coding-agent permission hook: it reviews unresolved shell approvals and leaves existing allow/deny decisions final. | Agents & orchestrationMore projects & source code | |
| jevwiregithub.com | 21 | Jev decision layer for agent harnesses: an MCP server with seven tools (rank, pick, verify, evaluate, gate action, next step, list models), an embeddable DecisionModel library and an escalate-only Claude Code plugin. | Agents & orchestrationMore projects & source code | |
| jevwire — READMEgithub.com | 21 | A Jev integration package combines MCP tools, an embeddable decision library and Claude Code hooks; judgments remain inactive without a configured key. | Agents & orchestrationMore projects & source code | |
| jevxgithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jevx — READMEgithub.com | 1 | An agent skill and CLI for bounded Jev-style yes/no, choices and ratings, usable with hosted TypeSafe Jev or a self-hosted System One endpoint. | Agents & orchestrationMore projects & source code | |
| jevymarket-directgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| jevymarket-direct — READMEgithub.com | 0 | jevymarket-direct is a Polymarket bot whose README says Jev supplies the pricing probability through TypeSafe’s direct API; the code handles scanning and execution. | Agents & orchestrationMore projects & source code | |
| jonathanavis96/jev-kitgithub.com | 1 | Everything you need to run TypeSafe's Jev with Claude Code: a tool-call guard, tier guard, file search, browser agent, review, belay, compaction and installers. | Agents & orchestrationMore projects & source code | |
| Jotgithub.com | 19 | General-purpose agent loop where Jev chooses tools and their arguments while Jot executes them, feeds real results back into context and asks again until Jev can reply, with every tool call visible. | Agents & orchestrationMore projects & source code | |
| JudgeJevgithub.com | 0 | DeepThink meets llm as judge. Use heavy prefill caching to run the same request in parallel and let Jev give you the best result. | Agents & orchestrationMore projects & source code | |
| JudgeJev — READMEgithub.com | 0 | Non-streaming LLM fan-out proxy asks several upstream models for candidates and uses Jev to choose a response for the client. | Agents & orchestrationMore projects & source code | |
| judging-with-typesafe — READMEgithub.com | 1 | This Letta-agent skill describes criterion-based judgments through TypeSafe System One/Jev. | Agents & orchestrationMore projects & source code | |
| karnakgithub.com | 0 | Jev picks Claude Code's reasoning effort before every step. Low for routine, high when recovering. | Agents & orchestrationMore projects & source code | |
| karnak — READMEgithub.com | 0 | A Claude Code plugin asks Jev how much reasoning the next step needs, then adjusts the request’s effort accordingly. | Agents & orchestrationMore projects & source code | |
| keeledgithub.com | 2 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| keeled — READMEgithub.com | 2 | An agent harness separates Jev action selection from LLM generation and tool execution; a stub controller supports offline examples/tests. | Agents & orchestrationMore projects & source code | |
| kibugithub.com | 4 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| kibu — READMEgithub.com | 4 | Early macOS desktop-pet assistant uses Jev for bounded choices in file organization, search and renaming, while other tasks can use a separate planner. | Agents & orchestrationMore projects & source code | |
| KISS Sorcar decide toolgithub.com | ≈554 | KISS Sorcar, a general-purpose agent framework for long-horizon tasks, gives the agent a decide tool that asks Jev typed questions to pick its next move. | Agents & orchestrationMore projects & source code | |
| kz-harnessgithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| kz-harness — READMEgithub.com | 1 | Kz-harness is a broader desktop agent harness whose Jev Auto plugin classifies a prompt as work or a question and routes project tasks to an agent. | Agents & orchestrationMore projects & source code | |
| laya-browser-agent — READMEgithub.com | 16 | A browser-driving agent uses local Laya as a TypeSafe Jev alternative; the README says it runs on the user’s own machine. | Agents & orchestrationMore projects & source code | |
| laya-jev — READMEgithub.com | 1 | The README describes Laya as a local inference engine using released weights and reproducing Jev’s request/response format; it does not claim the Laya checkpoint is Jev. | Agents & orchestrationMore projects & source code | |
| laya-jev-GraphRAGgithub.com | 47 | A database-agnostic Agentic GraphRAG framework using swappable System One models (local Laya / cloud Jev). A plug-and-play intelligence layer featuring a complete 4-phase pipeline, continuous evaluation and custom A traversal for any graph database. | Agents & orchestrationMore projects & source code | |
| laya-jev-GraphRAG — READMEgithub.com | 47 | GraphRAG framework separates graph storage from a swappable decision layer using local Laya or cloud Jev. | Agents & orchestrationMore projects & source code | |
| leepokai/jev-guardgithub.com | 50 | Auto mode for every coding agent, built on Jev: risk-scores every tool call with session context (deny / ask / allow), flags prompt injection in results, checks skills and plugins. Claude Code, Codex, Copilot, Gemini, Cursor, pi, OpenCode, ACP. | Agents & orchestrationMore projects & source code | |
| Levgithub.com | 26 | System One decision engine in Clojure on Chez Scheme that serves Laya ModernBERT/mmBERT encoders at about 100 ms on a laptop CPU, plus a llama.cpp thinker whose answers are scored by token log probabilities. | Agents & orchestrationMore projects & source code | |
| lev — READMEgithub.com | 26 | Lev implements typed state-question decisions using encoders from Laya checkpoints. | Agents & orchestrationMore projects & source code | |
| LifeOS Jev orchestrator experimentsgithub.com | — | Six offline experiments in the self-hosted LifeOS personal assistant testing, on recorded real turns, whether Jev typed judgments would help its chat orchestrator with routing, in-loop decisions and latency. | Agents & orchestrationMore projects & source code | |
| LifeOS Jev orchestrator experiments — repogithub.com | — | Six offline experiments in the self-hosted LifeOS personal assistant testing, on recorded real turns, whether Jev typed judgments would help its chat orchestrator with routing, in-loop decisions and latency. | Agents & orchestrationMore projects & source code | |
| lightjev — READMEgithub.com | 0 | Lightjev combines Lightpanda page rendering with Jev decisions about page relevance and which link to follow. | Agents & orchestrationMore projects & source code | |
| limpetgithub.com | 5 | Stop hook for Claude Code and Codex that has Jev score each attempt to stop against your plain-language rules in about 0.7 seconds and sends the agent back to work when a rule is violated. | Agents & orchestrationMore projects & source code | |
| limpet — READMEgithub.com | 5 | limpet is a Stop hook that asks Jev to score an agent’s pending stop against user-written rules. | Agents & orchestrationMore projects & source code | |
| Local Operator classification layergithub.com | — | Agent hub that runs a decision-model layer beside its heuristics once per message to recommend skills, MCP servers and roles; Jev scored 31/31 on labelled cases with the rubric in the options. | Agents & orchestrationMore projects & source code | |
| Local Operator classification layer — docsgithub.com | — | Agent hub that runs a decision-model layer beside its heuristics once per message to recommend skills, MCP servers and roles; Jev scored 31/31 on labelled cases with the rubric in the options. | Agents & orchestrationMore projects & source code | |
| LogiPulseAI_JEV — READMEgithub.com | 0 | LogiPulseAI is a logistics triage example where Jev answers typed parcel-routing questions and application guardrails produce the verdict; sample metrics are not independently confirmed. | Agents & orchestrationMore projects & source code | |
| loom-skill-selector — READMEgithub.com | 0 | An opt-in Loom extension ranks skill previews with batch Jev Noul questions; selection itself does not execute scripts or approve tool calls. | Agents & orchestrationMore projects & source code | |
| maestro-jevgithub.com | 0 | Goal-driven UI verification on iOS/Android/Chromium: Jev (TypeSafe) decides each step, Maestro executes. Installs as a pi / Claude Code / Codex skill. | Agents & orchestrationMore projects & source code | |
| maestro-jev — READMEgithub.com | 0 | maestro-jev pairs Jev’s per-step UI judgments with Maestro driving iOS, Android, or Chromium; it packages an agent-facing skill and verification loop. Compatibility and device behavior were not tested in this review. | Agents & orchestrationMore projects & source code | |
| Mario-And-Jevgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Mario-And-Jev — READMEgithub.com | 0 | A Mario game demo uses Jev for turn decisions and translates them into button presses. | Agents & orchestrationMore projects & source code | |
| Markus decide toolgithub.com | — | AI workforce platform that gives its agents a decision capability and a decide tool backed by Jev-style decision models, with a rule that all evidence must go into the state. | Agents & orchestrationMore projects & source code | |
| mastra-jev-moderation — READMEgithub.com | 7 | This Mastra input-moderation processor uses TypeSafe Jev only when TYPESAFE_API_KEY is set; without the key, no moderation is added. | Agents & orchestrationMore projects & source code | |
| McpMatchgithub.com | 0 | Web app that matches a user goal to the best MCP server or capability in a catalog, using a two-stage Jev pass when the catalog is large. | Agents & orchestrationMore projects & source code | |
| mcpmatch — product.tsgithub.com | 0 | McpMatch uses Jev to choose an MCP server or capability from a catalog described as about 40 fake-but-realistic tools; this is a routing demo. | Agents & orchestrationMore projects & source code | |
| mcts-agent — READMEgithub.com | 7 | An MCTS execution agent pairs TypeSafe Jev's structured Noul, Choice, and Score evaluations with Gemini action generation and a local run-inspection UI. | Agents & orchestrationMore projects & source code | |
| mem-jev — READMEgithub.com | 2 | memJev is a procedural-memory system; Jev is an optional judge for ambiguous candidates after deterministic eligibility gates, not the system’s required engine. | Agents & orchestrationMore projects & source code | |
| memory-router — READMEgithub.com | 0 | A self-hosted memory router uses Jev through OpenRouter to classify facts and grant write leases; its README warns the service has no authentication. | Agents & orchestrationMore projects & source code | |
| messenger-human-firewallgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| messenger-human-firewall — READMEgithub.com | 0 | The README’s documented Active Jev Mode uses Jev/System One to extract typed security signals from incoming Messenger messages. | Agents & orchestrationMore projects & source code | |
| MetaCoggithub.com | 9 | Metacognition loop where a small model branches into several thought paths and a System One judge, Jev or the open Reflex 4B, picks which one to continue; per-candidate Jev Nouls lifted HumanEval from 0.555 to 0.756. | Agents & orchestrationMore projects & source code | |
| MindRoom participation judgmentsgithub.com | — | Multi-agent runtime on Matrix where agents already in a thread decide whether to answer untagged messages, using an LLM or TypeSafe System One with a 0.8 threshold. | Agents & orchestrationMore projects & source code | |
| mja00/cmdc-auto-modegithub.com | 1 | Auto permission mode for Command Code: screens every tool call with TypeSafe Jev before it runs, and rejects anything out of scope. | Agents & orchestrationMore projects & source code | |
