Jev AtlasTypeSafe & System One
About the directory GitHub
THE COMMUNITY RESOURCE DIRECTORY

Small decisions.
Endless possibilities.

Find what people build with Jev. Explore tools, libraries, experiments and ideas from across the TypeSafe ecosystem.

Updated · 2026-10-01About the directory
11,066unique resources
29clear categories
220focused subcategories
Yourssave a personal reading list
YOUR NEXT BUILD STARTS HERE

Explore the collection

Download catalog

1,328 of 11,066 resources

Coding & developer tools1,328 of 1,328 matches
Code review & verification44 of 44 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
@mizchi/eslint-plugin-jevwww.npmjs.com—Experimental ESLint plugin whose verdicts come from Jev: write a rule as a node selector plus one sentence of prose, or get per-function scores against eight named defect classes.Coding & developer toolsCode review & verification
@mizchi/eslint-plugin-jev — demox.com—Experimental ESLint plugin whose verdicts come from Jev: write a rule as a node selector plus one sentence of prose, or get per-function scores against eight named defect classes.Coding & developer toolsCode review & verification
@mizchi/eslint-plugin-jev — repogithub.com19Experimental ESLint plugin whose verdicts come from Jev: write a rule as a node selector plus one sentence of prose, or get per-function scores against eight named defect classes.Coding & developer toolsCode review & verification
AIOSON jev:reviewgithub.com—Jev commands in the AIOSON AI operating framework for software projects: jev:review adds a typed semantic judgment to QA of UI deliveries and prototype critique, next to jev:judge and jev:agent-review.Coding & developer toolsCode review & verification
AIOSON jev:review — repogithub.com—Jev commands in the AIOSON AI operating framework for software projects: jev:review adds a typed semantic judgment to QA of UI deliveries and prototype critique, next to jev:judge and jev:agent-review.Coding & developer toolsCode review & verification
Celesto PR Review Labgithub.com—Example app on Celesto's microVM sandboxes where an agent prepares a public GitHub PR's test environment, runs checks, and compares an LLM reviewer with Jev on the candidate findings.Coding & developer toolsCode review & verification
Celesto PR Review Lab — discussionwww.reddit.com—Example app on Celesto's microVM sandboxes where an agent prepares a public GitHub PR's test environment, runs checks, and compares an LLM reviewer with Jev on the candidate findings.Coding & developer toolsCode review & verification
Celesto PR Review Lab — repogithub.com—Example app on Celesto's microVM sandboxes where an agent prepares a public GitHub PR's test environment, runs checks, and compares an LLM reviewer with Jev on the candidate findings.Coding & developer toolsCode review & verification
Clean Code Reviewgithub.com11Reviewer for public GitHub pull requests in which Jev answers a question set from Robert C. Martin's Clean Code for every changed file and Luna writes the review from those answers; built on eve and Next.js.Coding & developer toolsCode review & verification
diffguardgithub.com1Probabilistic pull-request risk triage using Jev and deterministic review policy.Coding & developer toolsCode review & verification
DiffJurygithub.com7Next.js app that fetches a public GitHub PR (title, body, diff, contributors) and returns a Jev merge-risk verdict with Score bars, Noul probabilities and a suggested review depth from one call.Coding & developer toolsCode review & verification
DiffJury — appdiffjury.up.railway.app—Next.js app that fetches a public GitHub PR (title, body, diff, contributors) and returns a Jev merge-risk verdict with Score bars, Noul probabilities and a suggested review depth from one call.Coding & developer toolsCode review & verification
five-linesgithub.com2Use Jev to review a PR diff against the ten rules of Five Lines of CodeCoding & developer toolsCode review & verification
hunk-triagegithub.com2Read the important changes first.Coding & developer toolsCode review & verification
if-aigithub.com1Plain-English pull request checks powered by Jev. One condition, a minimum confidence, one check.Coding & developer toolsCode review & verification
IFClite review lane Jev clientgithub.com—AI code-review lane of an open-source IFC toolkit that uses Jev to match review findings semantically against evals and merge duplicate findings, failing soft without a key.Coding & developer toolsCode review & verification
Jev PR reviewerx.com—PR reviewer that sends a diff to Jev in one call and gets 14 typed checks back as probabilities, mapped to block, security review, nits or merge, for $0.00007 per PR.Coding & developer toolsCode review & verification
Jev review MCPgithub.com230MCP server that rates an agent's diff on 19 quality dimensions, each with an applicability Noul, a Score, and a main-weakness Choice.Coding & developer toolsCode review & verification
Jev review MCP — demox.com—MCP server that rates an agent's diff on 19 quality dimensions, each with an applicability Noul, a Score, and a main-weakness Choice.Coding & developer toolsCode review & verification
Jev review MCP — videogithub.com—MCP server that rates an agent's diff on 19 quality dimensions, each with an applicability Noul, a Score, and a main-weakness Choice.Coding & developer toolsCode review & verification
jev-lens.nvimgithub.com1Neovim popup that shows the jev-lens verdict after Claude Code finishes editing: whether you need to look, which files and why, with keys to open the diff or strip flagged debris.Coding & developer toolsCode review & verification
jev-lens.nvim — plugingithub.com0Neovim popup that shows the jev-lens verdict after Claude Code finishes editing: whether you need to look, which files and why, with keys to open the diff or strip flagged debris.Coding & developer toolsCode review & verification
jev-needs-reviewgithub.com0Classifies PRs as "merge candidate" or "human review" required using Jev's typed decision making.Coding & developer toolsCode review & verification
jev-review — appblink.review—Staged code review that builds a Noul risk matrix, then profiles and rates each finding with Choices and Scores.Coding & developer toolsCode review & verification
jev-review — demox.com—Staged code review that builds a Noul risk matrix, then profiles and rates each finding with Choices and Scores.Coding & developer toolsCode review & verification
jev-review — discussionwww.reddit.com—Staged code review that builds a Noul risk matrix, then profiles and rates each finding with Choices and Scores.Coding & developer toolsCode review & verification
Meaning Diffx.com—Open-source patch reviewer that looks at what a diff means rather than its lines; on a broken checkout it flagged an agent hiding the payment failure, weakening an assertion and faking receipt delivery, in 343ms with Jev.Coding & developer toolsCode review & verification
meaning-aware-diffgithub.com0See the consequences hidden in a code diff. Jev-powered semantic reviews, a CLI, and an interactive demo.Coding & developer toolsCode review & verification
meko-jev-code-review-agentgithub.com0A code review check that learns from reviewer rulings without fine-tuning. Jev (TypeSafe AI) judges each diff hunk against your team's guidelines; a Meko datapack (powered by YugabyteDB) holds the guidelines, the rulings, and a trace of every decision. Companion repo for the blog post "Jev and Meko Make Code Review Feedback Stick".Coding & developer toolsCode review & verification
neuralintgithub.com1AI code review against your repository's rules, powered by TypeSafe AI's Jev.Coding & developer toolsCode review & verification
noulsgithub.com6Semantic linter and language server that splits files into functions with tree-sitter and asks Jev yes/no questions per function for issues like docstring drift, missing authorisation or non-idempotent retries.Coding & developer toolsCode review & verification
nouls — appbenomahony.github.io—Semantic linter and language server that splits files into functions with tree-sitter and asks Jev yes/no questions per function for issues like docstring drift, missing authorisation or non-idempotent retries.Coding & developer toolsCode review & verification
pi-reviewgithub.com15CLI, CI gate and agent skill that hands code and plan reviews to isolated read-only Pi reviewers; Jev powers a ~1 s screen against a defect catalog (1.2 s vs 30-50 s for a full round) and adjudicates panel consensus.Coding & developer toolsCode review & verification
pi-review — npmwww.npmjs.com—CLI, CI gate and agent skill that hands code and plan reviews to isolated read-only Pi reviewers; Jev powers a ~1 s screen against a defect catalog (1.2 s vs 30-50 s for a full round) and adjudicates panel consensus.Coding & developer toolsCode review & verification
pi-tool-supervisor TypeSafe backendgithub.com—Code-review backend for a Pi coding-agent supervisor that turns each numbered clause of a rule file into a Jev question about the diff; about US$0.00006 and 1-2 s per 3-rule review.Coding & developer toolsCode review & verification
pi-tool-supervisor TypeSafe backend — repogithub.com—Code-review backend for a Pi coding-agent supervisor that turns each numbered clause of a rule file into a Jev question about the diff; about US$0.00006 and 1-2 s per 3-rule review.Coding & developer toolsCode review & verification
pr-scoutgithub.com1Triage every open pull request of a GitHub repo in minutes: Jev typed questions, local Ollama descriptions, git test-merge, take / consider / skip boardCoding & developer toolsCode review & verification
roblox-jev-studiogithub.com0Review selected Roblox Studio Instances with typed Jev criteria and exact citations.Coding & developer toolsCode review & verification
slop-guardgithub.com0Guideline-driven code review bot for Ruby: code finds what to inspect, TypeSafe Jev judges it, explicit rules decide what to reportCoding & developer toolsCode review & verification
Sourdaw review-stances checkgithub.com—Review-process script in the Sourdaw open-source DAW that asks Jev, for each stance in a PR review bundle, whether its admission names a concrete failure mode rather than just the files touched, failing weak lines.Coding & developer toolsCode review & verification
Sourdaw review-stances check — repogithub.com—Review-process script in the Sourdaw open-source DAW that asks Jev, for each stance in a PR review bundle, whether its admission names a concrete failure mode rather than just the files touched, failing weak lines.Coding & developer toolsCode review & verification
tenetgithub.com5The review gate for code that agents write: rules in plain language, judged on every commit (TypeSafe Jev community).Coding & developer toolsCode review & verification
tracecheckgithub.com2Evidence-backed code review for AI coding agents, powered by Jev.Coding & developer toolsCode review & verification
typesafeai-reviewgithub.com0Code reviewer shipped as a dev dependency where code runs the checks and slices the diff, Jev answers narrow typed questions per hunk, and code composes the verdict, score and findings.Coding & developer toolsCode review & verification
Coding agents74 of 74 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
@elyracode/jev-toolswww.npmjs.com—Elyra coding-agent extension that adds a decide tool for Jev yes/no, choice and score questions, plus an opt-in bash gate that asks before running commands that look destructive.Coding & developer toolsCoding agents
Bambooedwww.youtube.com—Walkthrough of an architect-and-worker AI coding team in Bambooed, where an Astra architect plans, cheaper workers code, and optional Jev help picks workers, retrieves context, and reviews changes.Coding & developer toolsCoding agents
Bambooed — appbambooed.ai—Walkthrough of an architect-and-worker AI coding team in Bambooed, where an Astra architect plans, cheaper workers code, and optional Jev help picks workers, retrieves context, and reviews changes.Coding & developer toolsCoding agents
Claude Flow Novice Jev shadow pilotsgithub.com—Claude Code orchestration kit that runs Jev in shadow mode across its loops: pre-triaging review-vote suggestions, second-guessing destructive-command denies, and scoring night-mode report items.Coding & developer toolsCoding agents
Claude Flow Novice Jev shadow pilots — repogithub.com—Claude Code orchestration kit that runs Jev in shadow mode across its loops: pre-triaging review-vote suggestions, second-guessing destructive-command denies, and scoring night-mode report items.Coding & developer toolsCoding agents
Claude Lane Stack Jev skillsgithub.com—One-person multi-agent coding factory for Claude Code with Jev skills that review git diff hunks and run browser QA over a DOM table, via TypeSafe or OpenRouter.Coding & developer toolsCoding agents
Cloud Harness MCP skill suggestergithub.com—Remote coding harness exposed as an MCP server whose runner asks Jev, on prompt submit, whether a shortlisted skill fits the request before suggesting it.Coding & developer toolsCoding agents
Cloud Harness MCP skill suggester — appharness.agentkit.best—Remote coding harness exposed as an MCP server whose runner asks Jev, on prompt submit, whether a shortlisted skill fits the request before suggesting it.Coding & developer toolsCoding agents
Cloud Harness MCP skill suggester — repogithub.com—Remote coding harness exposed as an MCP server whose runner asks Jev, on prompt submit, whether a shortlisted skill fits the request before suggesting it.Coding & developer toolsCoding agents
Codex-Jevgithub.com1VS Code Codex Plugin for Jev-gated tool output integrationCoding & developer toolsCoding agents
Eva Jev featuresgithub.com—Cloud-sandbox coding agent orchestrator that uses Jev for task tags, findings triage, draft readiness, mention urgency and skill chips, with readers that fall back when an answer is missing.Coding & developer toolsCoding agents
Eva Jev features — appeva.vedantb.com—Cloud-sandbox coding agent orchestrator that uses Jev for task tags, findings triage, draft readiness, mention urgency and skill chips, with readers that fall back when an answer is missing.Coding & developer toolsCoding agents
foreman-jevgithub.com—Experimental supervisor where Jev oversees a Codex coding worker through Vercel AI Gateway while a programmer-chosen acceptance command decides pass or fail, plus a read-only Azure DevOps PR review flow.Coding & developer toolsCoding agents
GPT Researchergithub.com—Research agent that uses TypeSafe Jev to score scraped passages before context reaches its language model.Coding & developer toolsCoding agents
harlan-github-agent Jev classificationgithub.com—Claude Code agent kit with a service that works GitHub repos autonomously, using Jev classifications as a boundary where failures are returned as values so callers default to the safe path.Coding & developer toolsCoding agents
harlan-github-agent Jev classification — repogithub.com—Claude Code agent kit with a service that works GitHub repos autonomously, using Jev classifications as a boundary where failures are returned as values so callers default to the safe path.Coding & developer toolsCoding agents
Jackalope — projectgithub.com—Desktop workspace for running Codex, Claude Code, OpenCode, and other coding agents in parallel on Git worktrees, where Jev picks an agent per task, runs basic review checks, and supplies context.Coding & developer toolsCoding agents
Jauvexx.com—Two-way voice chat app for talking to and steering Claude and Codex coding agents while they work, with Jev classifying instructions and intents.Coding & developer toolsCoding agents
Jauvex — companyreindent.com—Two-way voice chat app for talking to and steering Claude and Codex coding agents while they work, with Jev classifying instructions and intents.Coding & developer toolsCoding agents
Jev + Claude Code = The Cheapest Agentic Coding Loop Yet — repogithub.com652Puts Jev inside a Claude Code agentic loop using the official skill, the skill-suggestion cookbook, and the jev-review workflow, and looks at what it costs to run.Coding & developer toolsCoding agents
Jev as debugging orchestratorwww.reddit.com—Debugging loop where Claude gathers evidence, Jev scores hypotheses and picks the next diagnostic, and Claude or Codex executes it, cutting repeated failed attempts.Coding & developer toolsCoding agents
Jev Guard (OpenHarness)github.com1,057Jev Guard agent in the OpenHarness store that watches your project folder, runs the tests on every edit, and has Jev score how close to the goal, how risky and how trusted each change is.Coding & developer toolsCoding agents
Jev Guard (OpenHarness) — appwww.autonomous.ai—Jev Guard agent in the OpenHarness store that watches your project folder, runs the tests on every edit, and has Jev score how close to the goal, how risky and how trusted each change is.Coding & developer toolsCoding agents
Jev × WebMCPgithub.com122Chrome side-panel extension that discovers the WebMCP tools a page exposes and uses Jev to pick and fill the right tool call as you type, showing confidence and latency with no site-specific setup.Coding & developer toolsCoding agents
jev-arenagithub.com1Write a fighter in plain English. Jev pilots it in real time. PR-driven ladder, swappable brains.Coding & developer toolsCoding agents
jev-autogithub.com—Letta Code mod that auto-approves low-impact tool calls and asks a person about risky or unclear ones, using Jev through OpenRouter; off by default and scoped to the current conversation.Coding & developer toolsCoding agents
jev-auto — repogithub.com—Letta Code mod that auto-approves low-impact tool calls and asks a person about risky or unclear ones, using Jev through OpenRouter; off by default and scoped to the current conversation.Coding & developer toolsCoding agents
jev-dev-harnessgithub.com0Open-source developer harness and runtime safety toolkit for AI coding agents powered by TypeSafe AI / JevCoding & developer toolsCoding agents
jev-harnessgithub.com4Minimal coding agent that filters every tool result through Jev before the main model sees it, with an A/B harness: 25/30 tasks passed with the filter versus 22/30 without.Coding & developer toolsCoding agents
jev-harness — pypipypi.org—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.Coding & developer toolsCoding agents
JevCodergithub.com—Pi coding agent extension where Jev selects each next action (a tool, code editing or finish) and the Pi model fills in arguments, with cache-stable tool schemas and SWE-bench Verified experiments.Coding & developer toolsCoding agents
Jivegithub.com19Terminal coding agent that replaces tool calls with graph calls: the LLM plans an executable DAG of tool calls and Jev calls, so fast intuitive decisions and bulk evaluations along the plan run without extra LLM calls.Coding & developer toolsCoding agents
Jive — demowww.reddit.com—Terminal coding agent that replaces tool calls with graph calls: the LLM plans an executable DAG of tool calls and Jev calls, so fast intuitive decisions and bulk evaluations along the plan run without extra LLM calls.Coding & developer toolsCoding agents
JMPgithub.com3Local desktop coding agent where Jev selects the next action from the request and real history, a chosen generator (DeepSeek, OpenAI or local Bonsai) fills in arguments, and OpenHands or MCP tools execute them.Coding & developer toolsCoding agents
kimaki opencode auto modegithub.com—The opencode-auto-mode plugin in kimaki, a Discord-based OpenCode orchestrator, runs a deny-all side-session classifier that uses Jev to decide each auto-mode turn.Coding & developer toolsCoding agents
kimaki opencode auto mode — appkimaki.dev—The opencode-auto-mode plugin in kimaki, a Discord-based OpenCode orchestrator, runs a deny-all side-session classifier that uses Jev to decide each auto-mode turn.Coding & developer toolsCoding agents
Kimi Code JevTriagegithub.com—JevTriage tool in an unofficial security-research variant of the Kimi Code CLI coding agent that asks Jev to triage a reported vulnerability finding from its type, endpoint, evidence and claimed impact.Coding & developer toolsCoding agents
Kimi Code JevTriage — repogithub.com—JevTriage tool in an unofficial security-research variant of the Kimi Code CLI coding agent that asks Jev to triage a reported vulnerability finding from its type, endpoint, evidence and claimed impact.Coding & developer toolsCoding agents
letcodegithub.com6Opencode-style terminal coding agent in Rust with a Ratatui TUI whose experimental auto permission mode can ask Jev instead of a chat reviewer whether to allow each tool call.Coding & developer toolsCoding agents
octomind evaluategithub.com—CLI-first AI coding agent runtime with an evaluate command that sends a JSON state and typed questions to an evaluation model such as Jev, including calibrated checks.Coding & developer toolsCoding agents
octomind evaluate — appoctomind.run—CLI-first AI coding agent runtime with an evaluate command that sends a JSON state and typed questions to an evaluation model such as Jev, including calibrated checks.Coding & developer toolsCoding agents
OpenHumangithub.com—Rust agent harness that uses Jev to rank tool-search candidates and select browser actions, with host-controlled approval gates for consequential actions.Coding & developer toolsCoding agents
OwnCodegithub.com6Experimental Go terminal coding agent with tool approvals, parallel Witch workers and pluggable context compaction, including an experimental Jev mode that scores and shortens eligible tool output via Compact Engine.Coding & developer toolsCoding agents
OwnCode — relatedgithub.com—Experimental Go terminal coding agent with tool approvals, parallel Witch workers and pluggable context compaction, including an experimental Jev mode that scores and shortens eligible tool output via Compact Engine.Coding & developer toolsCoding agents
phi Jev providergithub.com—A lean Go terminal coding agent that ports the synchronous System One surface of TypeSafe's SDK so it can ask Jev typed questions with the same retry policy as its other providers.Coding & developer toolsCoding agents