| mmmgc_modelroutergithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| mmmgc_modelrouter — READMEgithub.com | 0 | This Pi extension uses TypeSafe Jev to set each turn’s reasoning-effort tier while keeping the user-selected model fixed. | Agents & orchestrationMore projects & source code | |
| mnemon-memory-agent — READMEgithub.com | 3 | A memory agent separating Jev/System One read-time record screening and judgments from the LLM’s search planning and answer composition. | Agents & orchestrationMore projects & source code | |
| model-preflightgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| model-preflight — READMEgithub.com | 0 | The preflight skill uses TypeSafe System One/Jev for typed judgments before coding; without a key, the README says it falls back to session-model judgment. | Agents & orchestrationMore projects & source code | |
| model-proxygithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| model-proxy — READMEgithub.com | 0 | The proxy includes a separate System One decisions endpoint with a TypeSafe Jev provider. | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | 2023Anita/codex-jev-assistant — Codex Plugin with typed Jev judgments for risk review, evidence checks, context screening, and reranking. · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 10 | 24601/Augustus — Agent skill: design judgment-assisted systems with TypeSafe Jev (System One). Maps Choice/Score/Noul onto decision theory, reranking, and routing. Composition algebra, question design, validation gates. MIT. · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 4 | RahulBalakavi/claude-code-jev — Experimental Jev permission gate for Claude Code via OpenRouter, with reproducible latency and cost benchmarks · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | Umbylicus/umby-jev-stack — Portable agent skill: TypeSafe Jev as a cheap code-review classifier (HTTP + optional jev-review MCP) · model_id | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | XieChengYuan/jev-observatory — Jev 活动看板：自动接入 MCP，实时查看真实输入、判断结果与用量。Local-first, passive MCP observability. · model_id · JavaScript | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 2 | bojansandhaus/jev-decisions — Jev Decisions Plugin for Hermes (and other AI Agents): tool risk reviews, human approval recommendations, evidence checks, and a local decision journal. · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | connectedGraph/claude-jev-warden — Real-time quality gate and Art Director Warden for Claude Code powered by TypeSafe Jev 1.13 non-autoregressive decision model · model_id · HTML | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 147 | dbreunig/building-with-jev-skill — A skill for writing and improving programs that call Jev, TypeSafe's System One model · model_id | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | de-niji/jev-hermes — Jev for Hermes: cheap intent gates + verbatim tool compaction on OpenRouter · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 10 | doeixd/jev-pref — Turn your AGENTS.md preferences into a fast, Jev-powered AI linter. · model_id · JavaScript | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | emirbartu/opencode-system-one — Opencode plugin using Jev (system one model) as part of software development process. Not affiliated with Opencode team. · model_id · TypeScript | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 2 | nexibeo/jev-organize — Throw in a pile of company files and get them classified and organized by department, type, sensitivity, date, counterparty and PII, with an index for AI agents. Powered by TypeSafe's Jev on OpenRouter (17¢ per 1,000 files). Zero-dependency Node CLI + Claude skill + Codex agent. · model_id · JavaScript | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 3 | rsdkrasen/hermes-jev-router — TypeSafe/Jev router plugin for Hermes Agent — compact tool results, suppress duplicate tools, skip unnecessary main-model calls · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | vit-cerny/jev-browser-kit — Give any MCP-capable LLM harness an on-demand real-browser search tool (TypeSafe Jev) with per-run timing and cost tracking. · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | wahahaorg/jev-mcp — jev-mcp turns browser-use/jev-ultrafast into one local, session-aware MCP server for fast read-only product research. Jev selects an action and observed target from a DOM snapshot; · model_id · Python | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 11 | whyashthakker/beam-cli — Monitoring & Safety layer for all your agents. Open Source CLI & Skills for Claude Code, Codex, Cursor, Jev and your preferred agents. · model_id · TypeScript | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 1 | wotai-dev/typesafe-jev-tools — A Claude Code hook that asks whether the decision you are writing needs a model at all. Includes a measured 149-row comparison of TypeSafe Jev against Claude Haiku 4.5. · model_id · Shell | Agents & orchestrationMore projects & source code | |
| model_idgithub.com | 554 | wuyoscar/jev-skill — A Jev decision toolbox for agents and people: 56 recipes, an OpenRouter CLI, and transparent paired evaluations. · model_id · Python | Agents & orchestrationMore projects & source code | |
| morcoan/JevSeekgithub.com | 2 | A local coding workspace pairing Jev action routing with DeepSeek argument generation. Native tools, persistent sessions, React desktop, and documented research. | Agents & orchestrationMore projects & source code | |
| mozbz/jev-review-hookgithub.com | 1 | Jev powered code review hook for agents | Agents & orchestrationMore projects & source code | |
| mwijanarko1/pi-jev-agentgithub.com | 1 | pi-jev-agent is an experimental extension for the Pi coding agent. It uses TypeSafe Jev to choose the next tool before each language-model step. | Agents & orchestrationMore projects & source code | |
| n8n-nodes-typesafegithub.com | 0 | TypeSafe for n8n: ask yes/no, pick-one and rate-it questions about any item and route on calibrated answers (Pass / Fail / Review) | Agents & orchestrationMore projects & source code | |
| n8n-nodes-typesafe — READMEgithub.com | 0 | An n8n community node asks TypeSafe Noul, Choice, and Score questions about workflow text or JSON, then routes on typed answers. | Agents & orchestrationMore projects & source code | |
| n8n-nodes-typesafe — READMEgithub.com | 0 | An n8n node asks TypeSafe Jev typed yes/no, choice, or score questions and returns answers that workflows can branch on. | Agents & orchestrationMore projects & source code | |
| neo4jev — READMEgithub.com | 154 | Neo4jev documents graph traversal through one TypeSafe system_one call per hop, combining a Choice over relationships with a Noul goal check. | Agents & orchestrationMore projects & source code | |
| NetClaw jev-auditgithub.com | — | Skill-audit sweeps in the NetClaw network agent that use Jev to flag ambiguous triggers, missing failure behavior, stale assumptions and injection smells, and to map skill overlap and coverage gaps. | Agents & orchestrationMore projects & source code | |
| Nika System One seatgithub.com | — | Workflow language for repeatable AI work whose compiler can hand bounded decisions to TypeSafe System One, one Choice with a "none" option per question. | Agents & orchestrationMore projects & source code | |
| Nitrogithub.com | 5 | Grok Build variant where Jev decides once per turn which of the 25 tools a request needs, trimming tool schemas from about 11K to 2.9K tokens and costing 14 to 40% less on three tasks, two runs each. | Agents & orchestrationMore projects & source code | |
| nitro — READMEgithub.com | 5 | A Grok Build integration asks Jev once per turn which tools are needed and filters the chat model’s tool list; the chat model still writes commands and code. | Agents & orchestrationMore projects & source code | |
| nova-decidegithub.com | — | Typed-decision route in Nova Tools, a set of tools for AI agents on different models to message, wait and swarm together, where Jev answers small choice, score or yes/no judgments beside deterministic machinery. | Agents & orchestrationMore projects & source code | |
| nova-decide — repogithub.com | — | Typed-decision route in Nova Tools, a set of tools for AI agents on different models to message, wait and swarm together, where Jev answers small choice, score or yes/no judgments beside deterministic machinery. | Agents & orchestrationMore projects & source code | |
| oh-my-hermes Jev posturegithub.com | — | Module in the oh-my-hermes plugin that detects Jev-class Hermes plugins on a machine and reports what each declares about its tools, hooks and data flow, without claiming any ran. | Agents & orchestrationMore projects & source code | |
| omp-jev-compaction — READMEgithub.com | 9 | The omp plugin scores tool calls and results with Jev, then truncates those Jev marks as no longer needed. | Agents & orchestrationMore projects & source code | |
| omp-jev-extensionsgithub.com | 1 | OMP / pi-coding-agent extensions that delegate acceptance gating and subagent routing to the Typesafe Jev decision API | Agents & orchestrationMore projects & source code | |
| omp-jev-extensions — READMEgithub.com | 1 | This set of OMP coding-agent extensions connects to the Typesafe Jev typed-decision API. | Agents & orchestrationMore projects & source code | |
| omp-jevens-classifiergithub.com | 0 | Permission gate for the OMP coding agent that sends bash commands and process-spawning eval payloads that would run unseen in yolo mode to Jev, turning dangerous ones into a Run-or-Deny prompt. | Agents & orchestrationMore projects & source code | |
| omp-jevens-classifier — READMEgithub.com | 0 | This OMP plugin adds Jev-judged permission checks for bash commands and process-spawning eval payloads; it does not itself execute a permitted command. | Agents & orchestrationMore projects & source code | |
| omp-typesafegithub.com | 1 | Reviewer plugin for the omp coding agent that watches the session and posts adversarial or advisory notes, plus a typesafe_ask tool exposing Noul, Choice and Score to the agent. | Agents & orchestrationMore projects & source code | |
| open-jev-approvals — READMEgithub.com | 3 | A Claude Code approval hook sends proposed tool calls to Jev, then a local policy returns allow or deny. | Agents & orchestrationMore projects & source code | |
| Open-Science Jev classificationgithub.com | — | Classification service in the AIPOCH Open-Science research workbench that uses Jev to select which skills a request needs and whether a document should be read in full. | Agents & orchestrationMore projects & source code | |
| OpenAgents Jevgithub.com | — | OpenAgents builds general agent infrastructure where every turn starts with a Jev typed judgment that returns a choice and probabilities, foldable back into the round. | Agents & orchestrationMore projects & source code | |
| openclaw-jev-compaction — READMEgithub.com | 1 | An OpenClaw context engine asks Jev which tool calls/results to retain and drops the others without summarizing; user/assistant text remains unchanged. | Agents & orchestrationMore projects & source code | |
| openclaw-jev-plugingithub.com | 2 | Jev as a message gate to determine if agents should respond | Agents & orchestrationMore projects & source code | |
| openclaw-jev-plugin — READMEgithub.com | 2 | An OpenClaw plugin asks Jev whether a bot should answer a message and, optionally, which model tier should handle it; silenced messages can bypass generation. | Agents & orchestrationMore projects & source code | |
| openclaw-plugin-typesafe-aigithub.com | 0 | OpenClaw plugin that uses Jev for group-chat triage on whether to respond, tool-call safety checks, compaction curation and model routing, without modifying the OpenClaw core. | Agents & orchestrationMore projects & source code | |
| opencode-jev-orchestratorgithub.com | 3 | OpenCode plugin that keeps sessions on a cheap sticky parent model to keep the cache warm and escalates only hard turns to stronger models in temporary child subagents. | Agents & orchestrationMore projects & source code | |
| opencode-jev-plugingithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| opencode-jev-plugin — READMEgithub.com | 0 | An OpenCode plugin describes Jev screening for bash commands and user messages; its guardrails fail open on Jev errors, so broken screening does not block work. | Agents & orchestrationMore projects & source code | |
| opencode-jev-routergithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| opencode-jev-router — READMEgithub.com | 1 | This OpenCode plugin uses Jev just before model dispatch to select tools; the README says it has not been published to npm. | Agents & orchestrationMore projects & source code | |
| opencode-jev-router — READMEgithub.com | 8 | An OpenCode plugin asks Jev how much reasoning each Responses API request needs; the selected model carries out the request. | Agents & orchestrationMore projects & source code | |
| opencode-langsearch — READMEgithub.com | 0 | This work-in-progress OpenCode search plugin describes an optional Jev gate for judging result relevance, evidence, and injection attempts. | Agents & orchestrationMore projects & source code | |
| OpenCompanygithub.com | 196 | Agent hive for one-person businesses where Jev decides who a message needs and who picks up a broadcast, falling back to the desk lead without a key. | Agents & orchestrationMore projects & source code | |
| opencompany approval reviewgithub.com | — | Approval reviewer in the opencompany AI workspace where Jev auto-approves routine integration actions only when they match the user's request and look low-risk; destructive, metered or permission changes always ask. | Agents & orchestrationMore projects & source code | |
| opencompany approval review — repogithub.com | — | Approval reviewer in the opencompany AI workspace where Jev auto-approves routine integration actions only when they match the user's request and look low-risk; destructive, metered or permission changes always ask. | Agents & orchestrationMore projects & source code | |
| openharnessgithub.com | 1,057 | Follow your curiosity. Build across disciplines. Open-source software and hardware for polymaths in the making. | Agents & orchestrationMore projects & source code | |
| openharness — READMEgithub.com | 1,057 | OpenHarness contains Jev Browser and Jev Sheets agent modules: the browser module uses TypeSafe Jev for decisions, while Sheets sends rows to the Jev API. | Agents & orchestrationMore projects & source code | |
| openharness — READMEgithub.com | 1,057 | OpenHarness contains Jev Browser and Jev Sheets agent modules: the browser module uses TypeSafe Jev for decisions, while Sheets sends rows to the Jev API. | Agents & orchestrationMore projects & source code | |
| OpenSmokegithub.com | 1 | Find the AI agent runs that broke because their environment did: missing keys, tools, files, permissions, network, or context. Powered by TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| openthai-systemone — READMEgithub.com | 65 | OpenThai-SystemOne presents a Thai/English local typed-decision model whose README says its API contract mirrors TypeSafe’s Jev endpoint; compatibility is stated, not independently tested here. | Agents & orchestrationMore projects & source code | |
| openvons — READMEgithub.com | 13 | An independent local finite-choice model project inspired by Jev-like decisions; its README traces “open-Jev” to Jevons and describes probability-bearing options, not TypeSafe Jev itself. | Agents & orchestrationMore projects & source code | |
| paperclip-plugin-jevgithub.com | 1 | Community Paperclip plugin for Jev-powered issue triage and automatic routing. | Agents & orchestrationMore projects & source code | |
| paperclip-plugin-jev — READMEgithub.com | 1 | This Paperclip community plugin adds Jev-based issue routing, prioritization, and bounded browser-action decisions; its README says the plugin itself does not drive a browser session. | Agents & orchestrationMore projects & source code | |
| paseo-spacedockgithub.com | 4 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| paseo-spacedock — READMEgithub.com | 4 | The Paseo Spacedock plugin offers an advisory gate judgment using TypeSafe System One, Jev by default; the captain records the decision. | Agents & orchestrationMore projects & source code | |
| permit-mcp — READMEgithub.com | 1 | An MCP execution-control layer with deterministic policies, optional Jev decisions and one-use permits. Its local mocks and configured-upstream policy do not imply live Jev validation or a general OS sandbox. | Agents & orchestrationMore projects & source code | |
| Pi Agent Foremangithub.com | 0 | Pi extension that reviews the final exchange after every settled run and, when the agent admits it stopped before required work was done, sends one firm instruction to finish the job. | Agents & orchestrationMore projects & source code | |
| Pi Jev Guardgithub.com | 20 | Pi coding-agent plugin with rules you add by timing, local match or Jev judgment, and action, where Jev checks tool calls for destructiveness, data exfiltration, task drift, and rule conflicts. | Agents & orchestrationMore projects & source code | |
| pi-advisorgithub.com | 18 | Advisor and Executor plugin for the Pi coding agent that consults a stronger model for consequential decisions, with an optional Jev filter that skips low-stakes consultations and a Jev turn gate that pulls the Advisor in. | Agents & orchestrationMore projects & source code | |
| pi-ask-jeff — READMEgithub.com | 0 | pi-ask-jeff documents Jeff as a Jev-compatible System One advisor exposed through pi’s ask_jeff tool; compatibility is the project’s description. | Agents & orchestrationMore projects & source code | |
| pi-extensionsgithub.com | 638 | A monorepo of Pi Coding Agent extensions | Agents & orchestrationMore projects & source code | |
| pi-extensions — READMEgithub.com | 638 | The pi-typesafe-compact extension uses Jev to select older history before Pi's native compaction flow summarizes it. | Agents & orchestrationMore projects & source code | |
| pi-fast-jev-compaction — READMEgithub.com | 3 | pi-fast-jev-compaction asks Jev whether old tool calls and results are needed, pruning or truncating only those items. | Agents & orchestrationMore projects & source code | |
| pi-jevgithub.com | 59 | Pi coding-agent extension that uses Jev to activate only the tools a prompt needs, suggest matching SKILL.md skills and run typed Choice, Noul and Score judgments from inside the agent. | Agents & orchestrationMore projects & source code | |
| pi-jevgithub.com | 0 | Jev-powered semantic process conditions and skill suggestions for Pi. | Agents & orchestrationMore projects & source code | |
| Pi-Jevgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| pi-jevgithub.com | 0 | a Jev classification layer for the Pi coding agent | Agents & orchestrationMore projects & source code | |
| pi-jevgithub.com | 152 | Checks coding-agent tool calls with Nouls for destructive, exfiltrating, and out-of-scope actions plus an impact Score, and screens outputs for leaked secrets. | Agents & orchestrationMore projects & source code | |
| pi-jev (compaction and routing)github.com | 9 | Pi extensions that use Jev to compact context by removing obsolete tool outputs while keeping the rest verbatim, and to route each turn to a cheaper or stronger model by difficulty. | Agents & orchestrationMore projects & source code | |
| pi-jev — READMEgithub.com | 59 | pi-jev adds Jev-powered typed evaluation, tool and skill discovery to the Pi coding agent; its automatic routing features are opt-in and off by default. | Agents & orchestrationMore projects & source code | |
| pi-jev — READMEgithub.com | 0 | One Pi extension asks Jev to evaluate user-defined process-output conditions while the process is running. | Agents & orchestrationMore projects & source code | |
| pi-jev — READMEgithub.com | 9 | A Pi extension combines optional Jev-based context compaction with per-turn model routing; if Jev is unavailable, Pi retains its normal compaction or current model. | Agents & orchestrationMore projects & source code | |
| Pi-Jev — READMEgithub.com | 0 | A Pi plugin brings TypeSafe Jev’s typed decision tools and a /jev panel into the coding agent, and documents its terminology and integration design. | Agents & orchestrationMore projects & source code | |
| pi-jev — READMEgithub.com | 0 | The Pi extension makes coding-agent choices explicit Jev decisions; modules have no effect until promoted. | Agents & orchestrationMore projects & source code | |
| pi-jev — READMEgithub.com | 152 | pi-jev adds Jev judgments to Pi: a gate checks bash/write/edit calls before execution, and an output judge evaluates command output. | Agents & orchestrationMore projects & source code | |
| pi-jev-auto-modegithub.com | 30 | Auto mode for the Pi coding agent that auto-approves bash, write and edit tool calls after deterministic deny rules, using Jev for the semantic judgment and failing closed when unsure. | Agents & orchestrationMore projects & source code | |
| pi-jev-auto-mode — READMEgithub.com | 30 | A Pi coding-agent extension applies deterministic permission rules first, then asks Jev to judge unresolved bash/write/edit calls and fails closed on uncertainty. | Agents & orchestrationMore projects & source code | |
| pi-jev-compaction — READMEgithub.com | 1 | This Pi compaction extension asks Jev to judge complete read-only tool pairs, then prunes context without asking Jev to write a summary. | Agents & orchestrationMore projects & source code | |
| pi-jev-continuegithub.com | 0 | Goal-scoped continuous development for Pi, judged by Jev. | Agents & orchestrationMore projects & source code | |
| pi-jev-continue — READMEgithub.com | 0 | pi continuation extension asks Jev which development work to do next, whether to continue and how to answer bounded choices, while pi's main model performs implementation/investigation; human-input or undecidable cases stop forced continuation. | Agents & orchestrationMore projects & source code | |
| pi-jev-controlgithub.com | 1 | System-One control plane for Pi Coding Agent powered by TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| pi-jev-control — READMEgithub.com | 1 | pi-jev-control connects TypeSafe System One judgments to the Pi Coding Agent’s control plane. | Agents & orchestrationMore projects & source code | |
| pi-jev-helm — READMEgithub.com | 1 | Pi Jev Helm uses Jev classification to choose a Pi model; its README says OpenRouter is the default provider and TypeSafe is explicit. | Agents & orchestrationMore projects & source code | |
| pi-jev-permitgithub.com | 1 | Permission gate for the Pi coding agent that has Jev judge every bash, write and edit call before it runs, after local hard-deny rules and read-only fast paths. | Agents & orchestrationMore projects & source code | |
| pi-jev-permit — READMEgithub.com | 1 | A Pi permission gate asks Jev to judge Bash, write, and edit calls before they run. | Agents & orchestrationMore projects & source code | |
| pi-jev-reasoning-router-litegithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| pi-jev-reasoning-router-lite — READMEgithub.com | 0 | Pi extension asks Jev to select a thinking level while leaving the generation model and provider independently configurable. | Agents & orchestrationMore projects & source code | |
| pi-jev-routergithub.com | 0 | Jev model integration for pi coding agent. | Agents & orchestrationMore projects & source code | |
| pi-jev-router — READMEgithub.com | 16 | A Pi extension lets Jev choose a model and reasoning effort for the session, with generation delegated to the user’s existing Pi providers. | Agents & orchestrationMore projects & source code | |