phi Jev provider — docspulseaiclub.github.io—A lean Go terminal coding agent that ports the synchronous System One surface of TypeSafe's SDK so it can ask Jev typed questions with the same retry policy as its other providers.Coding & developer toolsCoding agents
Pi Codemodegithub.com27Pi extension that replaces many small tool calls with one type-checked TypeScript program run in a sandbox; when a TypeSafe key is set, programs can call jev.ask for Noul, Choice and Score answers and compose thresholds in code.Coding & developer toolsCoding agents
pi-approve-for-megithub.com—Pi coding-agent extension that gates the bash tool: a fast TypeSafe risk score auto-runs low-risk commands, while elevated or changed ones go to a context-aware Jev reviewer and then to human confirmation.Coding & developer toolsCoding agents
pi-approve-for-me — repogithub.com—Pi coding-agent extension that gates the bash tool: a fast TypeSafe risk score auto-runs low-risk commands, while elevated or changed ones go to a context-aware Jev reviewer and then to human confirmation.Coding & developer toolsCoding agents
pi-fabric Jev programsgithub.com≈268Jev support in pi-fabric, a programmable tool runtime for the Pi coding agent, where the reasoning model writes a TypeScript program that observes, asks Jev, acts and loops in QuickJS without further LLM turns.Coding & developer toolsCoding agents
pi-fabric Jev programs — npmwww.npmjs.com—Jev support in pi-fabric, a programmable tool runtime for the Pi coding agent, where the reasoning model writes a TypeScript program that observes, asks Jev, acts and loops in QuickJS without further LLM turns.Coding & developer toolsCoding agents
pi-fabric Jev programs — repogithub.com≈268Jev support in pi-fabric, a programmable tool runtime for the Pi coding agent, where the reasoning model writes a TypeScript program that observes, asks Jev, acts and loops in QuickJS without further LLM turns.Coding & developer toolsCoding agents
pi-jev — demox.com—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.Coding & developer toolsCoding agents
pi-jev-auto-mode — discussionnews.ycombinator.com—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.Coding & developer toolsCoding agents
pi-jev-auto-mode — write-updev.to—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.Coding & developer toolsCoding agents
pi-warden — discussionwww.reddit.com—Guardrails for the Pi coding agent that enforce project rules on every write, hold only hard-to-undo actions (3 per 1,000 calls on real sessions), and catch unverified done claims and stuck loops.Coding & developer toolsCoding agents
pier Jev decision layergithub.com—Optional Jev decision layer in pier, a Pi coding-agent extension with a todo loop and interactive subagents in herdr terminal panes, used for a diagnostic gate, notice ranking and excerpt windows.Coding & developer toolsCoding agents
pier Jev decision layer — repogithub.com—Optional Jev decision layer in pier, a Pi coding-agent extension with a todo loop and interactive subagents in herdr terminal panes, used for a diagnostic gate, notice ranking and excerpt windows.Coding & developer toolsCoding agents
rift-typesafegithub.com—Jev client crate in Rift, a Rust terminal coding agent for local models, used for swarm decisions it would otherwise parse out of prose, kept separate from chat providers.Coding & developer toolsCoding agents
rift-typesafe — repogithub.com—Jev client crate in Rift, a Rust terminal coding agent for local models, used for swarm decisions it would otherwise parse out of prose, kept separate from chat providers.Coding & developer toolsCoding agents
S1Codegithub.com5Rust terminal coding agent where Claude or OpenAI plans and writes code while optional Jev integration selects among fully specified actions and helps decide which evidence stays in the active context.Coding & developer toolsCoding agents
S1Code — demonews.ycombinator.com—Rust terminal coding agent where Claude or OpenAI plans and writes code while optional Jev integration selects among fully specified actions and helps decide which evidence stays in the active context.Coding & developer toolsCoding agents
SAYACODEgithub.com6Local terminal coding agent built on LangChain and LangGraph whose jev trust level auto-approves low-risk tool calls, hands uncertain ones to the user and rejects clear overreach, using the official TypeSafe SDK.Coding & developer toolsCoding agents
Seatworksgithub.com12Paseo plugin that runs a supervisor, lead, peer, reviewer and watcher team of coding agents, with Jev via OpenRouter as an optional paid reader that flags incidents in how the work is being done.Coding & developer toolsCoding agents
Smithers Jev flowsgithub.com—Jev integration in the Smithers codebase-maintainer agent: a Vercel AI Gateway client plus flows that use it to score repository issues, spot duplicates and reproductions, check wiki citations and watch agent sessions.Coding & developer toolsCoding agents
Smithers Jev flows — websitesmithers.sh—Jev integration in the Smithers codebase-maintainer agent: a Vercel AI Gateway client plus flows that use it to score repository issues, spot duplicates and reproductions, check wiki citations and watch agent sessions.Coding & developer toolsCoding agents
software-factorygithub.com10Local software factory that moves requests through YAML pipelines of coding-agent, shell, and Jev steps, where a typesafe step judges the diff with typed questions and routes it to review or back to coding.Coding & developer toolsCoding agents
SuperQode SystemOne Tunex.com—SuperQode's SystemOne harness for Jev-like models adds GEPA-based tuning that improves Jev's typed tool-gating and ticket-routing answers from unlabeled history without silently rewriting the trusted decision pack.Coding & developer toolsCoding agents
SuperQode SystemOne Tune — docsdocs.superqode.dev—SuperQode's SystemOne harness for Jev-like models adds GEPA-based tuning that improves Jev's typed tool-gating and ticket-routing answers from unlabeled history without silently rewriting the trusted decision pack.Coding & developer toolsCoding agents
SuperQode SystemOne Tune — repogithub.com—SuperQode's SystemOne harness for Jev-like models adds GEPA-based tuning that improves Jev's typed tool-gating and ticket-routing answers from unlabeled history without silently rewriting the trusted decision pack.Coding & developer toolsCoding agents
The Jev-enatorgithub.com0Claude Code hooks that use Jev inside the agent loop: a PreToolUse danger gate that blocks destructive calls, a PostToolUse failure notice and an optional completion check, tested offline with cassettes.Coding & developer toolsCoding agents
Varro Jev auto-approvegithub.com—VS Code workbench for OpenCode that can auto-approve agent permission requests by asking Jev to allow, ask or reject, alongside destructive and manipulation checks.Coding & developer toolsCoding agents
Varro Jev auto-approve — appmarketplace.visualstudio.com—VS Code workbench for OpenCode that can auto-approve agent permission requests by asking Jev to allow, ask or reject, alongside destructive and manipulation checks.Coding & developer toolsCoding agents
Varro Jev auto-approve — repogithub.com—VS Code workbench for OpenCode that can auto-approve agent permission requests by asking Jev to allow, ask or reject, alongside destructive and manipulation checks.Coding & developer toolsCoding agents
Terminal & CLI tools23 of 23 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
DLQ Inspectorgithub.com5Local-first CLI for inspecting and safely recovering dead-letter queue messages from RabbitMQ and Redis Streams, with an opt-in Jev assist that resolves failures the rule-based classifier marks INVESTIGATE.Coding & developer toolsTerminal & CLI tools
GitHub Star Organizergithub.com1Python CLI that uses Jev to classify your GitHub stars into your existing GitHub Lists, writes a JSON classification report, and can apply the assignments.Coding & developer toolsTerminal & CLI tools
guessworkgithub.com2Fish-style inline zsh autosuggestions that ask Jev which recent history command you are retyping, so abbreviations and fuzzy matches work instead of prefix matching only.Coding & developer toolsTerminal & CLI tools
guesswork — appguesswork.findmalek.com—Fish-style inline zsh autosuggestions that ask Jev which recent history command you are retyping, so abbreviations and fuzzy matches work instead of prefix matching only.Coding & developer toolsTerminal & CLI tools
jeqgithub.com10Go CLI in the spirit of jq that lets scripts and agents pipe JSON and NDJSON through typed Jev questions, composing map, reduce, rank, and rate steps with an explicit offline policy gate.Coding & developer toolsTerminal & CLI tools
jev-cligithub.com1Ask TypeSafe AI typed questions about JSON state from the command lineCoding & developer toolsTerminal & CLI tools
jev-cligithub.com3Command-line tool that runs question packs such as log triage and security audit over JSON, NDJSON and JSONC artifacts, returning line-anchored typed answers you can gate on.Coding & developer toolsTerminal & CLI tools
jev-cligithub.com3Command-line client for Jev that takes a state and typed questions from files or stdin and prints structured JSON answers, calling TypeSafe directly or through Vercel AI Gateway.Coding & developer toolsTerminal & CLI tools
jev-cli (jevctl)github.com22TypeScript CLI installed as jevctl that turns Jev judgments into pipeable, exit-code-gated commands: verify claims, screen for prompt injection, classify, extract, match, route, rerank and compact transcripts.Coding & developer toolsTerminal & CLI tools
jev-logtriagegithub.com1CLI that batches collapsed Loki logs per source into one Jev call of Noul, Score and Choice questions, then maps answers in code to suppress, watch, review, notify or page, with low confidence going to review.Coding & developer toolsTerminal & CLI tools
jev-shell-historygithub.com117Zsh plugin that shows fish-style autosuggestions by asking Jev which of your last 100 distinct history entries you are most likely completing, displayed in grey with its score.Coding & developer toolsTerminal & CLI tools
jev-shell-history — demox.com—Zsh plugin that shows fish-style autosuggestions by asking Jev which of your last 100 distinct history entries you are most likely completing, displayed in grey with its score.Coding & developer toolsTerminal & CLI tools
jev-tablegithub.com0Local-first CLI that adds Jev-backed AI columns to CSV or JSONL files, labeling every row with typed answers and probabilities, with deduplication, a review queue, resume and a dry-run cost preview.Coding & developer toolsTerminal & CLI tools
jev-triagegithub.com—CLI & Pipelines: Message triage CLI in Go, built on the Jev decision model from TypeSafe AI. Categorizes messages, scores urgency, and flags low-confidence ones for human review.Coding & developer toolsTerminal & CLI tools
jevotrongithub.com6CLI-first field-level anomaly detection for structured files and text, powered by JevCoding & developer toolsTerminal & CLI tools
jevyoumeangithub.com—CLI & Pipelines: Unlike an edit-distance \Did you mean?\, \jym\ matches on \intent\: it hands Jev the candidate subcommand names plus their help descriptions. \remove → rm\, \list → ps\, \undo → restore\ are close in meaning but far in spelling — that is the gap this experiment targets.Coding & developer toolsTerminal & CLI tools
jsortgithub.com25Command-line sort by meaning that shows Jev two texts at a time, asks which ranks higher on a plain-English dimension, and fits a Bradley-Terry scale to print lines or documents with scores and standard errors.Coding & developer toolsTerminal & CLI tools
Port Cleanupgithub.com—Native macOS utility that lists every listening TCP port with its process and project context, and recommends which are safe to stop; you confirm every stop.Coding & developer toolsTerminal & CLI tools
shell.online Jev assessmentsgithub.com—Service that shares any terminal process as an encrypted browser link, adding opt-in Jev advisory assessments of excerpts the owner explicitly discloses from a session.Coding & developer toolsTerminal & CLI tools
shell.online Jev assessments — appshell.online—Service that shares any terminal process as an encrypted browser link, adding opt-in Jev advisory assessments of excerpts the owner explicitly discloses from a session.Coding & developer toolsTerminal & CLI tools
superterminalgithub.com100GPU-rendered native multiplexer terminal whose command palette uses Jev to understand plain-words queries like "kill this tab", ranking rows about a quarter second after typing settles.Coding & developer toolsTerminal & CLI tools
watfilegithub.com2CLI that sorts PDFs and text files into category folders with one Jev Choice per document, reporting per-category probabilities and confidence, with a pluggable backend for a local Laya model.Coding & developer toolsTerminal & CLI tools
watfile — pypipypi.org—CLI that sorts PDFs and text files into category folders with one Jev Choice per document, reporting per-category probabilities and confidence, with a pluggable backend for a local Laya model.Coding & developer toolsTerminal & CLI tools
Git, CI & testing33 of 33 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
Adversarial browser release testingx.com—Massively parallel browser-based adversarial test suite that tries to break each software release, costing pennies per run.Coding & developer toolsGit, CI & testing
ai-workflows Scope Classifygithub.com—Reusable GitHub Actions step that gates how much CI a pull request gets (full or partial CI, AI review, release notes, e2e) with Jev Choices over a repository rubric, with deterministic always-full overrides.Coding & developer toolsGit, CI & testing
ai-workflows Scope Classify — repogithub.com—Reusable GitHub Actions step that gates how much CI a pull request gets (full or partial CI, AI review, release notes, e2e) with Jev Choices over a repository rubric, with deterministic always-full overrides.Coding & developer toolsGit, CI & testing
Clausuragithub.com203CI agent CLI that gates pipelines on structured findings, with optional Jev verification that asks per finding whether it is genuine and supported by its own evidence before gating.Coding & developer toolsGit, CI & testing
commit-changeloggithub.com0Turns free-form git commits into a Keep a Changelog section: each commit's own first line, placed by its conventional-commit prefix or by TypeSafe's Jev model, with unsure commits listed for review.Coding & developer toolsGit, CI & testing
DeliveryGuardgithub.com10Evidence-driven delivery gates for coding agents covering specs, acceptance, repairs and releases, with an optional Jev adapter that adds a summary-based auto-review signal for ordinary delivery candidates.Coding & developer toolsGit, CI & testing
DeliveryGuard — docsgithub.com10Evidence-driven delivery gates for coding agents covering specs, acceptance, repairs and releases, with an optional Jev adapter that adds a summary-based auto-review signal for ordinary delivery candidates.Coding & developer toolsGit, CI & testing
EntropyLab Jev adversarial testsgithub.com—Scheduled CI job for an offline Bitcoin wallet calculator that explores the built page with hostile inputs and has Jev judge the outcomes, degrading to invariant checks without a key.Coding & developer toolsGit, CI & testing
EntropyLab Jev adversarial tests — appentropylab.online—Scheduled CI job for an offline Bitcoin wallet calculator that explores the built page with hostile inputs and has Jev judge the outcomes, degrading to invariant checks without a key.Coding & developer toolsGit, CI & testing
hekajevgithub.com6A hundred hands through Git history — reproducible commit analytics powered by Jev.Coding & developer toolsGit, CI & testing
hushgithub.com1GitHub Action that labels new issues and flags spam, needs-more-info and possible duplicates in one Jev call, applying each only above a maintainer-set threshold and saying nothing below it; 202-530 ms per issue.Coding & developer toolsGit, CI & testing
jev-ci-selectorgithub.com2GitHub Action where you describe what each CI check verifies and Jev reads the PR diff to output booleans for which jobs to run, with a pure policy engine and shadow mode.Coding & developer toolsGit, CI & testing
jev-ci-triagegithub.com0Classify every failing CI test as regression, flaky, environment, test-data or unknown. Rules first, Jev (TypeSafe) for the ambiguous ones. Never reruns, never masks.Coding & developer toolsGit, CI & testing
jev-e2egithub.com4Plain-English end-to-end tests for web apps: Jev picks controls on the page, Playwright runs the steps, and each case ends PASS, FAIL or BLOCKED with evidence.Coding & developer toolsGit, CI & testing
jev-e2e — demox.com—Plain-English end-to-end tests for web apps: Jev picks controls on the page, Playwright runs the steps, and each case ends PASS, FAIL or BLOCKED with evidence.Coding & developer toolsGit, CI & testing
jev-playwrightgithub.com6Jev-powered Playwright test selectionCoding & developer toolsGit, CI & testing
jev-test-filtergithub.com30CLI and agent skill that scores every test against a git diff in one Jev round trip and prints filter arguments for vitest, Jest, node:test, Playwright, cargo test and go test so only plausibly affected tests run.Coding & developer toolsGit, CI & testing
jev-test-impactgithub.com0Run only the tests your change touched: import graph + ownership rules first, Jev decides the ambiguous rest, fail-safe to the full suite. Replay benchmark with real ground truth.Coding & developer toolsGit, CI & testing
jevassertgithub.com0Record/replay regression tests for Jev (TypeSafe System One) question packs — accuracy, calibration and cost gates in CICoding & developer toolsGit, CI & testing
Leanestgithub.com3Local-first test selector that sits in front of an existing test runner and skips tests judged unaffected by a code change, using classifier.dev by default or Jev/Laya.Coding & developer toolsGit, CI & testing
Leanest — siteleanest.pages.dev—Local-first test selector that sits in front of an existing test runner and skips tests judged unaffected by a code change, using classifier.dev by default or Jev/Laya.Coding & developer toolsGit, CI & testing
lisptc AGENTS.md checkergithub.com—CI script in the lisptc agent-oriented Lisp monorepo that uses Jev via OpenRouter to flag AGENTS.md files that explain how code works instead of stating rules and pointers, failing or warning on set probability thresholds.Coding & developer toolsGit, CI & testing
lisptc AGENTS.md checker — blogd4shi.com—CI script in the lisptc agent-oriented Lisp monorepo that uses Jev via OpenRouter to flag AGENTS.md files that explain how code works instead of stating rules and pointers, failing or warning on set probability thresholds.Coding & developer toolsGit, CI & testing
lisptc AGENTS.md checker — repogithub.com—CI script in the lisptc agent-oriented Lisp monorepo that uses Jev via OpenRouter to flag AGENTS.md files that explain how code works instead of stating rules and pointers, failing or warning on set probability thresholds.Coding & developer toolsGit, CI & testing
Paxeer X Jev checksgithub.com—CI advisory tool inside an agent execution network that uses Jev to check pull request and commit message coherence, triage cargo-deny advisories, test failures and README translation drift.Coding & developer toolsGit, CI & testing
pytest-jevgithub.com2Plugin for pytest that adds semantic assertions on LLM app output: claims about one text go to Jev in a single request, and each test reports the probability per claim, failing uncertain ones by default.Coding & developer toolsGit, CI & testing
semantic-assertgithub.com4Testing library for asserting plain-English claims about captured UI or text state, such as whether an alert tells the user how to recover, with Jev as the default judge and pass thresholds in code, plus Playwright helpers.Coding & developer toolsGit, CI & testing
semantic-assert — appmengdi.dev—Testing library for asserting plain-English claims about captured UI or text state, such as whether an alert tells the user how to recover, with Jev as the default judge and pass thresholds in code, plus Playwright helpers.Coding & developer toolsGit, CI & testing
shipit-gategithub.com0The AI deploy gate. Ask Jev if it's safe to ship — before you push. Blocks scary Friday deploys.Coding & developer toolsGit, CI & testing
Tester Army e2etester.army—Upcoming open-source TypeScript framework for agent-run end-to-end tests of web and mobile apps, demoed with Jev choosing the next action at each step while assertions stay in code.Coding & developer toolsGit, CI & testing
Tester Army e2e — demox.com—Upcoming open-source TypeScript framework for agent-run end-to-end tests of web and mobile apps, demoed with Jev choosing the next action at each step while assertions stay in code.Coding & developer toolsGit, CI & testing
TinaCMS issue dedupergithub.com—GitHub Actions script in the TinaCMS repo that checks each new issue against all open issues with Jev and, when confident, comments with the original and adds a Duplicate label.Coding & developer toolsGit, CI & testing
TinaCMS issue deduper — apptina.io—GitHub Actions script in the TinaCMS repo that checks each new issue against all open issues with Jev and, when confident, comments with the original and adds a Duplicate label.Coding & developer toolsGit, CI & testing
Coding model routers106 of 106 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
Agent-Workflow TypeSafe routinggithub.com—Optional TypeSafe provider in Agent-Workflow, a coding-agent orchestration tool, that asks Jev for advisory task class, missing-authorization and semantic-risk judgments while deterministic policy keeps final authority.Coding & developer toolsCoding model routers
Astra-Aresgithub.com294Has Jev pick the reasoning effort and how long to hold it for a running Codex task, on a patched Codex CLI built from upstream source.Coding & developer toolsCoding model routers
auto-model-routergithub.com6Cost-, cache- and quota-aware LLM router with a Claude Code gateway shim, where a Jev-class model (hosted Jev or local Laya) classifies each turn's topic, difficulty and tool needs before picking the cheapest capable model.Coding & developer toolsCoding model routers