| pi-jev-router — README.engithub.com | 0 | This Pi extension uses Jev-first classification to choose a model and reasoning level; its English README describes the official TypeSafe provider for direct calls and optional local Laya. Generated answers and tool execution stay with Pi’s selected provider. | Agents & orchestrationMore projects & source code | |
| pi-jev-sentinelgithub.com | 12 | Checks Pi tool calls, tool outputs, and replies for risky actions and prompt injection, with user approvals, context re-checks, secret scrubbing, and optional task pinning. | Agents & orchestrationMore projects & source code | |
| pi-jev-skill-picker — READMEgithub.com | 36 | A Pi extension scores each skill with Jev, then sorts and loads the selected skills. | Agents & orchestrationMore projects & source code | |
| pi-jev-task-router — READMEgithub.com | 3 | pi-jev-task-router uses Jev to choose a model tier for each Pi prompt, but falls back to a keyword heuristic when the key is absent or Jev fails. | Agents & orchestrationMore projects & source code | |
| pi-quiet-askgithub.com | 12 | Rule engine for the Pi coding agent whose packs bind hooks to Jev questions, such as whether a command is destructive, output leaked a key or a claim was verified, with every answer logged and failing open. | Agents & orchestrationMore projects & source code | |
| pi-quiet-ask — READMEgithub.com | 12 | A Pi extension sends selected closed questions to Jev, records judgments and keeps intervention optional; the README says it fails open. | Agents & orchestrationMore projects & source code | |
| pi-software-factorygithub.com | 1 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| pi-software-factory — READMEgithub.com | 1 | Pi Software Factory places Jev gates around semantic classification and routing in its engineering pipeline; other models perform the implementation work. | Agents & orchestrationMore projects & source code | |
| pi-warden — READMEgithub.com | 153 | Pi Warden can use Jev to judge writes against project rules; its README also documents an offline-only mode when no key is supplied. | Agents & orchestrationMore projects & source code | |
| PiJevgithub.com | 4 | Terminal coding agent built on Pi where Jev shortlists skills, ranks file excerpts and triages tool failures while your coding model writes the code; fewer tool calls than plain Pi on SWE-bench django tasks. | Agents & orchestrationMore projects & source code | |
| pikachugithub.com | 0 | Curated Pi coding agent configuration with Jev compaction, LSP, subagents, plan mode, safety tools, and automatic updates. | Agents & orchestrationMore projects & source code | |
| pikachu — READMEgithub.com | 0 | Pikachu is a reproducible Pi coding-agent configuration that bundles Jev-based context compaction and a TypeSafe-backed model-routing controller with other extensions; it is a setup/configuration project, not a new Jev model. | Agents & orchestrationMore projects & source code | |
| Pisper decision servicegithub.com | — | Multi-agent desktop and mobile app that adds a Jev decision service for auto-approving tool permissions above a 0.9 threshold and verifying computer-use actions. | Agents & orchestrationMore projects & source code | |
| pokemon-showdown-jev-agent — READMEgithub.com | 0 | The README describes an autonomous Gen 9 Random Battles agent for Pokémon Showdown using Jev decisions; this review did not execute or verify ladder behavior. | Agents & orchestrationMore projects & source code | |
| primitivesgithub.com | 1 | ThiagaoBR/typesafe_agent_gates — LangChain / Deep Agents middleware that uses TypeSafe's System One model (Jev) for typed judgments in unattended coding agents: a shell-command gate (database, production, destructive, secrets), issue triage and routing by severity and urgency, merge-request detection, and review of weakened tests. Measured with live probes. · primitives · Python | Agents & orchestrationMore projects & source code | |
| ProgressGategithub.com | — | Library that watches recent agent steps for semantic stagnation, where different actions keep chasing the same contradicted assumption; Jev reads the trajectory and a small deterministic policy decides what to do. | Agents & orchestrationMore projects & source code | |
| ps2-ai-agent — READMEgithub.com | 0 | The PS2-emulation agent architecture includes TypeSafe Jev System One as a modular decision engine alongside game telemetry and virtual-controller integration. | Agents & orchestrationMore projects & source code | |
| Public Browser Jev loop — repogithub.com | — | Example loop on the Public Browser Chrome library in which one Jev call per step picks the next action from the page's accessibility refs and gpt-4.1-nano writes text only for type actions, run on six benchmark cards. | Agents & orchestrationMore projects & source code | |
| pydantic-ai — typesafegithub.com | 20,310 | The Pydantic AI repository contains a TypeSafe provider and TypeSafeModel backed by the TypeSafe SDK for Jev typed decisions; this is one specific integration, not a claim that the whole framework runs on Jev. | Agents & orchestrationMore projects & source code | |
| pydantic-ai — typesafe.pygithub.com | 20,310 | The Pydantic AI repository contains a TypeSafe provider and TypeSafeModel backed by the TypeSafe SDK for Jev typed decisions; this is one specific integration, not a claim that the whole framework runs on Jev. | Agents & orchestrationMore projects & source code | |
| Rakazo Jev auto-reviewgithub.com | — | Auto-review adapter in the open-source Rakazo AI-teammate platform that asks Jev whether a bot tool call should auto-pass or be sent to the user for approval. | Agents & orchestrationMore projects & source code | |
| reapergithub.com | 1 | Semantic linter for AI coding agents and CI code review. Detects silent failures, weakened tests, scope creep, unnecessary abstractions, and other semantic code smells. | Agents & orchestrationMore projects & source code | |
| Reflexgithub.com | 5 | Coding agent and personal assistant built on Pi where each state-changing tool call goes through one Jev request of five risk Nouls plus a risk Score, mapped in code to allow, ask or block, and Jev picks the model tier. | Agents & orchestrationMore projects & source code | |
| reflexgithub.com | 3 | Pre-execution risk gate for AI coding agents (Claude Code, Codex, opencode, pi, Hermes). Blocks or asks on risky shell commands, guards against prompt injection in what agents read, escalates to System 2 and a human queue. Optional TypeSafe Jev. | Agents & orchestrationMore projects & source code | |
| reflex — READMEgithub.com | 5 | Reflex adds a TypeSafe Jev reflex layer to its assistant; the README says its code acts on Jev’s outputs. | Agents & orchestrationMore projects & source code | |
| reflex — READMEgithub.com | 3 | A pre-execution coding-agent hook combining deterministic command rules with an optional Jev engine for six typed risk questions. Policy retains execution control; safety claims are not independently certified. | Agents & orchestrationMore projects & source code | |
| reflex-stategithub.com | 0 | Jev-powered execution state for Pi coding agents. Track changes, checks, and blockers outside the main LLM, with evidence-backed updates and replay. Inspired by SKILL.state. | Agents & orchestrationMore projects & source code | |
| reflex-state — READMEgithub.com | 0 | ReflexState gives Pi coding agents an execution-state layer: Jev makes typed semantic decisions while deterministic code extracts facts and applies updates. | Agents & orchestrationMore projects & source code | |
| relay-macos-assistant — READMEgithub.com | 0 | Relay routes requests needing semantic selection through Jev into typed actions, while a local Ollama model handles text generation when needed. | Agents & orchestrationMore projects & source code | |
| RFQ_Routinggithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| RFQ_Routing — READMEgithub.com | 0 | Local fictional CNC quotation-inbox demo uses Jev to judge email and attachment text before Python applies routing rules, not to certify compliance. | Agents & orchestrationMore projects & source code | |
| rh-guardgithub.com | 3 | Coding-agent hook for Claude Code, Cursor, Codex, Pi and others that blocks reward hacking such as tampering with graders or hidden tests, combining structural denies with a Jev risk sidecar. | Agents & orchestrationMore projects & source code | |
| RikkaHub Plus Jev decisionsgithub.com | 19 | Android AI chat client and SillyTavern-compatible app that uses Jev to judge each memory's relevance to the conversation instead of embedding search, and gives the main model a Jev judge tool for yes/no, choice and score questions. | Agents & orchestrationMore projects & source code | |
| risk-analysis-server — READMEgithub.com | 2 | This risk-analysis MCP cookbook uses TypeSafe Jev Noul/Score/Choice gates for triage, query relevance, per-result scoring, and severity. | Agents & orchestrationMore projects & source code | |
| rocky — READMEgithub.com | 0 | Rocky is a macOS voice-control agent built on TypeSafe Jev; Jev selects only from options supplied by the app, while code handles the rest. | Agents & orchestrationMore projects & source code | |
| SAIVerse reflex judgmentgithub.com | — | Reflex-judgment layer in the SAIVerse AI-persona world that returns probabilities instead of prose, first used to rerank auto-recalled memories. It can target TypeSafe Jev, OpenRouter, a self-hosted localjev or a structured-output LLM. | Agents & orchestrationMore projects & source code | |
| SAIVerse reflex judgment — repogithub.com | — | Reflex-judgment layer in the SAIVerse AI-persona world that returns probabilities instead of prose, first used to rerank auto-recalled memories. It can target TypeSafe Jev, OpenRouter, a self-hosted localjev or a structured-output LLM. | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 74 | Bodila51/grok-bot-jev — Connect TypeSafe Jev to Grok Bot as a cheap decision layer - usage gates, skill template, examples · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 3 | CrowdLinker/JevPromptCoach — Claude Code plugin that scores how well you prompt a coding agent, and shows whether your habits are improving. Runs on TypeSafe's Jev model. Zero added latency. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 59 | TheoOliveira/pi-jev — Semantic tool routing and typed System One decisions for the Pi coding agent using TypeSafe Jev · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 25 | blakestone-x/jev-mcp — MCP server for TypeSafe Jev: typed classify, score, check, match and screen for any agent, with confidence on every answer · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 24 | buberlo/dsh-jev — Jev-powered decision layer for DeepSeek Harness · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 7 | buchmark/claude-jev — Claude Code plugin that scores review findings, debug hypotheses and design options with TypeSafe's Jev — calibrated probabilities instead of one more opinion. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 61 | burnigtm/jev-mcp — MCP server that puts TypeSafe Jev on the coding loop in Cursor, Codex, and any MCP client · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 118 | devagrawal09/stanley-code — Bounded TypeSafe Jev workflows for coding agents. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 2 | ibrahemid/git-jev-stage — Select Git changes for staging with a plain-language description. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 154 | jexp/neo4jev — Typesafe.ai System One Model Jev navigating a Neo4j graph by using a classifier over neighbouring relationships · sdk · Jupyter Notebook | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 472 | jkudish/jev-mcp — Fast, cheap, typed judgments from TypeSafe's Jev model, as MCP tools. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 30 | jomatsu/pi-jev-auto-mode — Jev (TypeSafe System One) backed auto mode for the Pi coding agent: semantically auto-approves bash, write, and edit tool calls and fails closed when a decision cannot be made. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 7 | lhemerly/mcts-agent — Discriminative Monte Carlo Tree Search using TypeSafe Jev System One Primitives and Gemini · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 2 | morcoan/JevSeek — A local coding workspace pairing Jev action routing with DeepSeek argument generation. Native tools, persistent sessions, React desktop, and documented research. · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 5 | savka777/jev-search — Fast deep research for the pi coding agent: reads up to 100 pages in full per round, Jev keeps only the passages that answer your questions. · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 1 | shubhangi013/prompt-oscilloscope — Analyze prompts with TypeSafe Jev before GitHub Copilot · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 1 | siddicky/omp-typesafe — TypeSafe AI (Jev) adversarial reviewer and typesafe\_ask tool for the omp coding agent · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 1 | stoopid-computers/jev-bot — Computer Use Agent developed with Jev · sdk · TypeScript | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 621 | thruwire/foreman — Software factory foreman based on TypeSafe's Jev model · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 2 | undeemed/Jcyber — Agent-driven bug bounty / pentest framework: one gated chain over five systems (Caido, HexStrike, Jev, Memgraph, TencentDB, Prometheus) · sdk · Python | Agents & orchestrationMore projects & source code | |
| sdkgithub.com | 107 | vinilana/jev-eval-agent — 🇺🇸 English · 🇧🇷 Leia em português · sdk · HTML | Agents & orchestrationMore projects & source code | |
| Sedumgithub.com | — | Browser end-to-end testing: in goal mode each turn asks one Jev Choice for the next operation (click, type, done or blocked) plus a speculative target among the page's offered elements, capped at 24 requests, 18 actions and 120 s, and the test passes only when an independent verify claim clears two Nouls (holds ≥ 0.75, contradicted flagged at ≥ 0.5), since the planner's done is never a verdict; authored-step tests reuse the same Choice to resolve each plain-English step, and with your own API key a 20-person team's PR suite costs $38–$91 a month vs $4,875 on a per-step AI platform. | Agents & orchestrationMore projects & source code | |
| Sentinelgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| Sentinel — READMEgithub.com | 0 | Sentinel is a local research prototype using typed Jev judgments for trajectory drift and related action-risk signals; its README says it is not a calibrated security product. | Agents & orchestrationMore projects & source code | |
| Sentio SMTPgithub.com | 246 | Multi-tenant Rust mail server that gives AI agents their own inboxes and can use Jev as a message classifier to label inbound mail without changing the spam score. | Agents & orchestrationMore projects & source code | |
| siegegithub.com | 1 | SIEGE: 200 people vs one agent. A typed action gate (TypeSafe System One) that learns from every breach, evaluated by W&B Weave, hardened by a defender loop. Built at CoreWeave Hacks: Agent Loops 2026. | Agents & orchestrationMore projects & source code | |
| siege — READMEgithub.com | 1 | SIEGE’s TypeSafe action gate is part of an agent-security loop, but the public website is an explicitly labeled simulation and the provider chain includes LLM/mock fallbacks. | Agents & orchestrationMore projects & source code | |
| sizeup — READMEgithub.com | 0 | Coding-agent router uses Jev confidence to delegate appropriately sized work to a smaller-model helper without switching the main session model; uncertainty, errors or timeout preserve the original flow. | Agents & orchestrationMore projects & source code | |
| skill-library-skillgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| skill-library-skill — READMEgithub.com | 0 | A skill-library workflow uses Jev to score the skill catalog, then loads only relevant skills. | Agents & orchestrationMore projects & source code | |
| skillboxgithub.com | 255 | Self-hosted, versioned skills library for AI agents served over MCP with scoped clients, where Jev can optionally recommend which of the skills a client may access fit the current task. | Agents & orchestrationMore projects & source code | |
| Skillfulgithub.com | 5 | Capability router for coding agents that matches each prompt against your installed skills, MCP servers, subagents and slash commands, injects at most one suggestion, and measures whether it helped. | Agents & orchestrationMore projects & source code | |
| skillpickgithub.com | 2 | Pick the right agent skill for every prompt with TypeSafe's Jev. Hooks for Claude Code, Codex, Gemini CLI, Droid, OpenCode, Amp. | Agents & orchestrationMore projects & source code | |
| skillpick — READMEgithub.com | 2 | skillpick asks Jev to choose which installed coding-agent skill fits each prompt, then injects a short hint into the agent context. | Agents & orchestrationMore projects & source code | |
| skillrankergithub.com | 1 | Rust CLI powered by Jev from TypeSafe.ai that ranks agent skills for the next step using live session context. Includes Claude Code hooks, structured JSON, abstention, and local feedback. Requires a TypeSafe API key. | Agents & orchestrationMore projects & source code | |
| skillranker — READMEgithub.com | 125 | SkillRanker integrates Jev into agent-skill selection by comparing available skills against live context; the README says ranking requires a user-provided TypeSafe key. | Agents & orchestrationMore projects & source code | |
| skillranker — READMEgithub.com | 1 | SkillRanker uses Jev to evaluate agent-context skills after local retrieval narrows the candidate list; fresh rankings require a TypeSafe API key. | Agents & orchestrationMore projects & source code | |
| smart-home-assistant — READMEgithub.com | 0 | Simulated smart-home app uses Jev to turn natural-language requests into typed room and device decisions and displays their probabilities. | Agents & orchestrationMore projects & source code | |
| SpendBrakegithub.com | 0 | Budget brake for agent runs that reads the budget, spend so far and remaining plan, then decides continue, downgrade the model or stop, with a value-for-cost score and over-budget checks. | Agents & orchestrationMore projects & source code | |
| spendbrake — READMEgithub.com | 0 | SpendBrake is an agent-budget dashboard that asks TypeSafe Jev whether to continue, downgrade the model, or stop. | Agents & orchestrationMore projects & source code | |
| sportsclaw Jev evidence verifiergithub.com | — | Opt-in evidence verifier in sportsclaw, a CLI and bot scaffold connecting LLMs to live sports data, that uses a typed JevDecisionClient for the final evidence check so high-confidence support skips the generative verification pass. | Agents & orchestrationMore projects & source code | |
| sportsclaw Jev evidence verifier — repogithub.com | — | Opt-in evidence verifier in sportsclaw, a CLI and bot scaffold connecting LLMs to live sports data, that uses a typed JevDecisionClient for the final evidence check so high-confidence support skips the generative verification pass. | Agents & orchestrationMore projects & source code | |
| stanley-code — READMEgithub.com | 118 | Stanley uses Jev to route requests to workflows, then asks bounded fixed-choice questions after gathering evidence; its README says code, not a model, makes the decision. | Agents & orchestrationMore projects & source code | |
| stepwarden — READMEgithub.com | 1 | Stepwarden is a proof-of-concept Claude Code plugin that asks Jev parallel questions about a pending tool call before it runs. | Agents & orchestrationMore projects & source code | |
| subagent-dispatchergithub.com | 0 | experimental subagent dispatcher | Agents & orchestrationMore projects & source code | |
| subagent-dispatcher — READMEgithub.com | 0 | The plugin explores deterministic subagent routing with Jev limited to semantic predicates; quality and cost effects are posed as research questions. | Agents & orchestrationMore projects & source code | |
| sunvaigithub.com | 0 | AI triage layer for MP CM Helpline 181: Jev typed decisions + Claude, built at Claude Code Build Day Bhopal | Agents & orchestrationMore projects & source code | |
| sunvai — READMEgithub.com | 0 | Sunvai documents a helpline-triage prototype: Jev supplies typed decisions, application code routes by confidence, and Claude drafts a brief. | Agents & orchestrationMore projects & source code | |
| super-jev — READMEgithub.com | 13 | Super Jev places Jev between an agent and its files for search, claim checks, and action permissions; its README requires a TypeSafe key for connect, ask, and check. | Agents & orchestrationMore projects & source code | |
| swarmrouter — jev.tsgithub.com | 0 | SwarmRouter asks Jev to choose a specialized agent or skill and score routing confidence; the missing-key path uses local heuristics instead of Jev. | Agents & orchestrationMore projects & source code | |
| swarmrouter — product.tsgithub.com | 0 | SwarmRouter asks Jev to choose a specialized agent or skill and score routing confidence; the missing-key path uses local heuristics instead of Jev. | Agents & orchestrationMore projects & source code | |
| switchboard — READMEgithub.com | 16 | A model-and-effort router for Claude Code and Codex uses Jev’s task assessment plus a user-defined policy, keeping the selected pair fixed through a conversation. | Agents & orchestrationMore projects & source code | |
| switchboard-ai — READMEgithub.com | 2 | Switchboard is an orchestration runtime with an optional Jev adapter for typed routing and verification; its README also documents a no-key rules fallback. | Agents & orchestrationMore projects & source code | |
| synthetic_colonygithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| synthetic_colony — READMEgithub.com | 0 | Experimental cognitive-graph platform studies compositional abstractions and documents confidence-gated escalation to Jev typed primitives; it offers offline/mocked operation separately from live Jev reasoning. | Agents & orchestrationMore projects & source code | |
| system-one-gemma — READMEgithub.com | 6 | A separate Gemma-based System One scorer presents itself as inspired by TypeSafe Jev, not as the hosted Jev service. | Agents & orchestrationMore projects & source code | |
| System-One-plus-Two-Ops-Agentgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| System-One-plus-Two-Ops-Agent — READMEgithub.com | 0 | Support-operations prototype uses Jev for triage, model routing and tool/reply checks alongside generative drafting; tools are in-memory mocks and replies are not actually sent. | Agents & orchestrationMore projects & source code | |
| system1-agents — READMEgithub.com | 136 | system1-agents offers configurable prebuilt or custom agents that can run with Jev, Laya, or Cua-S1; its README includes a blackjack run selecting Jev. | Agents & orchestrationMore projects & source code | |
| t-shiratori-ai-model-evaluation-routergithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| t-shiratori-ai-model-evaluation-router — READMEgithub.com | 0 | CLI supplies task descriptions and registered model-evaluation snapshots to Jev for model recommendations, without executing or automatically switching the recommended model. | Agents & orchestrationMore projects & source code | |
| tagtrimgithub.com | 0 | Tag every line of a command's output, then trim by tag, so your coding agent reads the signal, not the noise. Design stage. | Agents & orchestrationMore projects & source code | |
| tano_hackathongithub.com | 0 | tano\_hackathon - tano\_hackathon · Python | Agents & orchestrationMore projects & source code | |