cli-ck Jev task routergithub.com—Task router in cli-ck, an AI-native terminal and dev workspace built on Tauri, that asks Jev to choose one pre-authorized model tier for each agent task before the turn is dispatched.Coding & developer toolsCoding model routers
cli-ck Jev task router — appcli-ck.github.io—Task router in cli-ck, an AI-native terminal and dev workspace built on Tauri, that asks Jev to choose one pre-authorized model tier for each agent task before the turn is dispatched.Coding & developer toolsCoding model routers
cli-ck Jev task router — repogithub.com—Task router in cli-ck, an AI-native terminal and dev workspace built on Tauri, that asks Jev to choose one pre-authorized model tier for each agent task before the turn is dispatched.Coding & developer toolsCoding model routers
coderelaygithub.com5Router that scans the coding-agent CLIs installed on a machine (Claude Code, Codex, pi, omp) and dispatches each task to the best one, by local rules and scoring or, in jev mode, by Jev reading the task intent.Coding & developer toolsCoding model routers
Codex reasoning-effort routerx.com—Codex setup where Jev changes GPT-6's reasoning effort during a task, adding thinking when stuck and cutting it on routine steps, for 50% lower Astra costs in the author's tests.Coding & developer toolsCoding model routers
Codex Router Jev routegithub.com—Codex Router exposes an unlisted OpenRouter Decisions route for jev-latest, kept out of Codex's model picker and meant for an explicit local integration like a compaction pruner.Coding & developer toolsCoding model routers
codex-jev-routergithub.com4Portable Codex subagent-routing workflow using Jev Choice/Noul and local confidence gates, with a Sol fallback; invoked through project instructions rather than an automatic spawn hook.Coding & developer toolsCoding model routers
codex-jev-routergithub.com5Local bridge for the OpenAI Codex CLI that runs a loopback Responses API proxy, asks Jev to pick the Codex model and reasoning effort for each fresh turn, and falls back to the current model if Jev is unavailable.Coding & developer toolsCoding model routers
Cortexgithub.com5Meta-router that sits above coding-agent workflow systems in Claude Code, Cursor, Codex and others to pick the workflow, specialist and effort per task, with an optional Jev adviser for task classification.Coding & developer toolsCoding model routers
Epistemic Protocols Routegithub.com—Claude Code plugin that routes session context to the epistemic protocol matching its deficit, with an opt-in Jev advisory channel that nudges up to three protocols.Coding & developer toolsCoding model routers
Firstmate Jev dispatchgithub.com6,796Opt-in dispatch resolver in Firstmate, a crew manager for parallel coding agents, that matches each task brief to a dispatch rule with one Jev Choice before local policy picks the agent profile.Coding & developer toolsCoding model routers
Firstmate Jev dispatch — repogithub.com6,796Opt-in dispatch resolver in Firstmate, a crew manager for parallel coding agents, that matches each task brief to a dispatch rule with one Jev Choice before local policy picks the agent profile.Coding & developer toolsCoding model routers
Fleet AI Gateway Jev routinggithub.com—AI Gateway routing in Fleet, a local-first console for supervising Claude Code and other coding-agent sessions, where Jev can assign models and effort to delegated agent runs from task, benchmark and quota headroom data.Coding & developer toolsCoding model routers
Fleet AI Gateway Jev routing — appsbluemin.github.io—AI Gateway routing in Fleet, a local-first console for supervising Claude Code and other coding-agent sessions, where Jev can assign models and effort to delegated agent runs from task, benchmark and quota headroom data.Coding & developer toolsCoding model routers
Fleet AI Gateway Jev routing — repogithub.com—AI Gateway routing in Fleet, a local-first console for supervising Claude Code and other coding-agent sessions, where Jev can assign models and effort to delegated agent runs from task, benchmark and quota headroom data.Coding & developer toolsCoding model routers
Fleet Prime intent routergithub.com—Composer intent router in Fleet Prime, a local web workspace for the Prime Agent coding runtime, that asks Jev whether typed text requests one of the app's built-in commands or is an ordinary coding request.Coding & developer toolsCoding model routers
Fleet Prime intent router — appqredence.ai—Composer intent router in Fleet Prime, a local web workspace for the Prime Agent coding runtime, that asks Jev whether typed text requests one of the app's built-in commands or is an ordinary coding request.Coding & developer toolsCoding model routers
Fleet Prime intent router — repogithub.com—Composer intent router in Fleet Prime, a local web workspace for the Prime Agent coding runtime, that asks Jev whether typed text requests one of the app's built-in commands or is an ordinary coding request.Coding & developer toolsCoding model routers
flow-next Jev judgegithub.com—Optional Jev judge in flow-next, an agentic engineering workflow, that answers narrow routing, task-tier and memory questions in one request per decision point.Coding & developer toolsCoding model routers
flow-next Jev judge — appflow-next.dev—Optional Jev judge in flow-next, an agentic engineering workflow, that answers narrow routing, task-tier and memory questions in one request per decision point.Coding & developer toolsCoding model routers
Forgewright Jev skill routergithub.com—Optional, default-off Jev adapter in the Forgewright AI engineering harness that picks a skill for ambiguous task routing from a strict shortlist, with a budget cap, pinned model version and local-router fallback.Coding & developer toolsCoding model routers
Forgewright Jev skill router — repogithub.com—Optional, default-off Jev adapter in the Forgewright AI engineering harness that picks a skill for ambiguous task routing from a strict shortlist, with a budget cap, pinned model version and local-router fallback.Coding & developer toolsCoding model routers
Helm — discussionwww.reddit.com—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.Coding & developer toolsCoding model routers
Herdr Jev Routergithub.com—Policy command for Herdr's agent.spawn path that checks cached provider capacity, asks Jev for typed choices, validates them and writes a redacted audit record before a coding agent child is launched.Coding & developer toolsCoding model routers
herdr model routerwww.youtube.com—Terminal setup where a model-router CLI asks Jev which coding agent (Claude Code, Codex, or Cursor), which model, and how much reasoning effort each task gets.Coding & developer toolsCoding model routers
Jackalope Jev supportgithub.com—Jev task routing in Jackalope, a desktop workspace for coding agents: each candidate agent and model gets a reasoning-fit Score and a tool-support Noul for the task, and code picks the worker.Coding & developer toolsCoding model routers
Jackalope Jev support — appjackalope.dev—Jev task routing in Jackalope, a desktop workspace for coding agents: each candidate agent and model gets a reasoning-fit Score and a tool-support Noul for the task, and code picks the worker.Coding & developer toolsCoding model routers
Jackalope Jev support — repogithub.com—Jev task routing in Jackalope, a desktop workspace for coding agents: each candidate agent and model gets a reasoning-fit Score and a tool-support Noul for the task, and code picks the worker.Coding & developer toolsCoding model routers
Jev Auto Router — articleminilv.github.io—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.Coding & developer toolsCoding model routers
Jev Codex Routergithub.com277Per-call router for Codex where Jev picks the model and thinking effort for every model call, including tool continuations; a historical simulation estimated about 60% below full Astra on 237 turns.Coding & developer toolsCoding model routers
Jev Codex Router — demox.com—Per-call router for Codex where Jev picks the model and thinking effort for every model call, including tool continuations; a historical simulation estimated about 60% below full Astra on 237 turns.Coding & developer toolsCoding model routers
Jev for LLM model routingdev.classmethod.jp—Experiment replacing the LLM classifier in NeMo Switchyard-style coding-agent routing with a four-tier Jev Choice; 40 calls all matched the expected tier at about 0.66 s median and $0.000025 per call.Coding & developer toolsCoding model routers
Jev model router for Claude Codegithub.com30,825Hook that asks for a tier Choice, an effort Score, and a risk Noul in one request, then adjusts model and reasoning effort.Coding & developer toolsCoding model routers
Jev Model Router for Claude Codex.com—Claude Code mod that has Jev classify the subagent model, main model (only at session start to keep the cache) and effort level for every request, via the TypeSafe API or Vercel AI Gateway.Coding & developer toolsCoding model routers
Jev Model Router for Claude Code — docsaitmpl.com—Claude Code mod that has Jev classify the subagent model, main model (only at session start to keep the cache) and effort level for every request, via the TypeSafe API or Vercel AI Gateway.Coding & developer toolsCoding model routers
Jev Skill Router for Codexgithub.com2Codex plugin that has Jev pick which skills apply to each request and passes their paths and relevance probabilities to Codex, with a local analytics dashboard.Coding & developer toolsCoding model routers
Jev Skill Router for Codex — demox.com—Codex plugin that has Jev pick which skills apply to each request and passes their paths and relevance probabilities to Codex, with a local analytics dashboard.Coding & developer toolsCoding model routers
jev-affectedgithub.com1Semantic task routing for software development, powered by Jev.Coding & developer toolsCoding model routers
Jev-chooses-a-LLMgithub.com2Cursor router exposed over MCP where Jev picks a cost, balanced or intelligence route for each task before Cursor delegates to the matching sub-agent or model.Coding & developer toolsCoding model routers
jev-codex-model-and-effort-router — demox.com—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.Coding & developer toolsCoding model routers
jev-fitjev-fit.com—Hosted checker that reads a pasted software idea and says whether it suits plain code, Jev, or a reasoning LLM, answering "not sure" when confidence is low.Coding & developer toolsCoding model routers
jev-flash-router — npmwww.npmjs.com—MCP server for Cursor, Windsurf, Claude Desktop and Claude Code that lets coding agents offload small decisions, such as which file to edit or whether a diff breaks tests, to Jev via OpenRouter in about 150ms.Coding & developer toolsCoding model routers
jev-gateway — linkgithub.com1Local LLM gateway for Codex, Claude Code, OpenCode, and Gemini clients that hands the which-tool-to-call decision to Jev and forwards all other traffic to your usual model.Coding & developer toolsCoding model routers
jev-gateway — videowww.youtube.com—Local LLM gateway for Codex, Claude Code, OpenCode, and Gemini clients that hands the which-tool-to-call decision to Jev and forwards all other traffic to your usual model.Coding & developer toolsCoding model routers
jev-model-routergithub.com—CLI & Pipelines: Routes prompts to the right Claude tier (Haiku/Sonnet/Opus) using TypeSafe's Jev modelCoding & developer toolsCoding model routers
jev-model-router (Claude Code)github.com3Claude Code hook that asks Jev, before each turn, which tier from Haiku to Fable should answer and how hard it should think, then sends every model request in that turn to the chosen model.Coding & developer toolsCoding model routers
jev-review-modelgithub.com—Dotfiles script that reads a PR diff on stdin and asks Jev whether Opus, Sonnet or either model should review it.Coding & developer toolsCoding model routers
jev-review-model — repogithub.com—Dotfiles script that reads a PR diff on stdin and asks Jev whether Opus, Sonnet or either model should review it.Coding & developer toolsCoding model routers
jev-routegithub.com—CLI & Pipelines: Async, typed intent & tool routing for Python agents: Jev decides, a confidence gate stops weak decisions before they reach a privileged tool.Coding & developer toolsCoding model routers
jev-routergithub.com—CLI & Pipelines: Route any task to the right AI agent in under 1 second using Jev (TypeSafe System One). Supports Claude, ChatGPT, Cursor, and Antigravity with auto-launch on macOS, Windows, and Linux.Coding & developer toolsCoding model routers
jev-routergithub.com1Zero-dependency CLI and Claude Code/Codex skill that hands narrow agent judgments, like which of 184 files touch auth, to Jev instead of the LLM.Coding & developer toolsCoding model routers
jev-routergithub.com—Self-hosted model router for coding agents and OpenAI-compatible clients that has Jev assess each new task, then applies YAML rules to pick a model and reasoning effort and pins it for the session.Coding & developer toolsCoding model routers
jev-router (flaviusapop)github.com—Launcher for Claude Code, Codex, the Grok CLI and opencode that has Jev route each turn to the cheapest model and reasoning depth that can finish it, rewriting only the model field and reusing existing logins.Coding & developer toolsCoding model routers
jev-router (gmaxxxie)github.com—Pi extension that classifies every prompt with Jev and switches to the cheapest capable model and thinking depth on the existing Pi login, at about $0.00002 and 0.9 s per decision.Coding & developer toolsCoding model routers
jev-router (ianlintner)github.com—Loopback shadow-mode proxy that compares Jev feature extraction with Switchyard's coding_agent routing classifier without changing live routes, exporting Prometheus metrics and Grafana dashboards.Coding & developer toolsCoding model routers
jev-router — demox.com—Wraps Claude Code and Codex so each turn goes to the cheapest model tier that can finish it, chosen with one dynamically built Choice.Coding & developer toolsCoding model routers
jev-router — discussionnews.ycombinator.com—Wraps Claude Code and Codex so each turn goes to the cheapest model tier that can finish it, chosen with one dynamically built Choice.Coding & developer toolsCoding model routers
jevortgithub.com0Pi extension that selects model-compatible reasoning effort with TypeSafe JevCoding & developer toolsCoding model routers
kojo Jev effort classifiergithub.com—Per-turn effort classifier in kojo, a mobile remote control for Claude Code, Codex and Grok Build with persistent agents, that asks Jev whether a message needs low, medium or high reasoning effort, falling back to Claude.Coding & developer toolsCoding model routers
kojo Jev effort classifier — repogithub.com—Per-turn effort classifier in kojo, a mobile remote control for Claude Code, Codex and Grok Build with persistent agents, that asks Jev whether a message needs low, medium or high reasoning effort, falling back to Claude.Coding & developer toolsCoding model routers
Kungfu Kanban Jev shadow routinggithub.com—Shadow model routing in Kungfu Kanban, a local board where Claude Code agents work cards: Jev scores each card's difficulty and records which of Haiku, Sonnet or Opus it would pick, for later comparison with real outcomes.Coding & developer toolsCoding model routers
Kungfu Kanban Jev shadow routing — appkungfu-kanban.com—Shadow model routing in Kungfu Kanban, a local board where Claude Code agents work cards: Jev scores each card's difficulty and records which of Haiku, Sonnet or Opus it would pick, for later comparison with real outcomes.Coding & developer toolsCoding model routers
Kungfu Kanban Jev shadow routing — repogithub.com—Shadow model routing in Kungfu Kanban, a local board where Claude Code agents work cards: Jev scores each card's difficulty and records which of Haiku, Sonnet or Opus it would pick, for later comparison with real outcomes.Coding & developer toolsCoding model routers
Local harness and model routingwww.reddit.com—Claude Code hook that asks Jev to route each task before delegation: mechanical work to Haiku, harder work to Opus sub-agents, long implementation runs to Codex or OpenCode.Coding & developer toolsCoding model routers
Local harness and model routing — demox.com—Claude Code hook that asks Jev to route each task before delegation: mechanical work to Haiku, harder work to Opus sub-agents, long implementation runs to Codex or OpenCode.Coding & developer toolsCoding model routers
Milowww.npmjs.com—Routing gateway for coding harnesses such as pi, Codex and Claude Code: send model milo/auto and each request is routed through Jev to a fast, mid or frontier model on OpenRouter.Coding & developer toolsCoding model routers
model-pickergithub.com0Model Picker: JEV-powered Codex model and effort recommendations as task scope grows. Manual switching, private local state, MIT licensed.Coding & developer toolsCoding model routers
nanocodex Jev routinggithub.com—Building blocks for Rust OpenAI agents that use Jev for thread model routing, choosing the backend per thread with a fixed-vocabulary reliability and failure classifier.Coding & developer toolsCoding model routers
nanocodex Jev routing — docsdocs.rs—Building blocks for Rust OpenAI agents that use Jev for thread model routing, choosing the backend per thread with a fixed-vocabulary reliability and failure classifier.Coding & developer toolsCoding model routers
OpenChamber Jev routergithub.com10,105Optional automatic model router in OpenChamber, a workspace for OpenCode agents, where Jev classifies each message into a task category that maps to a model and reasoning level, and screens tool permissions.Coding & developer toolsCoding model routers
OpenChamber Jev router — appopenchamber.dev—Optional automatic model router in OpenChamber, a workspace for OpenCode agents, where Jev classifies each message into a task category that maps to a model and reasoning level, and screens tool permissions.Coding & developer toolsCoding model routers
OpenChamber Jev router — repogithub.com10,105Optional automatic model router in OpenChamber, a workspace for OpenCode agents, where Jev classifies each message into a task category that maps to a model and reasoning level, and screens tool permissions.Coding & developer toolsCoding model routers
OpenCode task routerdev.to—Tool for OpenCode that asks Jev three questions about each task (coordination, uncertainty, consequences) via OpenRouter and picks a strong or weak model against a consequence-based threshold.Coding & developer toolsCoding model routers
OpenCodexgithub.com—Codex and Claude Code provider proxy with an optional Jev route for selecting a model and reasoning effort.Coding & developer toolsCoding model routers
Pi Jev Routergithub.com9Task-boundary model routing extension for Pi Coding Agent: Jev judges task tier, context sufficiency and consequences, and code picks a model from what Pi can actually use.Coding & developer toolsCoding model routers
Pi Jev Router — demox.com—Task-boundary model routing extension for Pi Coding Agent: Jev judges task tier, context sufficiency and consequences, and code picks a model from what Pi can actually use.Coding & developer toolsCoding model routers
Pi-Bifrostgithub.com53Model router for the Pi coding agent that picks a model per prompt by tier and strategy, with optional TypeSafe/Jev judge tiers classifying each prompt's complexity.Coding & developer toolsCoding model routers
Pi-Bifrost — appiamaamir.github.io—Model router for the Pi coding agent that picks a model per prompt by tier and strategy, with optional TypeSafe/Jev judge tiers classifying each prompt's complexity.Coding & developer toolsCoding model routers
Pi-Bifrost — discussionwww.reddit.com—Model router for the Pi coding agent that picks a model per prompt by tier and strategy, with optional TypeSafe/Jev judge tiers classifying each prompt's complexity.Coding & developer toolsCoding model routers
pi-ext-model-routergithub.com0pi.dev extension to select models using TypeSafe JevCoding & developer toolsCoding model routers
pi-jev-effortgithub.com—Pi extension that sets each prompt's thinking level from a Jev Score of how hard it is (trivial to hard), capped by your remaining quota, at about 250 ms and $0.000015 per judgment.Coding & developer toolsCoding model routers
pi-jev-model-routergithub.com—CLI & Pipelines: Route pi prompts to task-appropriate model tiers with TypeSafe Jev typed judgments. Budget-aware, with automatic fallback.Coding & developer toolsCoding model routers
pi-jev-routergithub.com14Model router for the pi coding agent that classifies each task with Jev, then picks the knee of a Pareto frontier over quality, cost and latency across the OpenRouter catalog.Coding & developer toolsCoding model routers
pi-jev-router — demonews.ycombinator.com—Model router for the pi coding agent that classifies each task with Jev, then picks the knee of a Pareto frontier over quality, cost and latency across the OpenRouter catalog.Coding & developer toolsCoding model routers
pi-typesafe-router — npmwww.npmjs.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.Coding & developer toolsCoding model routers
RepoPrompt Jev task routergithub.com—RepoPrompt CE, a native macOS context-engineering app for coding agents, adds a Jev backend to its Agent Mode model router that routes each task without a text model.Coding & developer toolsCoding model routers
RepoPrompt Jev task router — apprepoprompt.com—RepoPrompt CE, a native macOS context-engineering app for coding agents, adds a Jev backend to its Agent Mode model router that routes each task without a text model.Coding & developer toolsCoding model routers
RUNE Jev routinggithub.com—Decision-API routing in RUNE, a local-model coding agent that only claims done when tests pass, where Jev answers bounded routing questions and file-role hints before the agent acts.Coding & developer toolsCoding model routers
RUNE Jev routing — repogithub.com—Decision-API routing in RUNE, a local-model coding agent that only claims done when tests pass, where Jev answers bounded routing questions and file-role hints before the agent acts.Coding & developer toolsCoding model routers
Sabigithub.com13Per-round model, effort and provider router for coding harnesses that uses trajectory evidence to pick cheap, mid or strong tiers and asks Jev only when a round is ambiguous.Coding & developer toolsCoding model routers
Sabi — npmwww.npmjs.com—Per-round model, effort and provider router for coding harnesses that uses trajectory evidence to pick cheap, mid or strong tiers and asks Jev only when a round is ambiguous.Coding & developer toolsCoding model routers
skill-routergithub.com23Hook-driven Claude Code router that picks the skill, agent, model and thinking depth before any tool fires, with Jev choosing over the whole skill index; in its eval Jev routed 39/63 prompts correctly vs 10/66 for the old pipeline.Coding & developer toolsCoding model routers
skill-router — evalgithub.com23Hook-driven Claude Code router that picks the skill, agent, model and thinking depth before any tool fires, with Jev choosing over the whole skill index; in its eval Jev routed 39/63 prompts correctly vs 10/66 for the old pipeline.Coding & developer toolsCoding model routers
Switchboard — demowww.reddit.com—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.Coding & developer toolsCoding model routers
Switchyardgithub.com6Subscription-aware model router for Pi, Oh My Pi and OpenCode that has Jev classify each coding task's kind and complexity via OpenRouter Decisions, then picks the cheapest capable model and reasoning effort.Coding & developer toolsCoding model routers
tuxevil-rotatorgithub.com63OpenAI-compatible gateway rotating free-tier LLM accounts for coding agents, where model "auto" can use Jev to pick a currently routable model with deterministic fail-open fallback.Coding & developer toolsCoding model routers
Vekil Jev policy routinggithub.com—Example policy for Vekil, a Go AI gateway that fronts GitHub Copilot and other providers for Claude Code and Codex CLI, where Jev classifies each request to a lightweight or a powerful model route.Coding & developer toolsCoding model routers
Vekil Jev policy routing — repogithub.com—Example policy for Vekil, a Go AI gateway that fronts GitHub Copilot and other providers for Claude Code and Codex CLI, where Jev classifies each request to a lightweight or a powerful model route.Coding & developer toolsCoding model routers
VexJoy Jev routinggithub.com424Jev router in the VexJoy toolkit for Claude Code and Codex that matches a plain-English /do request to one of 43 specialist agents, skills and workflows, injected by a hook before the model's first token.Coding & developer toolsCoding model routers
VexJoy Jev routing — appvexjoy.com—Jev router in the VexJoy toolkit for Claude Code and Codex that matches a plain-English /do request to one of 43 specialist agents, skills and workflows, injected by a hook before the model's first token.Coding & developer toolsCoding model routers
VexJoy Jev routing — repogithub.com424Jev router in the VexJoy toolkit for Claude Code and Codex that matches a plain-English /do request to one of 43 specialist agents, skills and workflows, injected by a hook before the model's first token.Coding & developer toolsCoding model routers
Xum auto model routinggithub.com—Xum, a desktop coding-agent multiplexer, sends each Auto-mode prompt to an evaluation model for a single difficulty-tier choice and runs the turn on that tier's model and thinking level.Coding & developer toolsCoding model routers
Xum auto model routing — appxum.coder.com—Xum, a desktop coding-agent multiplexer, sends each Auto-mode prompt to an evaluation model for a single difficulty-tier choice and runs the turn on that tier's model and thinking level.Coding & developer toolsCoding model routers
Coding context & memory65 of 65 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
AgentDirgithub.com22Local-first flight recorder for coding agents that records, replays and audits software tasks; an opt-in Jev memory reranker filters redacted context before it is handed back to the agent.Coding & developer toolsCoding context & memory
algal-skillswww.npmjs.com—Token-efficient agent skills for Devin, Claude Code and Codex that keep raw tool output out of context, with Jev optionally answering typed triage questions about diffs, sources and drafts.Coding & developer toolsCoding context & memory
Anvil Jev annotationsgithub.com—Local-first project state layer for multi-agent coding work with optional Jev annotations on requirements, tasks and evidence that are never allowed to mutate authoritative state.Coding & developer toolsCoding context & memory
Anvil Jev annotations — repogithub.com—Local-first project state layer for multi-agent coding work with optional Jev annotations on requirements, tasks and evidence that are never allowed to mutate authoritative state.Coding & developer toolsCoding context & memory
caliber-jev-compaction — apptrycaliber.ai—Claude Code function-hook plugin shipped with Caliber that replaces built-in context compaction with Jev decisions, loaded from disk as unbundled TypeScript.Coding & developer toolsCoding context & memory
chunk-standalonegithub.com0Finds RAG chunks that cannot be understood on their own and suggests keep, merge or split, using TypeSafe's Jev model.Coding & developer toolsCoding context & memory
codex-jev-compactiongithub.com0Codex plugin that restores Jev-selected tool evidence after context compactionCoding & developer toolsCoding context & memory
compact-advisergithub.com191Plugin for Pi, Claude Code, Codex CLI, and Grok that asks Jev whether the current unit of work is finished and whether it is hands-on or coordination, then hints or triggers /compact at safe boundaries.Coding & developer toolsCoding context & memory
compact-jevgithub.com0/compact-jev: Jev-guided verbatim compaction for Claude Code through Vercel AI Gateway, separate from /compact.Coding & developer toolsCoding context & memory
compactiogithub.com3System 1 for your coding agent: cut token use by deciding, not generating.Coding & developer toolsCoding context & memory
cursor-clijev-compactiongithub.com1Context recovery plugin for the Cursor CLI agent that captures tool calls and results, has Jev score each with atomic noul questions, and re-injects the kept facts verbatim after native compaction.Coding & developer toolsCoding context & memory
czipgithub.com6Compresses Claude Code, Codex, Grok and Hermes agent sessions into a searchable pack that new sessions query instead of reloading history; an optional Jev gate decides session merges and which large tool outputs to keep.Coding & developer toolsCoding context & memory
dsh-jev-prunegithub.com4DeepSeek Harness plugin that replaces size-based tool-result trimming and summary-based compaction with Jev keep or drop judgments, plus deterministic receipt compaction, with a pluggable judgment backend.Coding & developer toolsCoding context & memory
dsh-jev-toolsgithub.com11DeepSeek Harness plugin that uses Jev to prune tool output over 2,000 tokens to task-relevant segments, flag injected instructions in fetched pages, suggest a skill, and check completion claims against evidence.Coding & developer toolsCoding context & memory
dsh-jev-tools — npmwww.npmjs.com—DeepSeek Harness plugin that uses Jev to prune tool output over 2,000 tokens to task-relevant segments, flag injected instructions in fetched pages, suggest a skill, and check completion claims against evidence.Coding & developer toolsCoding context & memory
fast-compactgithub.com5/fc for Claude Code: shrink old tool output in about a second — Jev keeps what's still needed, every cut saved to a file. Your /compact stays untouched.Coding & developer toolsCoding context & memory
fast-compaction-dshgithub.com3Verdict-based context compaction for DeepSeek Harness — replaces lossy LLM summaries with fast keep/truncate/drop decisions from jev-latest; everything kept stays verbatim. Port of tamaratran/fast-jev-compaction.Coding & developer toolsCoding context & memory
fast-jev-codexgithub.com0Preserve exact context across Codex compaction with Jev. Open-source plugin, CLI, and library.Coding & developer toolsCoding context & memory
fast-jev-compactiongithub.com0Verbatim, Jev-guided context compaction for Claude Code (fork with local additions)Coding & developer toolsCoding context & memory
fast-jev-compactiongithub.com7,282Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction — demox.com—Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction — demo 2x.com—Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction — demo 3x.com—Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction — discussionwww.reddit.com—Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction — discussion 2www.reddit.com—Claude Code plugin that replaces the compaction summary with two Nouls per tool call, keep the call and keep its result; Nous Research found the 0.5 default kept none of 851 calls. The npm package is another publisher's.Coding & developer toolsCoding context & memory
fast-jev-compaction-layagithub.com6Key-free opencode plugin that compacts context by asking a local Laya server on Apple Silicon to keep, truncate or drop each old tool call, leaving the assistant's own messages intact, at about 7-15 ms per decision.Coding & developer toolsCoding context & memory
jevgithub.com0A wrapper that gets more value out of TypeSafe's JEV model: transcript compaction, cost-control modes, and a local usage ledger for Claude Code and Codex.Coding & developer toolsCoding context & memory
Jev Codex Token Savergithub.com3Codex plugin whose MCP server gathers search and log evidence locally, has Jev score a bounded candidate packet for relevance, and returns only the selected exact excerpts to keep large results out of context.Coding & developer toolsCoding context & memory
Jev Codex Token Saver — demowww.reddit.com—Codex plugin whose MCP server gathers search and log evidence locally, has Jev score a bounded candidate packet for relevance, and returns only the selected exact excerpts to keep large results out of context.Coding & developer toolsCoding context & memory
Jev memory filter for Codex and Claude Codex.com—Japanese write-up of a setup where Jev trims memory candidates passed to Codex and Claude Code from up to 8 to 3, a 62.5% cut, with a full copy-paste prompt for implementing it.Coding & developer toolsCoding context & memory
Jev selection for deepagentsx.com—Replaces LangChain deepagents' SummarizationMiddleware with Jev-based context selection; on a code benchmark it cut cost 83.9% for GLM-5.3 and 91.8% for Opus 4.8 with all 12 programs still passing.Coding & developer toolsCoding context & memory
Jev Skill Suggestion for Claude Codex.com—Claude Code mod that keeps skills out of the context window: for every request Jev picks the best-matching skill from the list and only that skill is injected.Coding & developer toolsCoding context & memory
Jev Skill Suggestion for Claude Code — docsaitmpl.com—Claude Code mod that keeps skills out of the context window: for every request Jev picks the best-matching skill from the list and only that skill is injected.Coding & developer toolsCoding context & memory
jev-compactgithub.com0Prune-not-summarise context compaction for Claude Code, guided by Jev (TypeSafe System One)Coding & developer toolsCoding context & memory
jev-compact — discussionwww.reddit.com—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.Coding & developer toolsCoding context & memory
jev-factkeep-compactiongithub.com0Fact-keeping Jev compaction for Claude Code: a maintained fork of tamaratran/fast-jev-compaction where dropped tool calls keep their facts instead of being erasedCoding & developer toolsCoding context & memory
jev-prunergithub.com154Trims long Bash output with Jev after the command runs and before the model sees it; short output, errors, and structured formats pass untouched.Coding & developer toolsCoding context & memory
jev-pruner — discussionwww.reddit.com—Prunes large shell outputs chunk by chunk with one Noul each before they enter the context, keeping the original on disk.Coding & developer toolsCoding context & memory
JevCompactgithub.com0Lossless LLM session compaction for Claude Code, Codex and ZCode — Jev keep/drop decisions plus a Chinese-optimized lossless policy. No paraphrasing, no hallucinated summaries; 9/9 HANDOVER-FREE on a private benchmark.Coding & developer toolsCoding context & memory
JevKeepgithub.com1Codex plugin that has Jev pick useful conversation passages before context compaction and restores their original text afterward alongside Codex's own summary, for manual and automatic compaction.Coding & developer toolsCoding context & memory
jevmemgithub.com104Automatic project memory for Claude Code. Also works with Cursor and Codex.Coding & developer toolsCoding context & memory
lccgithub.com7Local Context Compiler that cleans, dedupes and compacts prompt context before it reaches the model; lcc compact --provider jev asks Jev for keep-probabilities per block and drops blocks verbatim without summarizing.Coding & developer toolsCoding context & memory
lossless-compactgithub.com0Verbatim, reversible, provenance-tracked context compaction for Claude Code (fork of tamaratran/fast-jev-compaction)Coding & developer toolsCoding context & memory
Memorax Code Jev gategithub.com—Memory plugin for AI coding that uses Jev to decide whether the current prompt warrants a memory search, so it only retrieves when it is likely to help.Coding & developer toolsCoding context & memory
Memorax Code Jev gate — appcode.memorax.net—Memory plugin for AI coding that uses Jev to decide whether the current prompt warrants a memory search, so it only retrieves when it is likely to help.Coding & developer toolsCoding context & memory
Nanocode with Jev memory compactionx.com—Fork of the nanocode agent with tools to grep a memory.md file and Jev compaction that removes unnecessary memories.Coding & developer toolsCoding context & memory
omp-jev-compaction — demox.com—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.Coding & developer toolsCoding context & memory
opencode-context-prunergithub.com4Continuous verbatim context pruning for OpenCode, powered by TypeSafe Jev. Port of fast-jev-compaction adapted to OpenCode's context hook.Coding & developer toolsCoding context & memory
opencode-jev-compactiongithub.com1Jev-powered context compaction plugin for opencode: prune stale tool calls and results before every request, keep the rest verbatim.Coding & developer toolsCoding context & memory
Parcha recall Jev judgmentsgithub.com—Agent skill for recalling past coding-agent sessions that is moving its regex date parsing, duplicate detection and content triage into Jev judgments, starting with an inactive boundary and replay harness.Coding & developer toolsCoding context & memory
Parcha recall Jev judgments — plangithub.com—Agent skill for recalling past coding-agent sessions that is moving its regex date parsing, duplicate detection and content triage into Jev judgments, starting with an inactive boundary and replay harness.Coding & developer toolsCoding context & memory
pi-dcpgithub.com17Pi extension that cuts token spend in long coding sessions by deduplicating tool calls, stripping errored payloads and summarizing closed work-streams, with experimental Jev context selection of what to keep.Coding & developer toolsCoding context & memory
pi-jev-compactgithub.com0Smart Compact for Harnesses using Jev modelCoding & developer toolsCoding context & memory
pi-jev-compactiongithub.com4Automatic Pi extension that stops sending stale tool output to the model while keeping the conversation and letting you retrieve originals without rerunning commands; no summary model.Coding & developer toolsCoding context & memory
pi-jev-compactionwww.npmjs.com—Pi coding-agent extension that replaces LLM-summary compaction with Jev keep, drop or truncate decisions on tool calls and results, falling back to Pi's built-in summary on errors.Coding & developer toolsCoding context & memory
pi-jev-context-curatorgithub.com5Pi extension that curates the context before each LLM call, asking Jev one Noul per new message unit whether it is essential to a future response and reusing past judgments by content fingerprint.Coding & developer toolsCoding context & memory
pi-savergithub.com0Dynamic context filter for Pi Coder using Jev by TypeSafe to save 75%+ on context tokens.Coding & developer toolsCoding context & memory
Sandogithub.com54Local context-management plugin for Claude Code and Codex that caps oversized tool output, with an optional TypeSafe shadow judge that only measures and never changes the request.Coding & developer toolsCoding context & memory
save-token-jevgithub.com79Portable context compaction for coding agents where Jev decides which tool calls to keep in full, keep with a bounded result, or drop, with adapters for Codex, OpenCode, Claude Code, and raw transcripts.Coding & developer toolsCoding context & memory
Sigil Jev rerankgithub.com—Local-first memory for coding agents, shared over MCP, that uses Jev to decide which retrieved facts to keep before injecting context.Coding & developer toolsCoding context & memory
Sigil Jev rerank — repogithub.com—Local-first memory for coding agents, shared over MCP, that uses Jev to decide which retrieved facts to keep before injecting context.Coding & developer toolsCoding context & memory
sofar typed-judge calibrationgithub.com—Script in an event-sourced memory tool for coding agents that calibrates Jev against the project record, judging whether decisions are standing constraints and whether they bear on tasks, then reporting agreement and calibration.Coding & developer toolsCoding context & memory
sofar typed-judge calibration — appsofar.sh—Script in an event-sourced memory tool for coding agents that calibrates Jev against the project record, judging whether decisions are standing constraints and whether they bear on tasks, then reporting agreement and calibration.Coding & developer toolsCoding context & memory
sofar typed-judge calibration — repogithub.com—Script in an event-sourced memory tool for coding agents that calibrates Jev against the project record, judging whether decisions are standing constraints and whether they bear on tasks, then reporting agreement and calibration.Coding & developer toolsCoding context & memory
windows-save-token-jev-setupgithub.com4Windows Codex skill that installs and verifies save-token-jev hooks, where a PreCompact hook has Jev pick the tool records worth keeping and a SessionStart hook restores them after native compaction.Coding & developer toolsCoding context & memory
More projects & source code833 of 833 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
&jev test-file classifiergithub.com—Nushell script in a personal dotfiles repo that sends each source file to Jev and returns the probability that it contains tests, fixtures or snapshots.Coding & developer toolsMore projects & source code
&jev test-file classifier — repogithub.com—Nushell script in a personal dotfiles repo that sends each source file to Jev and returns the probability that it contains tests, fixtures or snapshots.Coding & developer toolsMore projects & source code
5chan Jev dev helpersgithub.com—Optional Node scripts for a peer-to-peer imageboard that use Jev for browser test planning and benchmarks, blind review handoffs and contrastive translation evaluation.Coding & developer toolsMore projects & source code
Abidegithub.com209Hook plugin for Claude Code, Codex and OpenCode in which Jev checks each agent edit against your AGENTS.md and CLAUDE.md rules, about 300 ms per check, and makes the agent fix violations.Coding & developer toolsMore projects & source code
action-scheduler-triagegithub.com0Group the failed actions of WordPress's Action Scheduler by hook and error, then ask Jev which groups are safe to run again unchanged and which need a fix first. Read-only: it prints WP-CLI commands and never runs them.Coding & developer toolsMore projects & source code
action-scheduler-triage — READMEgithub.com0Read-only Action Scheduler export triage groups failed jobs in code and uses Jev to suggest retry, fix or human review; it never retries jobs itself.Coding & developer toolsMore projects & source code
Agent Stack Jev QAgithub.com—QA engine in Agent Stack, a multi-agent delivery workflow for Codex, Claude Code, Cursor and OpenCode, where web and mobile QA agents compress checkpoints to text and Jev returns functional, view and business verdicts.Coding & developer toolsMore projects & source code
Agent Stack Jev QA — repogithub.com—QA engine in Agent Stack, a multi-agent delivery workflow for Codex, Claude Code, Cursor and OpenCode, where web and mobile QA agents compress checkpoints to text and Jev returns functional, view and business verdicts.Coding & developer toolsMore projects & source code
agent-airlockgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
agent-airlock — READMEgithub.com0Agent Airlock is a shadow-mode safety console where Jev scores tool-action risk as allow/review/deny; v1 displays hypothetical verdicts and blocks nothing.Coding & developer toolsMore projects & source code
agent-dispatchergithub.com46Claude Code and Codex dispatcher that routes each request to one of 27 specialist roles with matching skills, tools and a bounded context packet; an opt-in Jev engine can select roles, skills and tools from the catalog.Coding & developer toolsMore projects & source code
agent-gate-loopgithub.com0Reusable GitHub Action that turns an issue into a pull request: an AI agent writes the change, then guards, your checks, a read-only Claude reviewer and Jev gate it, with retries up to max_rounds.Coding & developer toolsMore projects & source code
agent-gate-loop — READMEgithub.com0A prototype GitHub Action loops issue fixes through checks, a reviewer, and Jev gates; its README documents security gaps before unattended use.Coding & developer toolsMore projects & source code
agmsg jev toolgithub.com—External tool driver in agmsg, a cross-vendor messaging layer for CLI coding agents, that lets a seat ask Jev typed questions about a state as a decision aid.Coding & developer toolsMore projects & source code