| tano_hackathon — READMEgithub.com | 0 | A hackathon assistant uses Jev for typed judgments to prepare creator-message and product drafts; the README says it never sends them. | Agents & orchestrationMore projects & source code | |
| the-llm-dispatcher — READMEgithub.com | 0 | An LLM dispatcher uses Jev to judge each request and route it to the cheapest suitable model. | Agents & orchestrationMore projects & source code | |
| ThiagaoBR/typesafe_agent_gatesgithub.com | 1 | LangChain / Deep Agents middleware that uses TypeSafe's System One model (Jev) for typed judgments in unattended coding agents: a shell-command gate (database, production, destructive, secrets), issue triage and routing by severity and urgency, merge-request detection, and review of weakened tests. Measured with live probes. | Agents & orchestrationMore projects & source code | |
| tiershift — READMEgithub.com | 3 | A model-tier router uses TypeSafe Jev to select a lower-cost tier and escalate when its configured signals warrant it. | Agents & orchestrationMore projects & source code | |
| tinyhivemind-typesafegithub.com | — | Rust crate that lets the TinyHiveMind multi-agent coordination library route handoffs with Jev, picking the best-placed teammate and also routing to any other option above 20%. | Agents & orchestrationMore projects & source code | |
| tinyhivemind-typesafe — repogithub.com | — | Rust crate that lets the TinyHiveMind multi-agent coordination library route handoffs with Jev, picking the best-placed teammate and also routing to any other option above 20%. | Agents & orchestrationMore projects & source code | |
| todo-jev — READMEgithub.com | 3 | todo-jev is a TypeSafe Jev task-routing experiment that recommends a local rule, structured decision, skill, or foundation model; it documents an offline fallback. | Agents & orchestrationMore projects & source code | |
| Tonone Jev decision layergithub.com | — | Shared decision layer for a 426-skill AI company toolkit that answers yes/no, one-of-N and rubric questions with Jev, or a local TF-IDF scorer when no key is set, and gates skill manifests. | Agents & orchestrationMore projects & source code | |
| toolboxgithub.com | 0 | One MCP endpoint in front of all your MCP servers: Jev routes each prompt to the right server and tool, and a risk gate decides run, ask or block. | Agents & orchestrationMore projects & source code | |
| toolbox — READMEgithub.com | 0 | MCP gateway uses Jev to select a server and tool per request, then applies a separate risk gate before execution. | Agents & orchestrationMore projects & source code | |
| toolgate — READMEgithub.com | 2 | ToolGate is an agent/MCP call gate that uses TypeSafe Jev risk scores to return allow, ask-human, or deny decisions. | Agents & orchestrationMore projects & source code | |
| trajectoire-ai/hermes-structured-aux-modelsgithub.com | 1 | A Hermes Agent model-provider plugin that routes selected auxiliary tasks through bounded Jev decision calls on OpenRouter instead of free-form chat prompts. | Agents & orchestrationMore projects & source code | |
| TrashCompact — READMEgithub.com | 0 | TrashCompact combines deterministic context cleanup with Jev scoring of assistant text and Jev-selected excerpts as separate stages. | Agents & orchestrationMore projects & source code | |
| triage-guardgithub.com | 0 | Worked example with three pipelines sharing one architecture: support-ticket triage behind a safety guardrail, observability alert triage with dedupe and auto-page decisions, and a deploy-risk gate. | Agents & orchestrationMore projects & source code | |
| triagedygithub.com | 1 | Rust UNIX filter for security-alert triage: JSONL alerts in, typed decisions out, with Jev or a local model answering five questions per alert and routing policy kept in code. | Agents & orchestrationMore projects & source code | |
| triagedy — READMEgithub.com | 1 | Unix-style security-alert filter uses Jev by default to turn JSONL alerts into typed triage decisions, with other model backends optional. | Agents & orchestrationMore projects & source code | |
| Turing Jailgithub.com | — | Three-level interrogation game where you write statements to an AI warden and must pass plea, logic, and paradox verdicts to earn release, with shareable sealed runs. | Agents & orchestrationMore projects & source code | |
| TypeSafe AI for Agent Zerogithub.com | 5 | Agent Zero plugin whose typesafe_query tool combines Choice, Noul and Score questions over the same evidence and renders probability cards in chat; bundles the official TypeSafe agent skill. | Agents & orchestrationMore projects & source code | |
| TypeSafe Code Guardgithub.com | 1 | MCP server for Antigravity, Cursor and Claude Code that replaces grep sweeps with semantic code discovery over 15-line snippets and adds a pre-commit diff sanity check. | Agents & orchestrationMore projects & source code | |
| Typesafe Migration Guardgithub.com | 2 | CI/CD gate that evaluates database migration DDL before it reaches a Supabase PostgreSQL database and blocks destructive operations such as DROP TABLE or DROP COLUMN with an HTTP 403. | Agents & orchestrationMore projects & source code | |
| typesafe-agent-frameworkgithub.com | 0 | TypeSafe Jev ecosystem repository. | Agents & orchestrationMore projects & source code | |
| typesafe-agent-framework — READMEgithub.com | 0 | A C# integration connects Microsoft Agent Framework patterns to TypeSafe AI, including a TypeSafe System One agent, confidence-gated routing, guardrails, and dynamic tool selection. | Agents & orchestrationMore projects & source code | |
| TypeSafe-as-a-Judgegithub.com | 2 | Unofficial Codex and Claude Code MCP plugin that gives coding agents bounded Jev tools to route work, rank candidates, extract, verify evidence and decide when to escalate for review, while the agent keeps authority. | Agents & orchestrationMore projects & source code | |
| typesafe-as-a-judge — READMEgithub.com | 2 | An unofficial Codex/Claude plugin provides bounded Jev route, rank, extract, verify, and judge tools while leaving policy and side effects to the agent. | Agents & orchestrationMore projects & source code | |
| typesafe-assist — READMEgithub.com | 3 | TypeSafe Conversation is a Home Assistant Assist integration that uses Jev to interpret commands and map answers to built-in intents. | Agents & orchestrationMore projects & source code | |
| typesafe-cli (geilt)github.com | 3 | Standard-library Python CLI and agent skill for Claude, Codex, and Grok that evaluates a state against typed Choice, Score, and Noul questions and prints structured answers. | Agents & orchestrationMore projects & source code | |
| typesafe-computer-use-windowsgithub.com | 1 | Windows port of typesafe-computer-use with Jev decisions, local OCR/UIA, and MCP host integration | Agents & orchestrationMore projects & source code | |
| typesafe-computer-use-windows — READMEgithub.com | 1 | An independent Windows port has Jev choose desktop operations and targets from window text and controls, with local OCR, UI Automation, and MCP support. | Agents & orchestrationMore projects & source code | |
| typesafe-jev-incident-router — READMEgithub.com | 8 | This demo pairs a TypeSafe Jev use-case guide with a Python example that routes synthetic data incidents. | Agents & orchestrationMore projects & source code | |
| typesafe-jev-toolsgithub.com | 1 | A Claude Code hook that asks whether the decision you are writing needs a model at all. Includes a measured 149-row comparison of TypeSafe Jev against Claude Haiku 4.5. | Agents & orchestrationMore projects & source code | |
| typesafe-mario — READMEgithub.com | 423 | TypeSafe Mario maps emulator telemetry into structured state and lets Jev choose legal controller actions; the inclusion basis is the documented controller concept. | Agents & orchestrationMore projects & source code | |
| typesafe-mcpgithub.com | 1 | MCP server exposing TypeSafe (Jev/System One) to the fleet: judge, rerank, systemone. | Agents & orchestrationMore projects & source code | |
| typesafe-mcpgithub.com | 339 | Go MCP server that exposes one evaluate tool so Claude Code, Claude Desktop, Codex and pi can send state plus Choice, Score or Noul questions to Jev and branch on the probabilities. | Agents & orchestrationMore projects & source code | |
| typesafe-mcp — READMEgithub.com | 0 | A multi-target agent plugin uses TypeSafe System One (Jev) typed judgments to gate multi-agent task dispatch; reported latency and quality claims were not verified. | Agents & orchestrationMore projects & source code | |
| typesafe-mcp-servergithub.com | 0 | MCP server built on @typesafe-ai/sdk that exposes Choice, Score and Noul judgments and model listing as agent-callable tools, bundled with the official TypeSafe agent skill. | Agents & orchestrationMore projects & source code | |
| typesafe-minesweeper — READMEgithub.com | 0 | The README describes an autonomous Minesweeper experiment in which TypeSafe System One selects the next operation. | Agents & orchestrationMore projects & source code | |
| typesafe-modgithub.com | 3 | Claude Code function hook that ranks installed skills for each prompt and shows Jev's probability for each option when the agent asks a this-or-that question, optionally answering it automatically. | Agents & orchestrationMore projects & source code | |
| typesafe-mod — READMEgithub.com | 3 | typesafe-mod is a Claude Code function hook that sends AskUserQuestion choices to Jev; automatic answers require TYPESAFE_AUTO_ANSWER=1. | Agents & orchestrationMore projects & source code | |
| typesafe-on-neon — READMEgithub.com | 5 | A Neon Function proxy lets Jev inspect a request body and optionally forward the unchanged bytes to a caller-selected HTTPS target. | Agents & orchestrationMore projects & source code | |
| typesafe-on-neon — typesafe-on-neongithub.com | 5 | Neon Function proxy for the Neon AI Gateway with TypeSafe Jev routing. | Agents & orchestrationMore projects & source code | |
| typesafe-pigithub.com | 0 | TypeSafe System One for Pi: a consent-gated judgment tool and workflow design skill | Agents & orchestrationMore projects & source code | |
| typesafe-pi — READMEgithub.com | 0 | A Pi extension packages TypeSafe System One as a structured-judgment tool and a skill for TypeSafe-powered development; compatibility was not tested here. | Agents & orchestrationMore projects & source code | |
| typesafe-skill-router — READMEgithub.com | 14 | This Hermes plugin sends a request to TypeSafe Jev (jev-latest) to select a fitting skill, then appends a one-line hint to the user message. | Agents & orchestrationMore projects & source code | |
| typesafe-triage-guard — READMEgithub.com | 0 | This README calls the project three composable Jev judgment pipelines and frames it as a test lab for TypeSafe primitives, not a finished product. | Agents & orchestrationMore projects & source code | |
| umby-jev-stackgithub.com | 1 | Portable agent skill: TypeSafe Jev as a cheap code-review classifier (HTTP + optional jev-review MCP) | Agents & orchestrationMore projects & source code | |
| UX3 Product Design Harness Jev gategithub.com | — | Decision gate in an agent-operated product design harness that routes tasks and evaluates product decisions with Jev, returning continue, verify or stop-and-reframe verdicts plus task boundaries. | Agents & orchestrationMore projects & source code | |
| UX3 Product Design Harness Jev gate — repogithub.com | — | Decision gate in an agent-operated product design harness that routes tasks and evaluates product decisions with Jev, returning continue, verify or stop-and-reframe verdicts plus task boundaries. | Agents & orchestrationMore projects & source code | |
| VCPChat global Jev servicegithub.com | — | VCPChat, the desktop terminal for the VCP AI-native engine, adds a global Jev service so its modules can make typed decisions through TypeSafe or OpenRouter. | Agents & orchestrationMore projects & source code | |