ai-config-kit typed decisionsgithub.com—CLI in a tool-neutral operating framework for AI coding agents that lets an agent ask Jev yes/no, pick-one or level questions about a diff or issue, then tier and log the answers in a decision ledger.Coding & developer toolsMore projects & source code
ai-config-kit typed decisions — repogithub.com—CLI in a tool-neutral operating framework for AI coding agents that lets an agent ask Jev yes/no, pick-one or level questions about a diff or issue, then tier and log the answers in a decision ledger.Coding & developer toolsMore projects & source code
ai-launchergithub.com0Interactive launcher for AI coding CLIs (Claude, Codex, Gemini)Coding & developer toolsMore projects & source code
ai-launcher — jev-server.pygithub.com0The launcher repository includes a separate TypeSafe Jev server that asks which AI CLI suits a task.Coding & developer toolsMore projects & source code
aidevopsgithub.com401OpenCode plugin and AI DevOps framework that ships opt-in Jev typed decisions: marketing decision batches, retrieval and shadow-triage pilots, and private evaluation reports.Coding & developer toolsMore projects & source code
aidevops — docsgithub.com401OpenCode plugin and AI DevOps framework that ships opt-in Jev typed decisions: marketing decision batches, retrieval and shadow-triage pilots, and private evaluation reports.Coding & developer toolsMore projects & source code
AIOS judgment gategithub.com—Local-first control plane for long-horizon coding agents with an opt-in System One judgment gate, exposed as a CLI and MCP tool, that stays off until explicitly enabled and fails closed.Coding & developer toolsMore projects & source code
AIOS judgment gate — bloggithub.com—Local-first control plane for long-horizon coding agents with an opt-in System One judgment gate, exposed as a CLI and MCP tool, that stays off until explicitly enabled and fails closed.Coding & developer toolsMore projects & source code
Albatrossgithub.com230Terminal coding agent that routes bounded questions through Jev before the main LLM, returning fixed direct answers only when confident, with off, shadow and active modes.Coding & developer toolsMore projects & source code
Albatross — docsgithub.com230Terminal coding agent that routes bounded questions through Jev before the main LLM, returning fixed direct answers only when confident, with off, shadow and active modes.Coding & developer toolsMore projects & source code
anti-dead-internetgithub.com0A Chrome Manifest extension that classifies and removes the text of posts on X/Twitter with TypeSafe's JEVCoding & developer toolsMore projects & source code
anti-dead-internet — READMEgithub.com0Chrome extension uses Jev to estimate AI-generated wording in X/Twitter posts and displays probabilities or hides posts at a configurable threshold.Coding & developer toolsMore projects & source code
Antigravity-mcp-semantic-search-with-TypeSafeAi — readmegithub.com1An MCP code-search and diff-review tool uses TypeSafe System One models to rank files and inspect changes semantically.Coding & developer toolsMore projects & source code
any-autogithub.com7Auto-approve daemon for the Pi and Antigravity coding agents that lets local rules pass read-only tools and block dangerous commands, then sends the rest to a reviewer backend such as Jev, Pi, agy or an API model.Coding & developer toolsMore projects & source code
Apparitegithub.com3Design-mock tool that turns a text prompt into a clickable app-like mock styled by your DESIGN.md, where Jev chooses the information architecture and anatomy, Gemini writes the copy and code assembles A2UI-style screens.Coding & developer toolsMore projects & source code
appstore-precheck Jev semantic reviewgithub.com—Optional, advisory Jev pass in appstore-precheck, an agent skill and Bash scanner for iOS App Store rejection risks, that reviews a locally collected bundle of source, copy and policy evidence in shadow mode.Coding & developer toolsMore projects & source code
appstore-precheck Jev semantic review — repogithub.com—Optional, advisory Jev pass in appstore-precheck, an agent skill and Bash scanner for iOS App Store rejection risks, that reviews a locally collected bundle of source, copy and policy evidence in shadow mode.Coding & developer toolsMore projects & source code
ask-jevgithub.com1Standard-library Python CLI that gives coding agents and pipelines fail-open advisory decisions and verbatim passage selection, while the calling agent keeps reasoning and action authority.Coding & developer toolsMore projects & source code
askgrep — READMEgithub.com4askgrep uses Jev for natural-language code-search questions and returns probabilities instead of prose.Coding & developer toolsMore projects & source code
Atomic Jev decision backendgithub.com—Optional Jev backend in the Atomic coding-agent runtime for schema-validated structured decisions, used by default to pick workflows and subagent models when TypeSafe credentials exist, with chat-model fallback.Coding & developer toolsMore projects & source code
Atomic Jev decision backend — repogithub.com—Optional Jev backend in the Atomic coding-agent runtime for schema-validated structured decisions, used by default to pick workflows and subagent models when TypeSafe credentials exist, with chat-model fallback.Coding & developer toolsMore projects & source code
Attocode Intelligence Jev scorergithub.com—Confidence scorer in Attocode Intelligence, a codebase-understanding MCP server, that replaces hardcoded rule confidences with Jev probabilities: ECE 0.07 vs 0.21 for the constants on its rule-accuracy eval.Coding & developer toolsMore projects & source code
Attocode Intelligence Jev scorer — repogithub.com—Confidence scorer in Attocode Intelligence, a codebase-understanding MCP server, that replaces hardcoded rule confidences with Jev probabilities: ECE 0.07 vs 0.21 for the constants on its rule-accuracy eval.Coding & developer toolsMore projects & source code
autoloop Jev routinggithub.com—Loop harness for long-running agent work, spun off from ralph-orchestrator, with opt-in Jev workflow selection that routes a request to a preset or no_match above a confidence floor.Coding & developer toolsMore projects & source code
autoloop Jev routing — docsgithub.com—Loop harness for long-running agent work, spun off from ralph-orchestrator, with opt-in Jev workflow selection that routes a request to a preset or no_match above a confidence floor.Coding & developer toolsMore projects & source code
backstage-jev-operations-supportgithub.com0Jev-powered operations decision support for Backstage: readiness checks, incident triage, change review, template and team recommendations, and catalog reranking.Coding & developer toolsMore projects & source code
backstage-jev-operations-support — READMEgithub.com0The Backstage plugin presents TypeSafe Jev-powered operational decision workflows; its screenshots use synthetic data and fixed demo results.Coding & developer toolsMore projects & source code
barrunto — READMEgithub.com2A browser extension asks Jev to label viewed content; its README cautions that these labels are opinions, not verdicts.Coding & developer toolsMore projects & source code
Better TypeScript semanticgithub.com8Go linter for TypeScript projects whose semantic command checks natural-language engineering policies against working-tree changes or a Git range with bounded TypeSafe judgments.Coding & developer toolsMore projects & source code
Better TypeScript semantic — codegithub.com8Go linter for TypeScript projects whose semantic command checks natural-language engineering policies against working-tree changes or a Git range with bounded TypeSafe judgments.Coding & developer toolsMore projects & source code
better-cheaper-llmgithub.com0Use Jev instead of an LLM for RAG decisions: relevance grading, query routing, hallucination checks and LLM-as-a-judge evals. Faster, cheaper, measured first.Coding & developer toolsMore projects & source code
better-cheaper-llm — READMEgithub.com0Audits existing LLM logs for decision calls, replays them with Jev and offers shadow/enforce checks with LLM fallback; migration is preceded by traffic-specific measurement.Coding & developer toolsMore projects & source code
blink — READMEgithub.com92Blink searches codebases with a filesystem-walker ensemble whose routing is guided by Jev scores on file and folder names.Coding & developer toolsMore projects & source code
blockly-jevgithub.com0Blockly extension: TypeSafe Jev judgments (Noul, Choice, Score) as first-class blocksCoding & developer toolsMore projects & source code
blockly-jev — READMEgithub.com0Blockly blocks expose Jev judgments as Boolean, Number, or String values.Coding & developer toolsMore projects & source code
bounce-routergithub.com8TUI over Claude Code, Codex and Muse that fails over between providers when usage limits hit, with an optional Jev layer for review verdicts and for routing work to worker profiles by tier.Coding & developer toolsMore projects & source code
bounce-router — codegithub.com8TUI over Claude Code, Codex and Muse that fails over between providers when usage limits hit, with an optional Jev layer for review verdicts and for routing work to worker profiles by tier.Coding & developer toolsMore projects & source code
brogithub.com7Launcher that runs Claude Code, omp, Pi, Codex or DeepSeek Harness against any model provider, with a switch that puts Jev Router in front to choose Haiku or Opus per turn.Coding & developer toolsMore projects & source code
bro — routergithub.com513Launcher that runs Claude Code, omp, Pi, Codex or DeepSeek Harness against any model provider, with a switch that puts Jev Router in front to choose Haiku or Opus per turn.Coding & developer toolsMore projects & source code
Buggogithub.com29🐛 Give your coding agent a detective. Buggo investigates bugs, builds a suspect list, and finds where to look.Coding & developer toolsMore projects & source code
Buggo — READMEgithub.com29Buggo ranks repository files as bug-investigation suspects using Jev through OpenRouter; its README cautions that rankings are hypotheses, not proof.Coding & developer toolsMore projects & source code
Cairngithub.com6Git-based record keeping that ties AI-assisted development to agreed requirements; at consequential choices the coding agent can have Jev score its own draft on five dimensions before deciding.Coding & developer toolsMore projects & source code
Calliope CLIgithub.com8Multi-backend terminal coding agent with a judgments feature that asks typed questions and returns probabilities and confidence to branch on, from any backend or the native TypeSafe engine.Coding & developer toolsMore projects & source code
Calliope CLI — docsgithub.com8Multi-backend terminal coding agent with a judgments feature that asks typed questions and returns probabilities and confidence to branch on, from any backend or the native TypeSafe engine.Coding & developer toolsMore projects & source code
Canny — READMEgithub.com106Canny adds Jev-based judgments to a deterministic coding-agent gate; the README says missing keys or uncertain answers leave the deterministic rule in control.Coding & developer toolsMore projects & source code
cc-settings Jev hooksgithub.com—Team Claude Code and Codex config whose hooks use Jev: a delegation detector for multi-file prompts, a drift fuse for unattended turns and verbatim compaction. On 1,102 real prompts Jev at 0.7 made 53 correct fires vs 10 for regex.Coding & developer toolsMore projects & source code
cc-settings Jev hooks — repogithub.com—Team Claude Code and Codex config whose hooks use Jev: a delegation detector for multi-file prompts, a drift fuse for unattended turns and verbatim compaction. On 1,102 real prompts Jev at 0.7 made 53 correct fires vs 10 for regex.Coding & developer toolsMore projects & source code
ce-ai Jev decisionsgithub.com—Jev decision provider in ce-ai, a Rust CLI that orchestrates the Compound Engineering plugin across ten coding-agent harnesses, answering questions such as model class and task risk at workflow stages.Coding & developer toolsMore projects & source code
ce-ai Jev decisions — repogithub.com—Jev decision provider in ce-ai, a Rust CLI that orchestrates the Compound Engineering plugin across ten coding-agent harnesses, answering questions such as model class and task risk at workflow stages.Coding & developer toolsMore projects & source code
Chat Seekgithub.com64VS Code extension that finds past Claude Code, Codex and OpenCode conversations from a plain-language description, reranking local matches with the open Laya decision model.Coding & developer toolsMore projects & source code
check-riskgithub.com0CLI and GitHub Action that turns a Git change into an explainable risk score, required checks and reviewer groups, combining deterministic path, dependency and diff rules with Jev semantic signals.Coding & developer toolsMore projects & source code
check-risk — READMEgithub.com0A code-change risk CLI combines deterministic repository rules with TypeSafe Jev semantic signals; without an API key, it provides an explicitly incomplete offline assessment.Coding & developer toolsMore projects & source code
CheshiAI/Cheshigithub.com18Jev-powered conversation memory: find past sessions and revisit decisions with original sources. A macOS workspace for OpenAI Codex. Manage AI conversations and agents, explore code with CodeGraph, and work with Git, Ghostty terminals, and Apple Notes in one app.Coding & developer toolsMore projects & source code
choxos/jev-reviewergithub.com37Data extraction for systematic reviews, quoted from the papers. Ask a trial report and its supplements your extraction form or a RoB 2, ROBINS-I, QUADAS-2 or TIDieR template; Jev points at the lines, every answer is a verbatim quote with its page, you check it and export the table. Files stay in your browser.Coding & developer toolsMore projects & source code
chrome-x-post-qualifiergithub.com0Chrome extension that scores X (Twitter) posts in real time with TypeSafe AI — collapses engagement bait and spam, blocks ads, tracks API cost per post.Coding & developer toolsMore projects & source code
chrome-x-post-qualifier — READMEgithub.com0A Chrome extension describes real-time X-post scoring through TypeSafe System One; the README identifies its model as jev-latest.Coding & developer toolsMore projects & source code
chunk-standalone — READMEgithub.com0RAG chunk auditor asks Jev about standalone comprehensibility and neighboring context, suggesting keep, merge or split while sending uncertain cases for review.Coding & developer toolsMore projects & source code
claude-code-tracegithub.com370Claude Code session log viewer for desktop, web and TUI that uses Jev to score agent progress, tool use, focus, exploration, recovery and token efficiency.Coding & developer toolsMore projects & source code
claude-jev — READMEgithub.com18A Claude Code plugin uses Jev at prompt, subagent, edit and turn hooks for routing and rule checks; its compaction hook is explicitly experimental.Coding & developer toolsMore projects & source code
claude-jev — READMEgithub.com7claude-jev turns code questions into typed TypeSafe Jev queries and returns calibrated probabilities.Coding & developer toolsMore projects & source code
claude-jev-funnelgithub.com2Claude Code plugin + zero-dependency CLI for TypeSafe's Jev: judge items in bulk with calibrated yes/no, pick-one and rubric answers; handle the confident ends in code, review only the uncertain band.Coding & developer toolsMore projects & source code
claude-jev-funnel — READMEgithub.com2A Claude Code batch-decision plugin using TypeSafe Jev for confident cases and handing uncertain cases to Claude or a person; deterministic code applies confidence policy.Coding & developer toolsMore projects & source code
clue — READMEgithub.com1Clue ranks CLI and SQLite records semantically using hosted Jev or local providers; its README requires explicit content-sharing permission for hosted Jev calls.Coding & developer toolsMore projects & source code
code-lens-aIgithub.com182Code-intelligence MCP server for AI agents built on Tree-sitter call graphs, with an opt-in --layers mode that tags every file with an architectural layer via Jev.Coding & developer toolsMore projects & source code
Codex-Jev — READMEgithub.com1A Codex tool-output filter using Jev judgments to shorten repetitive logs while retaining private copies of full originals. Eligible output is sent to TypeSafe only when filters and a key are active.Coding & developer toolsMore projects & source code
codex-jev-compaction — READMEgithub.com4A Codex context-handoff tool uses Jev while preserving original text, surfacing decisions, and documenting conservative fallbacks.Coding & developer toolsMore projects & source code
codex-jev-compaction — READMEgithub.com0Codex compaction plugin uses Jev to select completed tool evidence for a bounded external checkpoint and restore it after compaction, with a fail-open fallback and no transcript rewriting.Coding & developer toolsMore projects & source code
codex-jev-preflightgithub.com—CLI & Pipelines: Fail-open Codex UserPromptSubmit hook that injects TypeSafe Jev pre-task routing metadata.Coding & developer toolsMore projects & source code
CodexBar Plasma todo gategithub.com—Repository check for the CodexBar KDE Plasma widget that asks Jev, via OpenRouter Decisions, whether a change completes an entry in TODO.md, so finished work is removed from the list in the same change.Coding & developer toolsMore projects & source code
CodexBar Plasma todo gate — repogithub.com—Repository check for the CodexBar KDE Plasma widget that asks Jev, via OpenRouter Decisions, whether a change completes an entry in TODO.md, so finished work is removed from the list in the same change.Coding & developer toolsMore projects & source code
commentcopgithub.com1CLI that reads JS/TS comments with their surrounding code and scores accuracy and usefulness out of 100, listing the weakest comments first; 159 comments were checked for about $0.008.Coding & developer toolsMore projects & source code
commentcop — READMEgithub.com1commentcop is a Jev-powered CLI that scores JavaScript/TypeScript comments for accuracy and usefulness.Coding & developer toolsMore projects & source code
commit-changelog — READMEgithub.com0Changelog CLI uses Jev to categorize commits since the last tag, preserves each original first line and routes uncertain classifications to a review list.Coding & developer toolsMore projects & source code
commit-minergithub.com36Rust CLI that classifies Git commit messages and diffs into bug fixes, security fixes with CWE IDs, and change types, and writes HTML or CSV reports.Coding & developer toolsMore projects & source code
commit-miner — READMEgithub.com36A Rust CLI classifies commit diffs and messages with Jev and includes an agent skill for invoking those scans.Coding & developer toolsMore projects & source code
compact-adviser — READMEgithub.com191Compact Adviser asks Jev whether the current session is near a boundary that is safe to compact; the host applies the configured hint or compaction policy.Coding & developer toolsMore projects & source code
compact-jev — READMEgithub.com0Manual Claude Code compaction plugin asks Jev which old tool calls/results still matter, retaining user/assistant messages verbatim while removing stale calls and shortening results; a library exposes the same logic.Coding & developer toolsMore projects & source code
compactio — READMEgithub.com3Uses Jev after tool calls to choose how much output an agent needs; deterministic code applies the choice rather than Jev rewriting text.Coding & developer toolsMore projects & source code
comparaigithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
comparai — READMEgithub.com0Compares Claude Code skill instructions in a working tree with main using Jev via Vercel AI Gateway; skills are read, not executed, and presentation order is reversed to check bias.Coding & developer toolsMore projects & source code
Conflux judge commandgithub.com—Judge-command boundary in Conflux, a spec-driven orchestrator running AI coding agents in parallel worktrees, that asks a jev-cli-compatible command whether changes depend on each other as a non-authoritative observer.Coding & developer toolsMore projects & source code
Conflux judge command — repogithub.com—Judge-command boundary in Conflux, a spec-driven orchestrator running AI coding agents in parallel worktrees, that asks a jev-cli-compatible command whether changes depend on each other as a non-authoritative observer.Coding & developer toolsMore projects & source code
Convoygithub.com6Semantic end-to-end agent testing for iOS, Android, and web.Coding & developer toolsMore projects & source code
Convoy — READMEgithub.com6Convoy asks Jev which control matches intent, then performs a tap or type on a passing result.Coding & developer toolsMore projects & source code
CyberCode Fast Judgmentgithub.com—CyberCode, an agent combining Claude Code with Hermes self-evolution, adds a Fast Judgment service backed by Jev for quick classification and browser decisions.Coding & developer toolsMore projects & source code
dbt-assaygithub.com0Your dbt project has types nobody declared. assay infers them and finds where they contradict each other. sqlglot for structure, Jev for meaning.Coding & developer toolsMore projects & source code
dbt-assay — READMEgithub.com0dbt-assay combines SQL parsing with Jev judgments about semantic questions that a parser cannot settle in dbt models.Coding & developer toolsMore projects & source code
Decapodgithub.com233Repo-native governance kernel for AI coding agents with an optional Jev provider that records whether an agent's trajectory satisfies the stated intent as an observation, never as an approval.Coding & developer toolsMore projects & source code
deciogithub.com2Turn context into typed decisions and pre-declared actions.Coding & developer toolsMore projects & source code
decio — READMEgithub.com2Decio turns context into typed decisions and may dispatch only pre-declared actions; its score template identifies Jev as a provider.Coding & developer toolsMore projects & source code