| VCPToolBox Jev modulesgithub.com | — | The VCP agent infrastructure adds Jev modules for context pruning, fold filtering, reranking and a semantic tool-call experiment, with multi-key rotation over TypeSafe or OpenRouter. | Agents & orchestrationMore projects & source code | |
| wakegategithub.com | 3 | Fail-open gate for long-running agents on Workers, Durable Objects and Node that asks Jev, before a sleeping agent's LLM is resumed by a timer or event, whether the wakeup matters given the agent's own sleep note. | Agents & orchestrationMore projects & source code | |
| wakegate — READMEgithub.com | 3 | A fail-open wake gate asks Jev whether an event justifies resuming a sleeping agent for a full LLM turn; ordinary code controls sleep timing. | Agents & orchestrationMore projects & source code | |
| wechat-jev-assistantgithub.com | — | Decision Tools: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details. | Agents & orchestrationMore projects & source code | |
| wellposedgithub.com | 2 | Lints Jev requests before they are sent: offline checks for missing none-of-the-above options and broken state paths, then Jev itself for what structure cannot decide. | Agents & orchestrationMore projects & source code | |
| winnow — READMEgithub.com | 100 | Winnow is a Claude Code context sieve whose default judge is Jev; the README states it does nothing until a judge key is configured, so use is conditional. | Agents & orchestrationMore projects & source code | |
| XieChengYuan/jev-observatorygithub.com | 1 | Jev 活动看板：自动接入 MCP，实时查看真实输入、判断结果与用量。Local-first, passive MCP observability. · model_id · JavaScript | Agents & orchestrationMore projects & source code | |
| xjevboostgithub.com | — | Decision Tools: Add as much tabular data as you want to Jev models using adaptive ensembles that learn to query only the rows and columns needed. | Agents & orchestrationMore projects & source code | |
| y9Finsi/jev-mcpgithub.com | 2 | Universal Model Context Protocol (MCP) Server for TypeSafe Jev (System One) semantic code search and validation. | Agents & orchestrationMore projects & source code | |
| zcode-jevgithub.com | 0 | Typed judgment layer for coding agents with gates from PRD intake through coverage, CI triage and release, using any JSON-forced LLM today and a Jev backend written to the documented contract but not yet live-tested. | Agents & orchestrationMore projects & source code | |
| zcode-jev — READMEgithub.com | 0 | zcode-jev defines typed gates for coding-agent workflows, but the README says its working backend is JSON-forced LLM and the Jev backend is only a contract stub. | Agents & orchestrationMore projects & source code | |
| zocomputer-jevgithub.com | 0 | A Zo skill for situational script writing and execution using Vercel AI Gateway and TypeSafe AI Jev. | Agents & orchestrationMore projects & source code | |
| zocomputer-jev — READMEgithub.com | 0 | This versioned Zo Rule describes Jev as an internal decision delegate: Zo treats typed answers as evidence while retaining implementation, authorization, and side-effect control locally. | Agents & orchestrationMore projects & source code | |
| zurfyx/jev-browser-skill-demogithub.com | 1 | Let Jev, TypeSafe's ~100ms decision model, drive your browser. A plug-and-play skill for Claude Code and Codex. | Agents & orchestrationMore projects & source code |
More posts & discussions24 of 24 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| agent-handoff-gatex.com | — | Experimental protocol for evidence-aware agent handoffs, with Jev-assisted review before results reach the lead agent. | Agents & orchestrationMore posts & discussions | |
| AgentRunx.com | — | Harness from Grep.ai for repetitive knowledge work that learns the job as it runs, moving steps from LLM calls to code; 100,000 compliance alerts cost under $26K versus over $290K on Opus 5. | Agents & orchestrationMore posts & discussions | |
| AgentRun — articlex.com | — | Harness from Grep.ai for repetitive knowledge work that learns the job as it runs, moving steps from LLM calls to code; 100,000 compliance alerts cost under $26K versus over $290K on Opus 5. | Agents & orchestrationMore posts & discussions | |
| ask-jev gate for Grok botsx.com | — | Personal desk of Grok bots and assistants that now asks Jev first, via OpenRouter's Decisions API, for small yes/no, bucket and 1-5 score calls instead of waking Grok. | Agents & orchestrationMore posts & discussions | |
| ask-jev-skillx.com | — | Skill for Hermes, and other agents, to ask typesafe's jev | Agents & orchestrationMore posts & discussions | |
| Chat bot with no LLMx.com | — | Chat assistant built without any LLM: Jev picks the tool and its arguments across web search, Wikipedia, weather, Todoist and Home Assistant, so cited answers arrive instantly. | Agents & orchestrationMore posts & discussions | |
| foremanx.com | — | Software Factory Foreman: an agent supervisor that uses Jev decisions to keep coding agents on task. | Agents & orchestrationMore posts & discussions | |
| Group chat agent pickerx.com | — | Compares three ways to decide which agent in a group chat should answer: @mentions, an LLM orchestrator at 4-7s and $0.00046 per message, and Jev at 0.29s median and $0.00002. | Agents & orchestrationMore posts & discussions | |
| Hermes Jev skills — demox.com | — | Routing, memory, skill-selection, and triage skills for Hermes, Claude Code, and Codex that start in shadow mode and publish their failures. | Agents & orchestrationMore posts & discussions | |
| Hermes Jev skills — postx.com | — | Routing, memory, skill-selection, and triage skills for Hermes, Claude Code, and Codex that start in shadow mode and publish their failures. | Agents & orchestrationMore posts & discussions | |
| Jev as MCP workflow orchestratorx.com | — | Two patterns for combining Jev, LLMs and MCP: Jev as an outer loop that chooses valid actions to build a workflow AST from MCP output schemas, with a single LLM call writing the final answer. | Agents & orchestrationMore posts & discussions | |
| Jev in the elvex harness — discussionnews.ycombinator.com | — | Agent platform write-up on wiring Jev into an LLM harness as a callable tool for search, approvals and context, including 2,000 expense reports categorized in 20 seconds for 5 cents. | Agents & orchestrationMore posts & discussions | |
| jev-system-architectx.com | — | System-architecture skill for TypeSafe AI Jev/System One — find fuzzy semantic judgment and turn it into small Choice/Score/Noul primitives. | Agents & orchestrationMore posts & discussions | |
| jevc — demonews.ycombinator.com | — | Compiler that turns natural-language agent rules and JSON Schemas into Jev programs: a few narrow typed evidence questions plus a code reducer that computes the verdict, so a rule like never commit unless asked becomes testable. | Agents & orchestrationMore posts & discussions | |
| Jevmind — demox.com | — | Zero-dependency Python toolkit that pulls an agent's decisions out of prose into typed, confidence-scored answers behind a code gate, logged in a graded ledger, with a local rules brain or Jev behind one flag. | Agents & orchestrationMore posts & discussions | |
| JevScopex.com | — | Watches an AI agent work and plots per-step judgments of task alignment, progress, repetition and being stuck, shown catching a coding agent looping on a race condition. | Agents & orchestrationMore posts & discussions | |
| Just-in-time model selection in Goosex.com | — | Prototype for the Goose agent harness that has Jev read each prompt and pick which model should run the turn just before it starts. | Agents & orchestrationMore posts & discussions | |
| Monid tools with Jevx.com | — | Monid integration that lets an agent use Jev through OpenRouter across 2,000 tools, for jobs like scoring 2,000 leads, scanning TikTok hooks, or sorting Reddit threads, with a claimed 30x speed-up. | Agents & orchestrationMore posts & discussions | |
| OpenMausBot chief of staffx.com | — | Multi-agent setup on OpenMausBot where Jev reads each task, wakes the right agents and assigns each one a model from the user's existing subscriptions. | Agents & orchestrationMore posts & discussions | |
| Prompt difficulty classifierx.com | — | Chat UI check that classifies the difficulty of the prompt each time the user stops typing and, for simple prompts, offers a fast mode before they hit send. | Agents & orchestrationMore posts & discussions | |
| Slack agent pre-routingx.com | — | Slack agent made 2x faster by first asking Jev to pick the best skill, tool and parameters from the prompt before the agent runs. | Agents & orchestrationMore posts & discussions | |
| typesafe-jev-mcp — demowww.reddit.com | — | Rust MCP server with one evaluate tool that lets Claude Code, Codex, OpenCode, Antigravity or Cursor send a state and typed questions to Jev and get noul, choice or score answers with probabilities. | Agents & orchestrationMore posts & discussions | |
| typesafe-mcp — demox.com | — | Go MCP server that exposes one evaluate tool so Claude Code, Claude Desktop, Codex and pi can send state plus Choice, Score or Noul questions to Jev and branch on the probabilities. | Agents & orchestrationMore posts & discussions | |
| typesafe-mcp — discussionwww.reddit.com | — | Go MCP server that exposes one evaluate tool so Claude Code, Claude Desktop, Codex and pi can send state plus Choice, Score or Noul questions to Jev and branch on the probabilities. | Agents & orchestrationMore posts & discussions |
More videos & channels1 of 1 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| Treg + Jev automationwww.youtube.com | — | Demo of pairing Jev with Treg, an OpenRouter-style registry of agent tools, to build automation workflows where Jev makes the quick decisions. | Agents & orchestrationMore videos & channels |
More packages & releases5 of 5 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| jev-harness-router — npmwww.npmjs.com | — | Per-turn router for agent harnesses where one Jev call of about 350 ms sets the model tier, effort, tools and skill behind a hard deadline; it picks the right skill 90.0% of the time versus 59.4% for keywords. | Agents & orchestrationMore packages & releases | |
| jev-layer — npmwww.npmjs.com | — | Portable decision layer for agent harnesses such as Hermes, OMP, Codex and generic MCP that routes bounded choices through Jev and records receipts for replay, while the host keeps execution, permissions and approvals. | Agents & orchestrationMore packages & releases | |
| jevc — npmwww.npmjs.com | — | Compiler that turns natural-language agent rules and JSON Schemas into Jev programs: a few narrow typed evidence questions plus a code reducer that computes the verdict, so a rule like never commit unless asked becomes testable. | Agents & orchestrationMore packages & releases | |
| Sniff Test — npmwww.npmjs.com | — | Prose linter for Markdown and plain text that runs countable regex rules locally and sends paragraphs to Jev for judgment rules such as restating closings, stacked hedges or clichés, one probability per rule. | Agents & orchestrationMore packages & releases | |
| typesafe-jev-mcp — cratescrates.io | — | Rust MCP server with one evaluate tool that lets Claude Code, Codex, OpenCode, Antigravity or Cursor send a state and typed questions to Jev and get noul, choice or score answers with probabilities. | Agents & orchestrationMore packages & releases |
More guides & websites44 of 44 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| abmind System One judgments — appaksika.github.io | — | Pluggable System One judgment provider in abmind, a persistent memory and RAG engine for AI agents, that sends recall and diagnostics questions to Jev or a locally served Laya model as advice to its combining code. | Agents & orchestrationMore guides & websites | |