decio — score.yamlgithub.com2Decio turns context into typed decisions and may dispatch only pre-declared actions; its score template identifies Jev as a provider.Coding & developer toolsMore projects & source code
delegategithub.com—Claude Code plugin that decides who runs a task: this session, a sub-agent, or an outside CLI harness. It detects installed harnesses, keeps a user-controlled allowlist and asks Jev which entry fits, then prints the command.Coding & developer toolsMore projects & source code
delegate — repogithub.com—Claude Code plugin that decides who runs a task: this session, a sub-agent, or an outside CLI harness. It detects installed harnesses, keeps a user-controlled allowlist and asks Jev which entry fits, then prints the command.Coding & developer toolsMore projects & source code
deletion-test skillgithub.com—Claude Code skill that tests whether a module can be regenerated from its spec, with optional Jev triage of surviving mutants after behavioral-diff scoring.Coding & developer toolsMore projects & source code
deletion-test skill — docsgithub.com—Claude Code skill that tests whether a module can be regenerated from its spec, with optional Jev triage of surviving mutants after behavioral-diff scoring.Coding & developer toolsMore projects & source code
Dev Command Center decision providergithub.com—Local-first desktop workbench for coding agents with an opt-in decision provider that asks Jev to judge which retrieved ai-memory candidates are relevant.Coding & developer toolsMore projects & source code
Dev Command Center decision provider — repogithub.com—Local-first desktop workbench for coding agents with an opt-in decision provider that asks Jev to judge which retrieved ai-memory candidates are relevant.Coding & developer toolsMore projects & source code
dev-tab-autopilotgithub.com0Chrome extension: sort tabs into groups with TypeSafe JevCoding & developer toolsMore projects & source code
dev-tab-autopilot — READMEgithub.com0The Chrome extension groups tabs by asking TypeSafe Jev to choose a parent from candidates proposed from open tabs and existing group names.Coding & developer toolsMore projects & source code
DevOpsWorker port classifiergithub.com—Multi-agent Azure DevOps pipeline that asks Jev whether a pull request is a port of another recent PR, catching cases title regexes miss; over 100 PRs those misses cost 9 ports and $29.30 of full reviews.Coding & developer toolsMore projects & source code
DevOpsWorker port classifier — repogithub.com—Multi-agent Azure DevOps pipeline that asks Jev whether a pull request is a port of another recent PR, catching cases title regexes miss; over 100 PRs those misses cost 9 ports and $29.30 of full reviews.Coding & developer toolsMore projects & source code
diffguard — READMEgithub.com1A pull-request risk-triage experiment combines Jev's typed signals with deterministic review policy; its public demo is fixture-backed, separate from the live API path.Coding & developer toolsMore projects & source code
diffjury — READMEgithub.com7A web app fetches public GitHub PR details and asks TypeSafe Jev for risk judgments; a GitHub token is optional for higher API limits.Coding & developer toolsMore projects & source code
Distill Jev supportgithub.com—Jev module in Distill, a token-saving coding-agent harness and TUI for Grok, Codex and OpenAI-compatible models, that selects model tier and effort, routes bounded utility tasks and judges what context to keep.Coding & developer toolsMore projects & source code
Distill Jev support — repogithub.com—Jev module in Distill, a token-saving coding-agent harness and TUI for Grok, Codex and OpenAI-compatible models, that selects model tier and effort, routes bounded utility tasks and judges what context to keep.Coding & developer toolsMore projects & source code
docfrictiongithub.com1Automated friction logs for online documentation, scored section by section with TypeSafe JevCoding & developer toolsMore projects & source code
docfriction — READMEgithub.com1docfriction walks documentation by section and asks TypeSafe Jev fixed typed questions to produce friction logs; deterministic code handles link and formatting checks.Coding & developer toolsMore projects & source code
drift-guardgithub.com0Use Jev (TypeSafe AI) with Claude Code and Codex: hooks that warn when you or your agent drift off taskCoding & developer toolsMore projects & source code
drift-guard — READMEgithub.com0Claude Code/Codex hook pair uses Jev to compare user and agent turns against session intent and flag conversational drift.Coding & developer toolsMore projects & source code
DSCODE Jev approval plugingithub.com2Terminal DeepSeek coding agent plugin that answers automatic permission reviews with Jev over OpenRouter's Decisions API instead of a chat model.Coding & developer toolsMore projects & source code
dsh-auto-review-jevgithub.com1DeepSeek Harness plugin: per-tool-call Auto-permission review powered by TypeSafe Jev, with account usage and API-key management inline on its settings pageCoding & developer toolsMore projects & source code
dsh-auto-review-jev — READMEgithub.com1When a DeepSeek Harness session uses Auto, supported native and PTC tool calls pass through Jev before their bodies execute.Coding & developer toolsMore projects & source code
dsh-jev-advisorgithub.com0Jev (TypeSafe System One) gives the human a second opinion, not the agent. When DSH's model asks you a multiple-choice question, dsh-jev builds a structured Jev request and floats the recommendation, probability spread and confidence beside the options - one click to adopt. API key lives in Settings. No prompt changes.Coding & developer toolsMore projects & source code
dsh-jev-advisor — READMEgithub.com0DeepSeek Harness plugin turns questions addressed to the human into Jev requests and displays recommendations and probabilities beside the choices.Coding & developer toolsMore projects & source code
dsh-jev-autogithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
dsh-jev-auto — READMEgithub.com0This dsh integration auto-approves a command only when Jev selects allow above the confidence threshold; otherwise it leaves approval to dsh’s native human-review path.Coding & developer toolsMore projects & source code
dsh-jev-prune — READMEgithub.com4Adds a pluggable Jev judgment backend to DeepSeek Harness two-layer context compaction without replacing its compaction algorithms. Rules and self-hosted models remain separate backend options.Coding & developer toolsMore projects & source code
dsh-jev-tools — READMEgithub.com11A DeepSeek Harness plugin inserts Jev judgments before tool results enter context; README describes output trimming, injection screening, skill suggestions, and delivery gates, with fail-open behavior.Coding & developer toolsMore projects & source code
Dub malicious link checkgithub.com≈24,800Screens every new short link on the Dub link platform with Jev after a domain blacklist check, blocking phishing, malware, cloaking redirectors, gambling and adult destinations.Coding & developer toolsMore projects & source code
duckdb-semantic-profilegithub.com0what is this?Coding & developer toolsMore projects & source code
duckdb-semantic-profile — READMEgithub.com0DuckDB extension uses Jev to profile table-value meanings and select semantic probes for suspicious values and cross-column inconsistencies beyond statistical summaries.Coding & developer toolsMore projects & source code
dure jevgithub.com—CLI command in Dure, an open-source workspace coordinating Claude Code, Codex and Pi agents across projects, that evaluates a JSON request of typed Noul, Choice and Score questions against TypeSafe Jev from any machine.Coding & developer toolsMore projects & source code
dure jev — repogithub.com—CLI command in Dure, an open-source workspace coordinating Claude Code, Codex and Pi agents across projects, that evaluates a JSON request of typed Noul, Choice and Score questions against TypeSafe Jev from any machine.Coding & developer toolsMore projects & source code
ErisLintgithub.com17Rust linter with a VS Code extension where you define code-quality rules in JSON, such as whether a function is needlessly complicated or a name is misleading, and map Jev's answers to warnings or errors.Coding & developer toolsMore projects & source code
eslint-plugin-jevgithub.com3Today's linters read the shape of your code. This one would read what it means.Coding & developer toolsMore projects & source code
eslint-plugin-jev — READMEgithub.com3eslint-plugin-jev sends functions to Jev for probability judgments, including whether a function name promises a different action from its body.Coding & developer toolsMore projects & source code
Esposter Jev repo automationgithub.com—Repository scripts in the Esposter Nuxt monorepo that use Jev to label issue triage and read CodeRabbit finding severities and release gates, with a 0.6 confidence floor and 0.85 for writes outside the checkout.Coding & developer toolsMore projects & source code
Esposter Jev repo automation — repogithub.com—Repository scripts in the Esposter Nuxt monorepo that use Jev to label issue triage and read CodeRabbit finding severities and release gates, with a 0.6 confidence floor and 0.85 for writes outside the checkout.Coding & developer toolsMore projects & source code
everygithub.com7Grep whose pattern is a yes/no question: judges every function in a codebase and ranks them by probability of yes, measured at 1,302 functions of gin-gonic/gin in 3.7 s for $0.018.Coding & developer toolsMore projects & source code
every — READMEgithub.com7A code-search CLI asks Jev yes/no questions about every function and ranks probabilities; scanned source text is sent to TypeSafe for judgment.Coding & developer toolsMore projects & source code
evotgithub.com328Lightweight terminal coding agent whose Jev prune step deletes stale context instead of summarising it, asking of every old tool call whether the task still depends on it.Coding & developer toolsMore projects & source code
expect-semanticgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
expect-semantic — READMEgithub.com0A Jest/Vitest assertion library lets tests express semantic expectations that are judged with Jev rather than ordinary exact-string matching.Coding & developer toolsMore projects & source code
factlabelgithub.com1A nutrition label for AI-written content: audits what AI agents say about data, blocks what doesn't hold up, and shows readers why. Powered by Jev.Coding & developer toolsMore projects & source code
falcongithub.com0🦅 Native macOS Jev proxy and decision observability, in developmentCoding & developer toolsMore projects & source code
falcon — READMEgithub.com0Native macOS HTTP/MCP proxy and decision-observation app for Jev, with per-source local keys and inspection of recorded request states and outcomes.Coding & developer toolsMore projects & source code
fast-compact — READMEgithub.com5Adds a separate Claude Code compaction command: Jev selects tool outputs to keep intact, while code truncates others and archives full text. It does not replace built-in conversation compaction.Coding & developer toolsMore projects & source code
fast-compaction-dsh — READMEgithub.com3A DeepSeek Harness plugin uses Jev to score tool calls/results for keep, truncate, or drop; on failure it falls back to the basic compactor.Coding & developer toolsMore projects & source code
fast-jev-codex — READMEgithub.com0A Codex plugin scores tool history with Jev, stores selected verbatim context, and restores it after compaction; README says tool-result bodies are omitted from the request.Coding & developer toolsMore projects & source code
fast-jev-compaction — READMEgithub.com0The compaction plugin uses Jev to remove unneeded tool calls and results; user and assistant text remains verbatim and in order.Coding & developer toolsMore projects & source code
fast-jev-compaction — READMEgithub.com7,282A Claude Code plugin asks Jev to score tool calls and results for its compaction summary; README describes its design, not independently measured savings.Coding & developer toolsMore projects & source code
fast-jev-compaction-codex — READMEgithub.com1The Codex port uses Jev to remove unneeded tool calls/results during compaction while keeping user and assistant text verbatim.Coding & developer toolsMore projects & source code
Feather Wandgithub.com60AI agent plugin for Apache JMeter whose optional Jev Smart Routing classifies each request's intent to give Agent Mode a focused tool pack, with a visible route card and full-tool fallback.Coding & developer toolsMore projects & source code
five-lines — READMEgithub.com2The Five Lines review tool asks Jev only for pull-request judgments that cannot be counted mechanically.Coding & developer toolsMore projects & source code
Fleet auto-mode Jev classifiergithub.com—Option in Fleet, a desktop terminal multiplexer for running several AI coding agents, to use Jev 1.13 as the auto-mode classifier model that answers yes or no about each agent command instead of a text model.Coding & developer toolsMore projects & source code
Fleet auto-mode Jev classifier — repogithub.com—Option in Fleet, a desktop terminal multiplexer for running several AI coding agents, to use Jev 1.13 as the auto-mode classifier model that answers yes or no about each agent command instead of a text model.Coding & developer toolsMore projects & source code
fleet-core TypeSafe clientgithub.com—Shared TypeSafe client for a set of Claude Code plugins that enforces a data rule in code: state must pass redaction and truncation before any Jev request, and every verdict is logged.Coding & developer toolsMore projects & source code
fleet-core TypeSafe client — docsgithub.com—Shared TypeSafe client for a set of Claude Code plugins that enforces a data rule in code: state must pass redaction and truncation before any Jev request, and every verdict is logged.Coding & developer toolsMore projects & source code
fleet-core TypeSafe client — repogithub.com—Shared TypeSafe client for a set of Claude Code plugins that enforces a data rule in code: state must pass redaction and truncation before any Jev request, and every verdict is logged.Coding & developer toolsMore projects & source code
Foremangithub.com621Supervises Codex and OpenCode workers with ten Nouls, such as whether a worker is stuck or looping.Coding & developer toolsMore projects & source code
GenOfficegithub.com—Desktop office suite with optional Jev reranking for local file searches.Coding & developer toolsMore projects & source code
get-fable Reflex (Fable-Jev)github.com—Reflex layer in get-fable, a skills-and-harness pack for existing coding agents, that uses Jev for semantic task routing with probability margins and for verbatim transcript compaction that drops obsolete tool results.Coding & developer toolsMore projects & source code
get-fable Reflex (Fable-Jev) — repogithub.com—Reflex layer in get-fable, a skills-and-harness pack for existing coding agents, that uses Jev for semantic task routing with probability margins and for verbatim transcript compaction that drops obsolete tool results.Coding & developer toolsMore projects & source code
git-jev-stage — READMEgithub.com2git-jev-stage asks Jev to classify Git diff hunks for a plain-language staging request.Coding & developer toolsMore projects & source code
git-judge-commitsgithub.com1⚖️ Judge git commits with Jev: is it breaking, does it deserve attention, and does its message tell the truth?Coding & developer toolsMore projects & source code
git-judge-commits — READMEgithub.com1git-judge-commits uses Jev to judge commit diffs across typed review questions, including whether the commit message matches the diff.Coding & developer toolsMore projects & source code
git-star-list-sortgithub.com0Sort your GitHub stars into your existing GitHub Lists with TypeSafe Jev. One command: refresh List descriptions, classify, then gated additive apply.Coding & developer toolsMore projects & source code
git-star-list-sort — READMEgithub.com0This CLI uses one Jev choice per repository to sort GitHub stars into existing Lists; classification writes a report, while applying assignments is a separate gated step.Coding & developer toolsMore projects & source code
github-star-organizer-jev — READMEgithub.com1A command-line app uses TypeSafe Jev to classify GitHub stars against existing Lists and can optionally apply the assignments.Coding & developer toolsMore projects & source code
Glowbom OSS Jev toolgithub.com—Optional tool in a Project Book coding-agent workflow that lets the agent ask Jev small decisions, such as whether build and test results look healthy or need another fix, via OpenCode.Coding & developer toolsMore projects & source code
gmail-jev-guardgithub.com0JevGuard — estensione Chrome MV3 che mostra nella lista di Gmail la probabilita' che un messaggio sia spam o phishing, calcolata da TypeSafe JevCoding & developer toolsMore projects & source code
gmail-jev-guard — READMEgithub.com0A Gmail-list Chrome extension sends the sender, subject, and preview visible in each row to Jev for spam/phishing triage; its README says it does not read message bodies.Coding & developer toolsMore projects & source code
Goddard Jev evaluation featuresgithub.com—Evaluation-model client in Goddard, a Rust GUI for orchestrating Codex, Claude, OpenCode and other coding agents, that uses Jev for per-turn status markers and experimental next-action prediction chips.Coding & developer toolsMore projects & source code
Goddard Jev evaluation features — repogithub.com—Evaluation-model client in Goddard, a Rust GUI for orchestrating Codex, Claude, OpenCode and other coding agents, that uses Jev for per-turn status markers and experimental next-action prediction chips.Coding & developer toolsMore projects & source code
Graphlingithub.com—Live architecture and activity diagrams in the browser while Claude Code or Codex explores and edits your code, with optional Jev classification of agent activity.Coding & developer toolsMore projects & source code
grevgithub.com54Thinking coreutilsCoding & developer toolsMore projects & source code
grev — READMEgithub.com54Unix-style text filters use TypeSafe Jev’s typed questions to select and organize input records by meaning.Coding & developer toolsMore projects & source code
grill-me-with-jevgithub.com—Agent skill that stress-tests an engineering plan through a decision-tree interview, using Jev to advise whether each concern should be asked now, investigated first, or skipped.Coding & developer toolsMore projects & source code
Harness MCP failure triagegithub.com—Official Harness.io MCP server whose harness_diagnose tool adds advisory failure-category triage for pipeline failures using a single TypeSafe Choice.Coding & developer toolsMore projects & source code
hekajev — READMEgithub.com6HekaJev answers Git-history questions by filtering and classifying commits, and describes itself as powered by Jev.Coding & developer toolsMore projects & source code
herdr-jevgithub.com0Herdr plugin: know why each coding agent stopped. Triage blocked and finished agents with TypeSafe's Jev model: risky approvals, questions, failed runs, red checks.Coding & developer toolsMore projects & source code
herdr-jev — READMEgithub.com0Herdr plugin uses Jev to inspect terminal tails when coding agents stop or block and show the reason in the sidebar.Coding & developer toolsMore projects & source code
herdr-mcpgithub.com9MCP/OAuth edge that lets ChatGPT and other web planners run code, Git and tests on local machines through Herdr, using Jev routes first for post-turn semantic judgments before LLM and script fallbacks.Coding & developer toolsMore projects & source code
here-we-go-jevgithub.com1Local playground and test bench for TypeSafe's Jev System One model: typed questions, calibrated-probability answers, and side-by-side comparison with an LLM baseline.Coding & developer toolsMore projects & source code
hey-jevgithub.com49TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
Hunch — linkgithub.com—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.Coding & developer toolsMore projects & source code
hunk-triage — READMEgithub.com2The Hunk extension uses Jev to classify changed files and order them for review.Coding & developer toolsMore projects & source code
Hunkpickgithub.com1Resolve git merge conflicts by enumeration and judgment: code enumerates every valid resolution, a TypeSafe System One model picks, code gates the pick.Coding & developer toolsMore projects & source code
Hunkpick — READMEgithub.com1Hunkpick generates valid merge-resolution candidates and asks Jev to select among them; the tool describes selection, not code generation, as its role.Coding & developer toolsMore projects & source code
huntedman/JevLintgithub.com11Configurable semantic linting powered by Jev, with file-level NOUL judgments and a magic-strings plugin.Coding & developer toolsMore projects & source code
IDA Pro MCP Jev intelligencegithub.com—MCP server for IDA Pro reverse engineering with an opt-in Jev provider that scores candidate pools for semantic function search and reorders the next analysis targets.Coding & developer toolsMore projects & source code
IDA Pro MCP Jev intelligence — docsgithub.com—MCP server for IDA Pro reverse engineering with an opt-in Jev provider that scores candidate pools for semantic function search and reorders the next analysis targets.Coding & developer toolsMore projects & source code
IDA Pro MCP Jev intelligence — repogithub.com—MCP server for IDA Pro reverse engineering with an opt-in Jev provider that scores candidate pools for semantic function search and reorders the next analysis targets.Coding & developer toolsMore projects & source code
if-ai — READMEgithub.com1A GitHub Action checks pull requests against plain-English rules with Jev and returns a typed verdict; it does not use a chat-completion model.Coding & developer toolsMore projects & source code
Inbox Zero TypeSafe Decision Providergithub.com12,265Email assistant with an optional TypeSafe decision provider.Coding & developer toolsMore projects & source code
interlinked-cli Jev checksgithub.com—Opt-in Jev checks in interlinked, a local guard layer for coding agents, that flag tests whose body does not test their title and docs that claim a module is live when nothing imports it.Coding & developer toolsMore projects & source code
interlinked-cli Jev checks — repogithub.com—Opt-in Jev checks in interlinked, a local guard layer for coding agents, that flag tests whose body does not test their title and docs that claim a module is live when nothing imports it.Coding & developer toolsMore projects & source code
IssueCrushgithub.com17Tinder-style app for swiping through your GitHub issues to close or keep them; optional structured triage asks Jev for a recommended action, staleness and effort scores and an actionable probability for each card.Coding & developer toolsMore projects & source code
japanese-jev-lintgithub.com1Go CLI (jjl) that asks Jev per sentence for the probability of typos, twisted subject-predicate structure, overlong sentences or repetition in Japanese text, with reviewdog and textlint output formats.Coding & developer toolsMore projects & source code
japanese-jev-lint — READMEgithub.com1A Japanese-text linter asks Jev to score sentence-level issues; it flags threshold-crossing sentences rather than generating corrections.Coding & developer toolsMore projects & source code
Jarvisgithub.com137Collaborative AI development platform whose config can assign Jev as a structured evaluation model for choosing among methodology and rule candidates.Coding & developer toolsMore projects & source code
jastgithub.com1JAST is an experimental SAST (static application security testing) desktop app that uses TypeSafe AI's Jev System One model.Coding & developer toolsMore projects & source code
jast — READMEgithub.com1JAST is an experimental desktop SAST app that asks Jev explicit questions about bounded source regions; uploading a snapshot requires explicit user authorization.Coding & developer toolsMore projects & source code
jcode Jev memory and browser handoffgithub.com—The jcode coding-agent harness uses Jev for memory recall relevance and for a fast browser handoff, where Jev picks each offered action and hands back to the main LLM for code or text.Coding & developer toolsMore projects & source code
jcolgithub.com1Apply natural-language codebooks to tables: a CLI and Python API with resumable annotation, exports, and label evaluation.Coding & developer toolsMore projects & source code
jcol — READMEgithub.com1A table-coding CLI/API turns natural-language codebook definitions into Jev judgment columns, with checkpoint/resume support; local fake-API tests do not validate model accuracy.Coding & developer toolsMore projects & source code
jeffgithub.com38Read-only Go CLI that checks files against your own rules, such as hidden side effects or weak error handling, with Jev judging each file semantically, locally or in CI.Coding & developer toolsMore projects & source code
jeq — READMEgithub.com10A composable terminal tool lets scripts and agents ask Jev typed questions, then locally apply policy gates to the returned answers.Coding & developer toolsMore projects & source code
jevgithub.com—Claude Code plugin where you ask in plain language, Claude turns it into typed Jev questions via OpenRouter, and you get a typed answer with a probability, with low-confidence results labeled.Coding & developer toolsMore projects & source code
jev (Claude Code plugin)github.com13Go Claude Code plugin that takes over searches and large file reads, using Jev to find relevant files and answer bounded questions about code so whole files never enter the agent's context.Coding & developer toolsMore projects & source code
Jev as Policygithub.com47MuJoCo studio that reproduces the Jev as Policy control structure: Jev turns a text state into an intent and motor choices for a Franka Panda that local IK executes to carry a block over a barrier.Coding & developer toolsMore projects & source code
Jev Chat Assistantgithub.com—Android chat assistant that uses Jev to classify conversations and rank candidate replies before manual insertion.Coding & developer toolsMore projects & source code
JEV Codex Pilotgithub.com2Local control center with a Kanban flow that turns software tickets into Codex runs, choosing model and reasoning effort per ticket and compacting long threads, estimated to cut tokens 20-40% on mixed workloads.Coding & developer toolsMore projects & source code
JEV DataOpsgithub.com37Workbench that screens uploaded training data with Jev, evaluates data quality, trains a LoRA automatically and evaluates the result on held-out data, via a browser app, CLI and Python and HTTP APIs.Coding & developer toolsMore projects & source code
Jev Frontend QAgithub.com5Evidence-driven frontend QA harness where Jev picks browser operations and targets while test code checks authored contracts against DOM, HTTP and SQLite evidence, demonstrated on a synthetic todo app.Coding & developer toolsMore projects & source code
Jev Issue Radargithub.com1Read-only dashboard for GitHub issue triage that retrieves candidate pairs and shows side-by-side evidence for whether two reports are duplicate, related, distinct, or under-documented.Coding & developer toolsMore projects & source code
Jev Linter Actiongithub.com1GitHub Action that reviews repository files against your own yes/no questions, passing only when Jev's probability for every expected answer meets its threshold, with glob and per-file modes.Coding & developer toolsMore projects & source code
jev moderation botgithub.com47A Discord moderation bot built with Python and TypeSafe AI Jev System One.Coding & developer toolsMore projects & source code
JEV OAS Sentinelgithub.com5GitHub Action and CLI that compares two OpenAPI documents with deterministic structural checks plus semantic review of changed descriptions, defaults, retries, pagination, auth, and error behavior.Coding & developer toolsMore projects & source code
JEV OAS Sentinel — marketplacegithub.com—GitHub Action and CLI that compares two OpenAPI documents with deterministic structural checks plus semantic review of changed descriptions, defaults, retries, pagination, auth, and error behavior.Coding & developer toolsMore projects & source code
Jev PR Labelergithub.com4Labels GitHub pull requests by type, area, platform and size through OpenRouter's Decisions API, where size reflects conceptual scope rather than lines changed.Coding & developer toolsMore projects & source code
Jev Realtime Code Check — projectgithub.com—Short demo of a VS Code and Cursor extension that checks local Git changes against 370 Markdown coding rules with Jev on every save, in under 2 seconds.Coding & developer toolsMore projects & source code
Jev Realtime Code Check — repogithub.com—Short demo of a VS Code and Cursor extension that checks local Git changes against 370 Markdown coding rules with Jev on every save, in under 2 seconds.Coding & developer toolsMore projects & source code
Jev Reviewgithub.com0Autonomous pull-request review bot that feeds PR metadata and patches to Jev as finite typed questions, applies deterministic approval gates and routes uncertain or risky reviews to trusted owners.Coding & developer toolsMore projects & source code
Jev Review Actiongithub.com2GitHub Action that reviews catalog submissions against pinned repository evidence or classifies pull requests with Jev, then updates one templated PR comment, with no text-generation model.Coding & developer toolsMore projects & source code
JEV shared judgment layergithub.com—Shared runner in a coding-agent skill set that turns a target state plus a battery of typed questions into a calibrated assessment, first used for pull-request assessment.Coding & developer toolsMore projects & source code
JEV shared judgment layer — repogithub.com—Shared runner in a coding-agent skill set that turns a target state plus a battery of typed questions into a calibrated assessment, first used for pull-request assessment.Coding & developer toolsMore projects & source code
Jev Skill Suggestergithub.com30Codex skill and standard-library Python CLI that recommends one installed skill for the current task with Jev Choice and Noul checks over skill descriptions and snippets, allowing no skill or unsure as answers.Coding & developer toolsMore projects & source code
Jev skill suggestion for Claude Codegithub.com30,825Hides the skill roster and lets Jev attach at most one skill to each prompt.Coding & developer toolsMore projects & source code
Jev Turn Analysisgithub.com—Rust CLI that reads finished Claude Code and Codex sessions, scores them with Jev for success, verification and wasted effort, and suggests changes to instructions, skills and tools.Coding & developer toolsMore projects & source code
jev — READMEgithub.com0This independent wrapper uses Jev to decide which coding-agent tool calls survive transcript compaction and how much work an agent should spend.Coding & developer toolsMore projects & source code
jev-affected — READMEgithub.com1jev-affected asks Jev whether a Git diff affects configured semantic conditions and turns the judgments into a task plan; the README demo is explicitly an offline fixture.Coding & developer toolsMore projects & source code
jev-agent-toolsgithub.com—CLI & Pipelines: Jev transport/provider layer: multi-provider transport layer that supports fail-closed validation. Used by jkudish/jev-browser and jkudish/jev-mcp.Coding & developer toolsMore projects & source code
jev-ai-slop-detectorgithub.com4Chrome extension that flags AI-slop posts on X, LinkedIn and Reddit in real time using the JEV ClassifierCoding & developer toolsMore projects & source code
jev-ai-slop-detector — READMEgithub.com4This browser extension labels feed posts with the TypeSafe JEV classifier; its screenshots use fictional posts even though Jev results are real.Coding & developer toolsMore projects & source code
jev-aligngithub.com301Finds the examples a Jev function is least sure about, asks you to label them, and improves the function with GEPA.Coding & developer toolsMore projects & source code
jev-align — READMEgithub.com301jev-align is Sutro's experimental CLI for building AI Functions with TypeSafe Jev.Coding & developer toolsMore projects & source code
jev-assertgithub.com—Agent skill that turns a natural-language UI expectation into a verdict by asking Jev whether the outcome holds in a page's captured text state, for use in automated UI verification.Coding & developer toolsMore projects & source code
jev-assert — repogithub.com—Agent skill that turns a natural-language UI expectation into a verdict by asking Jev whether the outcome holds in a page's captured text state, for use in automated UI verification.Coding & developer toolsMore projects & source code
jev-assistgithub.com4Agent skill and CLI that has Jev ask every tracked file the same relevance question for a one-line task, so a coding agent starts from the few files it needs; a validate command grades rankings against past commits.Coding & developer toolsMore projects & source code
jev-assist — READMEgithub.com4jev-assist uses TypeSafe Jev to rank repository files by relevance for coding tasks.Coding & developer toolsMore projects & source code
jev-auditgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-audit — READMEgithub.com0jev-audit uses TypeSafe Jev for an initial repository audit.Coding & developer toolsMore projects & source code
jev-audit — READMEgithub.com1A pre-commit auditor submitting code changes as state to TypeSafe Jev for four atomic typed questions. It gates review rather than writing fixes; safety and detection accuracy remain unverified.Coding & developer toolsMore projects & source code
jev-auto-approvegithub.com7GitHub Action that asks whether a pull request needs a human reviewer and approves it only when every question confidently clears your threshold, naming the question that held back a skip.Coding & developer toolsMore projects & source code
jev-auto-approve — linkgithub.com—GitHub Action that asks whether a pull request needs a human reviewer and approves it only when every question confidently clears your threshold, naming the question that held back a skip.Coding & developer toolsMore projects & source code
jev-axigithub.com25Agent-ergonomic CLI that gives coding agents Jev judgments from the shell, including a PreToolUse hook that blocks risky Bash calls, prompt-injection screening of fetched text, log triage and ranking.Coding & developer toolsMore projects & source code
jev-axi — READMEgithub.com25A CLI installs Jev-backed PreToolUse checks for Claude Code or Codex, providing an additional decision gate for coding-agent actions.Coding & developer toolsMore projects & source code
jev-belaygithub.com19Claude Code Stop hook that, when files changed and no check has passed, asks Jev four questions on whether the closing message is an unverified done; on 100 labeled stops it scored AUROC 0.976.Coding & developer toolsMore projects & source code
jev-belay — READMEgithub.com19A Claude Code Stop hook reads the local transcript and makes one Jev call only when files changed without a passing check.Coding & developer toolsMore projects & source code
jev-blindspotgithub.com—CLI & Pipelines: A side panel for Claude Code and Codex CLI that shows the blind spots of each prompt you submit: what the request would have needed to consider and shows no sign of. Jev decides, in one call per prompt, whether the prompt is worth a second look; only then does the agent's own headless mode read the project and return the items. Nothing blocks the prompt, nothing is added to the session.Coding & developer toolsMore projects & source code
jev-bookmarksgithub.com0Chrome履歴から目的に合うページを選び、Jevで操作して役立ったURLだけをプロジェクトごとに記録するCLICoding & developer toolsMore projects & source code
jev-bookmarks — READMEgithub.com0Preview CLI chooses browser-history entry pages with Jev, operates them and stores only useful URLs in each Git project's bookmark directory.Coding & developer toolsMore projects & source code
jev-browsegithub.com8Fast, cheap browser sub-tasks for Claude and other agents: TypeSafe Jev decisions on top of browser-harnessCoding & developer toolsMore projects & source code
jev-browser-qagithub.com2Browser QA tool where Playwright drives and records while Jev judges natural-language assertions, heals locators, and runs goal-driven click loops, with a JSON-flow CLI for agents and a run dashboard.Coding & developer toolsMore projects & source code
jev-browser-skillgithub.com—CLI & Pipelines: This Codex Skill lets Codex drive Chrome through Jev to complete multi-step browser tasks from a goal description with preset inputs.Coding & developer toolsMore projects & source code
jev-calibrategithub.com31CLI that tunes Jev question criteria on your labeled examples, confirms them on a held-out set, and says whether each question can gate a decision on its own, only rank, or carries no usable signal.Coding & developer toolsMore projects & source code
jev-checkgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-check — READMEgithub.com0Pre-commit CLI asks Jev yes/no questions about staged Git files and summarizes the checks as a pass or fail gate.Coding & developer toolsMore projects & source code
jev-chrome-blocker-extensiongithub.com0A chrome extension that uses JEV to dynamically block content per user requestCoding & developer toolsMore projects & source code
jev-chrome-blocker-extension — READMEgithub.com0The Chrome extension hides matching page blocks and uses Jev/System One yes/no judgments for rules.Coding & developer toolsMore projects & source code
jev-ci-triage — READMEgithub.com0CI failure triage uses rules first and Jev for ambiguous Playwright/JUnit failures, adding labels without rerunning tests or changing red builds to green.Coding & developer toolsMore projects & source code
jev-claude-code-hooksgithub.com0Claude Code hooks: TypeSafe Jev guards destructive Bash commands and grades promptsCoding & developer toolsMore projects & source code
jev-claude-code-hooks — READMEgithub.com0Claude Code hooks use Jev to grade prompts and flag destructive Bash commands for permission review, failing open on missing keys or request errors.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com1jev-cli is a Bun command-line client for sending typed JSON-state questions to the TypeSafe AI API.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com1An unofficial shell CLI for TypeSafe's Jev API provides typed questions, machine-readable answers, and offline policy checks over saved judgments.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com1jev-cli is a typed-decision CLI that defaults to TypeSafe’s direct Jev API and also documents OpenRouter and a self-hosted NewAPI gateway.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com3A CLI wraps Jev's state-and-typed-question interface and returns structured JSON for evaluation or integration workflows.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com0This thin CLI wraps the TypeSafe System One API with typed decisions and JSON input/output; its runtime compatibility was not tested here.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com22The jev CLI wraps TypeSafe Jev typed judgments in commands for fixed-answer questions.Coding & developer toolsMore projects & source code
jev-cli — READMEgithub.com34An unofficial Rust CLI turns Jev yes/no, choice and rubric questions into probabilities usable by shells, CI, scripts and MCP clients.Coding & developer toolsMore projects & source code
jev-codegithub.com20Coding CLI and TypeScript SDK where Jev writes Python or Bash one grammar rule at a time by choosing AST productions, then reads back, fixes, runs and reports the program.Coding & developer toolsMore projects & source code
jev-code — READMEgithub.com20jev-code is a TypeScript SDK in which callers define options and effects while Jev selects an option.Coding & developer toolsMore projects & source code
jev-codesgithub.com5CLI and agent commands that audit a git diff hunk by hunk against editable YAML coding-standards packs, so the agent fixes only flagged hunks; a 500-line diff audits in seconds for a fraction of a cent.Coding & developer toolsMore projects & source code
jev-codes — READMEgithub.com5A CLI and agent commands audit Git diffs against editable standards packs with Jev, scoring hunks; the agent, not Jev, edits flagged code.Coding & developer toolsMore projects & source code
jev-codexgithub.com0Delegation-first Jev routing, deterministic code review, and preventive checkpoints for Codex Desktop.Coding & developer toolsMore projects & source code
jev-codexgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-codex — jev_decide.pygithub.com0Codex decision-helper skill includes a Python HTTP script for Jev typed judgments; its README explicitly says script testing and real API requests had not been performed.Coding & developer toolsMore projects & source code
jev-codex — READMEgithub.com0Codex decision-helper skill includes a Python HTTP script for Jev typed judgments; its README explicitly says script testing and real API requests had not been performed.Coding & developer toolsMore projects & source code
jev-codex — READMEgithub.com0Community Codex Desktop plugin uses Jev for subagent model and reasoning-effort selection, alongside deterministic review policy and task checkpoints.Coding & developer toolsMore projects & source code
Jev-Cognigy-QA-Suitegithub.com0Score every Cognigy conversation against rubrics you write, using TypeSafe Jev instead of an LLM. CLI, local web UI, and an agent skill.Coding & developer toolsMore projects & source code
Jev-Cognigy-QA-Suite — READMEgithub.com0A Cognigy conversation QA tool asks Jev typed rubric questions about transcripts; the README describes Jev as a classifier rather than a text generator.Coding & developer toolsMore projects & source code
jev-commitgithub.com12Pre-commit hook that asks Jev five questions in one call about the commit message and staged diff, such as filler messages, contradictions, debug leftovers and scope creep, at about four cents per thousand commits.Coding & developer toolsMore projects & source code
jev-commit — READMEgithub.com12A commit-msg hook uses one Jev call before commit to check whether the commit message agrees with the code changes.Coding & developer toolsMore projects & source code
jev-compact — READMEgithub.com0Claude Code compactor prunes context with local rules and optionally asks Jev to rank remaining tool outputs for relevance before dropping them.Coding & developer toolsMore projects & source code
jev-compilergithub.com0Compile decision policies into inspectable, measurable TypeSafe Jev programs.Coding & developer toolsMore projects & source code
jev-compiler — READMEgithub.com0Decision-policy compiler combines typed Jev questions with ordinary code into inspectable programs, using a separate teacher for proposals rather than training Jev weights.Coding & developer toolsMore projects & source code
jev-curate — READMEgithub.com93jev-curate is a dataset-filtering pipeline that sends per-row typed rubrics to Jev; the advertised 24 rows/sec is explicitly from a local mock benchmark and is not repeated here.Coding & developer toolsMore projects & source code
jev-debtgategithub.com1Technical-debt gate for coding agents and CI that collects local metrics from a diff or file, asks Jev a fixed set of typed questions, and applies policy to return allow, review or block.Coding & developer toolsMore projects & source code
jev-debtgate — READMEgithub.com1This technical-debt gate asks Jev a fixed set of questions about diffs or files, then applies local policy; it does not rewrite code.Coding & developer toolsMore projects & source code
jev-demogithub.com—CLI & Pipelines: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details.Coding & developer toolsMore projects & source code
jev-design-testgithub.com31Jev shadcn-block generatorCoding & developer toolsMore projects & source code
jev-design-test — READMEgithub.com31A UI-layout sampler asks Jev to route a prompt among a fixed vocabulary and samples real shadcn/Base UI blocks; Jev does not generate JSX or copy.Coding & developer toolsMore projects & source code