| Agent action approvalopenrouter.ai | — | OpenRouter Labs recipe that runs four checks (off task, could destroy, untrusted input, ask first) on each of 24 coding or ops agent tool calls in one request, so risky steps stop for a person: 96 answers in 0.5 s. | Agents & orchestrationMore guides & websites | |
| Agent action approval at Weathernewszenn.dev | — | Weathernews replaced the LLM that approved operations in its internal AI agent with Jev, keeping classification accuracy while cutting latency and cost, and shares the design and comparison. | Agents & orchestrationMore guides & websites | |
| Agent! Jev advisor — appagentiloop.ai | — | Jev advisor in Agent!, a native macOS agent app, that gives a second opinion on shell commands already passed by pattern rules and refuses those whose data-destruction probability exceeds a user-set threshold. | Agents & orchestrationMore guides & websites | |
| AgenticX Jev group routing — appwww.agxbuilder.com | — | Multi-agent platform whose group chat routes un-mentioned messages with Jev, dispatching by label when confident and falling back to an LLM, and shows a decision card each turn. | Agents & orchestrationMore guides & websites | |
| AgentX decision backends — appagentx-docs.pages.dev | — | Self-hosted AI agent mesh for small businesses whose decision layer can run Jev-style typed decisions on a local LLM or a simple-jev server, recomputing confidence as normalized entropy. | Agents & orchestrationMore guides & websites | |
| AIMEAT decide — appaimeat.io | — | Federated, self-hosted AI operating system whose nodes carry Jev beside the text model, so apps and agents can ask closed triage, routing, scoring and pick-one questions via a JS library, MCP tools and a skill. | Agents & orchestrationMore guides & websites | |
| AIWG Jev decision adapter — appaiwg.io | — | Decision framework in an AI-augmented dev toolkit where versioned rulesets and bindings let workflows switch between Jev and an LLM subagent, with a hardened Jev transport. | Agents & orchestrationMore guides & websites | |
| Arbos Jev router — apparbos.life | — | Structured router in the Arbos file-system-native agent coordinator where Jev picks the next mechanical tool move from a situation card, while the LLM handles writing, planning and talking; act=llm hands the turn to the chat model. | Agents & orchestrationMore guides & websites | |
| AutoGPT TypeSafe blocks — appagpt.co | — | Seven no-code blocks, including a five-exit router, a yes, no, or unsure split, and a score filter. | Agents & orchestrationMore guides & websites | |
| Deco Studio Jev task dedupe — appdecocms.com | — | Agent control plane whose task board first asks Jev whether incoming tasks duplicate existing cards, falling back to a fast LLM when Jev is unavailable or unsure. | Agents & orchestrationMore guides & websites | |
| dsh-auto-modegit.allen-software.com | — | Auto-mode permission preset for DeepSeek Harness where Jev answers the open questions an agent leaves at each end prompt and hands the turn back to the human when unsure. | Agents & orchestrationMore guides & websites | |
| EchoFlow Jev router — appechoflow.adityavs.tech | — | Per-turn Jev router in the EchoFlow privacy-first Android AI chat app that decides whether to recall or skip memory, force web search, or save a memory before the answering model runs, for cloud chats only. | Agents & orchestrationMore guides & websites | |
| evoke — siteevoke.build | — | Rust CLI, git-based package manager and TypeScript SDK where a spoken or typed sentence becomes a call to an installed reflex program chosen by Jev with bounded arguments, gated to run, confirm, ask or abstain. | Agents & orchestrationMore guides & websites | |
| ExploitHunter Jev decision specialist — appjustsml.github.io | — | Offensive-security research harness that uses Jev as a bounded decision specialist to pick tools and render formats, with a defer option and a persisted receipt for every call. | Agents & orchestrationMore guides & websites | |
| fabricioctelles/skills websiteskilldev.pro | — | A collection of skills for AI agents (Kiro, Cursor, Windsurf, Claude Code, and others). Each skill is a reusable module that teaches the agent to perform complex tasks with context, structure, and best practices. | Agents & orchestrationMore guides & websites | |
| HarnessJudge — appharnessjudge.vercel.app | — | Paste an agent step trace and get an ok, retry, escalate or stop decision for agent builders. | Agents & orchestrationMore guides & websites | |
| Hersona decide — docsshiro-0x.github.io | — | Persona attribute library for AI agents whose hersona decide command uses TypeSafe Jev to recommend reply, ask, search, use a tool or hold, with confidence, persona alignment and risk. | Agents & orchestrationMore guides & websites | |
| iii-directory Jev search — appworkers.iii.dev | — | Worker for the iii engine that searches registered functions, installed skills and triggers with Jev relevance judgments, falling back to hybrid search. | Agents & orchestrationMore guides & websites | |
| Imajin TypeSafe connector — appwww.imajin.ai | — | Service connector in the reference implementation of the MJN trust protocol that lets identities call Jev with a sealed, kernel-held API key under a scoped typesafe:decide grant. | Agents & orchestrationMore guides & websites | |
| JarvisCore Jev decisions — websitejarviscore.developers.prescottdata.io | — | Native Jev support in the JarvisCore multi-agent runtime: agents ask typed Choice, Score and Noul questions through a decision client separate from the text model, and the kernel picks specialist subagents by Choice. | Agents & orchestrationMore guides & websites | |
| Jev Decisions — apphuggingface.co | — | Plugin for Hermes and other agents that adds tools to review risky plans, recommend human approval, check answers against sources, assess whether a task is really finished, and keep a local decision journal. | Agents & orchestrationMore guides & websites | |
| Jev in the elvex harnesswww.elvex.com | — | Agent platform write-up on wiring Jev into an LLM harness as a callable tool for search, approvals and context, including 2,000 expense reports categorized in 20 seconds for 5 cents. | Agents & orchestrationMore guides & websites | |
| Jev recipes on Tregtreg.to | — | Interactive guide to Jev with a live LLM-versus-Jev race and copyable GTM automation recipes built on Treg, such as signup fraud screening, buying-signal triage, and viral content monitoring. | Agents & orchestrationMore guides & websites | |
| jev-search — appjev.s1.dev | — | Natural-language web search that picks the time range and query with Choices and keeps each source only if its Noul passes (needs a Search1API key). | Agents & orchestrationMore guides & websites | |
| jevifythursdai.news | — | An agent skill to discover TypeSafe Jev opportunities, design typed questions, and learn from recent community experiments. | Agents & orchestrationMore guides & websites | |
| LiteLLM guardraildocs.litellm.ai | — | Asks one Noul per finished tool exchange and clears results that are no longer needed before each call. | Agents & orchestrationMore guides & websites | |
| Markus decide tool — appwww.markus.global | — | AI workforce platform that gives its agents a decision capability and a decide tool backed by Jev-style decision models, with a rule that all evidence must go into the state. | Agents & orchestrationMore guides & websites | |
| MindRoom participation judgments — docsdocs.mindroom.chat | — | Multi-agent runtime on Matrix where agents already in a thread decide whether to answer untagged messages, using an LLM or TypeSafe System One with a 0.8 threshold. | Agents & orchestrationMore guides & websites | |
| Monid tools with Jev — appmonid.ai | — | Monid integration that lets an agent use Jev through OpenRouter across 2,000 tools, for jobs like scoring 2,000 leads, scanning TikTok hooks, or sorting Reddit threads, with a claimed 30x speed-up. | Agents & orchestrationMore guides & websites | |
| Nika System One seat — appnika.sh | — | Workflow language for repeatable AI work whose compiler can hand bounded decisions to TypeSafe System One, one Choice with a "none" option per question. | Agents & orchestrationMore guides & websites | |
| oh-my-hermes Jev posture — docsrlaope.github.io | — | Module in the oh-my-hermes plugin that detects Jev-class Hermes plugins on a machine and reports what each declares about its tools, hooks and data flow, without claiming any ran. | Agents & orchestrationMore guides & websites | |
| Open-Science Jev classification — appaipoch.com | — | Classification service in the AIPOCH Open-Science research workbench that uses Jev to select which skills a request needs and whether a document should be read in full. | Agents & orchestrationMore guides & websites | |
| OpenAgents Jev — appopenagents.com | — | OpenAgents builds general agent infrastructure where every turn starts with a Jev typed judgment that returns a choice and probabilities, foldable back into the round. | Agents & orchestrationMore guides & websites | |
| opencompany approval review — appopencompany.cloud | — | Approval reviewer in the opencompany AI workspace where Jev auto-approves routine integration actions only when they match the user's request and look low-risk; destructive, metered or permission changes always ask. | Agents & orchestrationMore guides & websites | |
| OpenCompany — apptinyhumans.ai | — | Agent hive for one-person businesses where Jev decides who a message needs and who picks up a broadcast, falling back to the desk lead without a key. | Agents & orchestrationMore guides & websites | |
| Pi Agent Foreman — packagepi.dev | — | Pi extension that reviews the final exchange after every settled run and, when the agent admits it stopped before required work was done, sends one firm instruction to finish the job. | Agents & orchestrationMore guides & websites | |
| Pisper decision service — apppisper.cc | — | Multi-agent desktop and mobile app that adds a Jev decision service for auto-approving tool permissions above a 0.9 threshold and verifying computer-use actions. | Agents & orchestrationMore guides & websites | |
| Rakazo Jev auto-review — apprakazo.com | — | Auto-review adapter in the open-source Rakazo AI-teammate platform that asks Jev whether a bot tool call should auto-pass or be sent to the user for approval. | Agents & orchestrationMore guides & websites | |
| Sentio SMTP — apptruespar.com | — | Multi-tenant Rust mail server that gives AI agents their own inboxes and can use Jev as a message classifier to label inbound mail without changing the spam score. | Agents & orchestrationMore guides & websites | |
| SpendBrake — appspendbrake.vercel.app | — | Budget brake for agent runs that reads the budget, spend so far and remaining plan, then decides continue, downgrade the model or stop, with a value-for-cost score and over-budget checks. | Agents & orchestrationMore guides & websites | |
| Supercov — websitesupercov.com | — | CLI that tells a coding agent what to fix and what to test by scoring code-quality properties with Jev, about a cent per megabyte of source, and turning uncovered paths from your existing test command into targets. | Agents & orchestrationMore guides & websites | |
| Tonone Jev decision layer — appsecond.tonone.ai | — | Shared decision layer for a 426-skill AI company toolkit that answers yes/no, one-of-N and rubric questions with Jev, or a local TF-IDF scorer when no key is set, and gates skill manifests. | Agents & orchestrationMore guides & websites | |
| VCPToolBox Jev modules — appwww.vcptoolbox.com | — | The VCP agent infrastructure adds Jev modules for context pruning, fold filtering, reranking and a semantic tool-call experiment, with multi-key rotation over TypeSafe or OpenRouter. | Agents & orchestrationMore guides & websites |
No matches this time
Try a broader search or reset your filters to explore the full collection.
Showing 1,475 of 1,475 resources