jev-dev-harness — READMEgithub.com0README presents a safety harness/runtime toolkit for coding agents powered by TypeSafe AI / Jev; no execution result is inferred here.Coding & developer toolsMore projects & source code
jev-diffgithub.com0Compare saved Jev decisions before changing a model, question, or threshold. Offline CLI, zero runtime dependencies.Coding & developer toolsMore projects & source code
jev-diff — READMEgithub.com0An offline CLI compares saved Jev responses by case ID and shows decision changes; README says it makes no model calls and its example is synthetic.Coding & developer toolsMore projects & source code
jev-enforcegithub.com9📏 Claude Code plugin that makes Claude follow your AGENTS.md: every reply and edit checked by TypeSafe Jev ✅Coding & developer toolsMore projects & source code
jev-enforce — READMEgithub.com9A Claude Code plugin using TypeSafe Jev to check replies and file edits against CLAUDE.md or AGENTS.md rules. Rule text, edits and replies are sent to the API; this is a compliance-check integration, not a guarantee that an agent obeys every rule.Coding & developer toolsMore projects & source code
jev-enhanced-plugingithub.com0A Claude Code plugin that takes skills already in use and adds one thing to each: a typed judgment from TypeSafe Jev (System One) at the points where the skill previously relied on the agent's own impression.Coding & developer toolsMore projects & source code
jev-enhanced-plugin — READMEgithub.com0The agent plugin adds advisory Jev probability packs alongside existing skill procedures; people can use or ignore the added judgments.Coding & developer toolsMore projects & source code
jev-fingerprintgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-fingerprint — READMEgithub.com0An alpha design groups error reports by normalized fingerprints and asks Jev whether candidate incidents share a cause; README says the Jev transport remains unwired.Coding & developer toolsMore projects & source code
jev-flow-harnessgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-flow-harness — READMEgithub.com0A local node editor prototypes wiring an LLM to Jev evaluators with a threshold gate; the README says it does not touch firmware.Coding & developer toolsMore projects & source code
jev-focus-guardgithub.com0Jev Focus Guard: a local Chrome extension that asks Jev (System One) whether page elements are ads or distractions, then hides them.Coding & developer toolsMore projects & source code
jev-focus-guard — READMEgithub.com0Chrome distraction-removal extension sends reduced element descriptors to Jev, hides only allowed high-confidence candidates after local primary-content checks and supports undo/restore.Coding & developer toolsMore projects & source code
jev-frontend-qa — READMEgithub.com5A browser-based frontend QA tool asks Jev to select operations and targets from observed controls; its reported acceptance covers synthetic scenarios, with exact inputs and checks kept in code.Coding & developer toolsMore projects & source code
jev-frontier-100github.com—CLI & Pipelines: \\Jev scores 77.0%; Qwen3.5 2B with a 2,048-token thinking budget scores 82.0%; Qwen3.5 4B with the same budget scores 96.7%.\\ This small benchmark makes Jev's observed reasoning limits tangible through nine local-model settings.Coding & developer toolsMore projects & source code
jev-gategithub.com0Claude Code plugin: Jev (TypeSafe) checks Claude's plan against your request before it builds, and its final summary before it stops; gaps go back to Claude until Jev passes it.Coding & developer toolsMore projects & source code
jev-gategithub.com0Calibrated PR review gates powered by TypeSafe Jev: a GitHub Action, a local CLI, and an OpenCode pluginCoding & developer toolsMore projects & source code
jev-gate — jev.tsgithub.com0JEV Gate’s code-review path uses the TypeSafe client to request Jev/System One decisions for candidate changes.Coding & developer toolsMore projects & source code
jev-gate — READMEgithub.com0Claude Code plugin uses Jev to compare plans and final itemized summaries with user requirements, blocking edits before plan coverage and running bounded correction rounds; it reviews claimed delivery, not proven execution.Coding & developer toolsMore projects & source code
jev-gate — READMEgithub.com0JEV Gate’s code-review path uses the TypeSafe client to request Jev/System One decisions for candidate changes.Coding & developer toolsMore projects & source code
jev-gatesgithub.com0Claude Code hook gates, judged by Jev, that catch edits breaking CLAUDE.md rules, out-of-scope changes, unaddressed asks, false "tests pass" claims and commit messages the diff does not support; none ever approves.Coding & developer toolsMore projects & source code
jev-gates — READMEgithub.com0Claude Code hook gates use Jev to escalate questionable changes, commands and completion claims rather than automatically approve actions.Coding & developer toolsMore projects & source code
jev-github-demogithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-github-demo — READMEgithub.com0The README describes an internal PR-contribution dashboard combining fixed rules with Jev judgments. GitHub import is read-only and does not call Jev; after dry-run review, only unresolved category, impact, complexity, or risk judgments are sent. This is not a tested personnel-evaluation system.Coding & developer toolsMore projects & source code
jev-github-quality-gategithub.com1A Jev-powered quality gate for GitHub issues, pull requests, and commits.Coding & developer toolsMore projects & source code
jev-github-quality-gate — READMEgithub.com1GitHub activity quality gate combines optional Jev judgments with policy thresholds to classify issues, PRs and commits for human review.Coding & developer toolsMore projects & source code
jev-graph-searchgithub.com11Jev-assisted retrieval and evidence-preserving inspection for local Markdown, Obsidian vaults, and Logseq Markdown graphs.Coding & developer toolsMore projects & source code
jev-graph-search — READMEgithub.com11Jev Graph Search makes a local notes shortlist and uses Jev to rank candidates against a question.Coding & developer toolsMore projects & source code
jev-graphraggithub.com1Small demos + use-case backlog: TypeSafe AI's Jev as a calibrated decision layer for GraphRAG pipelines on Neo4j.Coding & developer toolsMore projects & source code
jev-groundsgithub.com1Chrome extension: a badge rating how well-argued an article is, via the Jev API.Coding & developer toolsMore projects & source code
jev-grounds — READMEgithub.com1Jev Grounds is a browser extension that uses Jev for ambiguity checks and a rubric score on extracted article text.Coding & developer toolsMore projects & source code
jev-guardgithub.com0Static code scanner (SAST linter) that finds insecure usage of Jev / TypeSafe guardrails in Python. Not a guardrail itself — build-time analysis, offline, no API key.Coding & developer toolsMore projects & source code
jev-guard — READMEgithub.com0jev-guard statically scans source for insecure Jev/System One patterns when those models are used as guardrails.Coding & developer toolsMore projects & source code
jev-hooksgithub.com0Claude Code commit reviewer driven by Jev-style typed decisions, with a measured question bench (self-hosted rizzo-flow or TypeSafe Jev)Coding & developer toolsMore projects & source code
jev-hooks — READMEgithub.com0Claude Code commit-review plugin sends diff chunks to a System One backend and applies editable probability policy for review lanes; effort routing and GitHub Action remain planned, with synthetic thresholds not a validated gate.Coding & developer toolsMore projects & source code
jev-issue-radar — READMEgithub.com1A read-only GitHub duplicate-issue dashboard uses Jev via OpenRouter in the local app; the public sample shows synthetic, hand-authored decisions.Coding & developer toolsMore projects & source code
jev-judgegithub.com0Jev CLI for the jev-judge Claude/Codex skill. Wraps the official @typesafe-ai/sdk.Coding & developer toolsMore projects & source code
jev-judge — READMEgithub.com0The agent skill wraps TypeSafe Jev's typed yes/no, choice, and score judgments for agent clients.Coding & developer toolsMore projects & source code
jev-kiyafet-bulgithub.com—CLI & Pipelines: Kod yerleri: \lib/filtre.ts\ (Adım 1), \lib/regex.ts\, \lib/daralt.ts\, \lib/sirala.ts\ (Adım 2 + sıralama), \lib/ara.ts\ (hat), \lib/jev.ts\ (istemci), \lib/kota.ts\, \app/api/ara/route.ts\, \components/\ (arayüz).Coding & developer toolsMore projects & source code
jev-labeler-actiongithub.com0Zero-config AI issue labeling with TypeSafe's Jev. No generated text. Unofficial.Coding & developer toolsMore projects & source code
jev-labeler-action — READMEgithub.com0Jev Labeler asks one yes/no Jev question per repository label using its description, then adds labels that clear a threshold.Coding & developer toolsMore projects & source code
jev-leftpadgithub.com74Joke npm package that left-pads strings by asking Jev to choose between space_0 and space_10 options, one API request per call, so it can add at most 10 spaces and costs more than padStart().Coding & developer toolsMore projects & source code
jev-lens — READMEgithub.com0A Claude Code companion asks Jev yes/no questions about changes since the last review and writes a verdict; the README says it neither blocks Claude nor edits files.Coding & developer toolsMore projects & source code
jev-linkedin-slopgithub.com0Chrome extension that rings AI slop in your LinkedIn feed, judged by typesafe/jev-1.13 — a structured decision modelCoding & developer toolsMore projects & source code
jev-linkedin-slop — READMEgithub.com0This Chrome extension marks LinkedIn posts using Jev 1.13 served through OpenRouter; the cited model route is not a direct TypeSafe endpoint.Coding & developer toolsMore projects & source code
jev-linkedin-slop-filtergithub.com5Slams a BAIT, CORP or BRAG stamp onto LinkedIn engagement-bait, judged live by Jev (TypeSafe System One).Coding & developer toolsMore projects & source code
jev-linkedin-slop-filter — READMEgithub.com5Uses TypeSafe Jev to stamp LinkedIn posts with BAIT, CORP or BRAG labels and confidence while leaving posts readable.Coding & developer toolsMore projects & source code
jev-lintgithub.com6A fuzzy linter for coding agents. It checks the code your agent writes against your team's best practices while the agent is still working, not at code review.Coding & developer toolsMore projects & source code
jev-lintgithub.com113Linter where each rule is one sentence: ast-grep picks the code to check and Jev scores the sentence against each match, catching drift like comments that promise behavior the body does not have.Coding & developer toolsMore projects & source code
jev-lintgithub.com0An experiment with semantic code linting using Jev from TypeSafe AICoding & developer toolsMore projects & source code
jev-lintgithub.com2Linter that checks JavaScript and TypeScript against project conventions written in plain English, inline or in Markdown, with Jev reporting possible violations in compact output built for coding agents.Coding & developer toolsMore projects & source code
jev-lint — READMEgithub.com6A fuzzy linter sending agent-written code and team rules to TypeSafe Jev for rule judgments. Included as a review hook, not proof of perfect rule compliance or measured latency.Coding & developer toolsMore projects & source code
jev-lint — READMEgithub.com113Jev-lint documents using Jev probabilities to check code against natural-language rules, alongside syntax-oriented tooling.Coding & developer toolsMore projects & source code
jev-lint — READMEgithub.com0jev-lint is a spike exploring semantic linting with TypeSafe Jev's System One model.Coding & developer toolsMore projects & source code
jev-linter-action — READMEgithub.com1The GitHub Action reviews selected repository files with user-written yes/no questions and uses Jev probabilities against configured thresholds.Coding & developer toolsMore projects & source code
jev-logtriage — READMEgithub.com1jev-logtriage collapses log batches, asks Jev whether they warrant action, then applies code gates to the structured judgments.Coding & developer toolsMore projects & source code
jev-lspgithub.com0An LSP server whose ambient pass runs the rules a repository states in .jev/rules/.json: each candidate line is sent to Jev, a classifier that answers one typed question with a value and a probability, never prose.Coding & developer toolsMore projects & source code
jev-lsp — READMEgithub.com0A proof-of-concept LSP server exposes repository .jev/rules as diagnostics, code actions, and code lenses; the README labels the project WIP.Coding & developer toolsMore projects & source code
jev-msw — READMEgithub.com1jev-msw supplies contract-aware MSW handlers for tests that exercise the TypeSafe JavaScript SDK without making live Jev calls.Coding & developer toolsMore projects & source code
jev-needs-review — READMEgithub.com0A prototype classifies pull requests as merge candidates or requiring human review; the README says it does not merge, approve, comment on, or run repository code.Coding & developer toolsMore projects & source code
jev-nlgrep — READMEgithub.com4nlgrep searches code, docs, and logs by meaning, asking Jev to score matches; its README distinguishes a preview plan that does not call Jev.Coding & developer toolsMore projects & source code
jev-oas-sentinel — READMEgithub.com5JEV OAS Sentinel asks TypeSafe Jev bounded semantic questions about changed OpenAPI behavior and describes itself as an independent community project.Coding & developer toolsMore projects & source code
jev-oxlintgithub.com1Builds a project-specific Oxlint plugin from an agent skill, combining deterministic AST checks with narrow Jev questions for rules that need contextual judgment.Coding & developer toolsMore projects & source code
jev-plannergithub.com—CLI & Pipelines: With \\N\\ agents, \--mode ultra\ makes \\2N + 1\\ agent calls: drafts, reviews, and final synthesis, plus \\N\\ if Jev requests another review. The default \balanced\ makes as few as \\N + 1\\ and never more than \ultra\; \fast\ makes \\N\\, or \\N + 1\\ when it merges. Each Jev evaluation is a separate TypeSafe call.Coding & developer toolsMore projects & source code
jev-playground — READMEgithub.com0A TypeScript playground tests Jev structured answers through TypeSafe direct, OpenRouter, or Vercel AI Gateway.Coding & developer toolsMore projects & source code
jev-playwright — READMEgithub.com6A Playwright reporter using TypeSafe Jev yes/no decisions to select existing tests relevant to a code change before execution.Coding & developer toolsMore projects & source code
jev-policy-studiogithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-policy-studio — READMEgithub.com0Visual Jev policy workspace designs typed questions, fuzzes payloads and exports integrations; without an API key it uses a local mock evaluator.Coding & developer toolsMore projects & source code
jev-proxygithub.com0Local observability proxy and playground for TypeSafe JevCoding & developer toolsMore projects & source code
jev-proxygithub.com0Local proxy for the TypeSafe Jev API: records, caches and replays calls, with a web UI to inspect them. Unofficial.Coding & developer toolsMore projects & source code
jev-proxy — READMEgithub.com0A local Jev observability proxy records System One requests and responses with their payloads; the README warns that stored payloads are cleartext on the machine.Coding & developer toolsMore projects & source code
jev-proxy — READMEgithub.com0Rust local proxy records, caches and replays TypeSafe Jev System One calls and provides a web interface to inspect them.Coding & developer toolsMore projects & source code
jev-qa-demosgithub.com—CLI & Pipelines: Demos for the video "Jev explained for testers": TypeSafe AI's System One model, called through \\Vercel AI Gateway\\ with the AI SDK's \experimental\_evaluate\.Coding & developer toolsMore projects & source code
jev-rcagithub.com0Prototype of Jev-powered Root Cause Analysis for eventsCoding & developer toolsMore projects & source code
jev-rca — READMEgithub.com0jev-rca sends candidate root-cause, co-symptom, or unrelated judgments to Jev for CMDB event analysis.Coding & developer toolsMore projects & source code
jev-req-gategithub.com0A quality gate for AI-generated requirements, powered by Jev (TypeSafe System One). Routes each requirement to PASS / REVIEW / BLOCK with calibrated probabilities. Real benchmark vs LLM: 0 false blocks, 9x cheaper.Coding & developer toolsMore projects & source code
jev-req-gate — READMEgithub.com0Requirements-quality gate batches Jev atomic questions across five dimensions and routes by code thresholds to accept, human review or return; keyless demos simulate the routing rather than call the API.Coding & developer toolsMore projects & source code
jev-reranker — READMEgithub.com4A CLI uses TypeSafe Jev to rerank retrieved JSON documents, filter out candidates without usable evidence, or extract query-specific passages.Coding & developer toolsMore projects & source code
jev-review (MCP) — READMEgithub.com230Jev Review is a local MCP server that gives coding agents structured quality scores from Jev while leaving diagnosis and code changes to the agent.Coding & developer toolsMore projects & source code
jev-review — READMEgithub.com652Jev Review keeps review orchestration in code and uses Jev for bounded judgments over changes.Coding & developer toolsMore projects & source code
jev-review — READMEgithub.com0This autonomous PR-review bot is built around Jev, while its README explicitly keeps approval writes disabled.Coding & developer toolsMore projects & source code
jev-review-action — READMEgithub.com2A GitHub Action accepts configurable criteria, gets typed Jev judgments for submissions or PRs, and lets code apply policy and post a comment.Coding & developer toolsMore projects & source code
jev-review-mcp — READMEgithub.com1A single-purpose MCP code-review gate turns a git diff into a typed Jev decision; without an API key, it runs in offline mock mode.Coding & developer toolsMore projects & source code
jev-rulesgithub.com62Claude Code plugin that asks Jev one yes/no question per standing rule, whether the request is about it, and passes Claude only the matching rules and map documents in well under a second.Coding & developer toolsMore projects & source code
jev-rules — READMEgithub.com62Claude Code plugin uses Jev to select project rules relevant to each prompt and pass the matching rules to Claude.Coding & developer toolsMore projects & source code
jev-search — READMEgithub.com0A Pi code-search extension asks Jev to reorder FFF matches by the user’s intent, adding intent to ffgrep while leaving fffind unchanged.Coding & developer toolsMore projects & source code
jev-semgrep — READMEgithub.com145This semantic grep uses Jev to judge whether each line matches a natural-language meaning and then applies a threshold; the README warns that searched lines are sent to TypeSafe.Coding & developer toolsMore projects & source code
jev-sentinelgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-sentinel — READMEgithub.com0An infrastructure watchdog sends telemetry snapshots to Jev for typed advice and can alert through Telegram; its README says suggested self-healing actions are logged, not executed.Coding & developer toolsMore projects & source code
jev-seo — READMEgithub.com91Rust SEO/GEO CLI and MCP tools use TypeSafe Jev for page, site and keyword judgments alongside local audit functions.Coding & developer toolsMore projects & source code
jev-skillgithub.com—CLI & Pipelines: Jev skill for Claude Code and Codex: TypeSafe System One guide, Python CLI, 5 tested recipes, and findings from 19,367 Jev calls.Coding & developer toolsMore projects & source code
jev-skill-gate — READMEgithub.com7Uses TypeSafe Jev relevance scores to select Claude Code skills and write skillOverrides, limiting which descriptions enter the model’s context.Coding & developer toolsMore projects & source code
jev-skill-scoutgithub.com0Audit CLI that replays Claude Code transcripts through Jev to find turns where a skill should have loaded and did not, plus a mod that suggests the skill live; the author found 72% of skill-needing turns loaded none.Coding & developer toolsMore projects & source code
jev-skill-scout — READMEgithub.com0A CLI replays Claude Code prompts through Jev to flag turns where a skill should have loaded but did not; README-reported counts are not revalidated.Coding & developer toolsMore projects & source code
jev-smart-retrygithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-smart-retry — READMEgithub.com0A GitHub Action retries a failed CI command unchanged only when Jev judges a retry could help; successful commands skip the Jev call.Coding & developer toolsMore projects & source code
jev-sortgithub.com1TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-sort — READMEgithub.com1An archived Jev row-classification CLI now points to jev-tools; historical documentation shows typed classification/score columns, not a currently maintained standalone implementation.Coding & developer toolsMore projects & source code
jev-sort — READMEgithub.com1An archived Jev row-classification CLI now points to jev-tools; historical documentation shows typed classification/score columns, not a currently maintained standalone implementation.Coding & developer toolsMore projects & source code
jev-specgithub.com10CLI that checks code against the requirements in Markdown specs, asking Jev one question per requirement and failing a pre-commit hook or CI build when the probability falls below your threshold.Coding & developer toolsMore projects & source code
jev-startergithub.com3Visually configure Jev questions, test them live, export ready-to-use code.Coding & developer toolsMore projects & source code
jev-starter — READMEgithub.com3A visual Jev console for configuring Noul, Choice and Score questions, trying decisions and exporting Python, JavaScript or cURL requests.Coding & developer toolsMore projects & source code
jev-stop-guardgithub.com—Stop hook for Codex, Cursor and Devin CLI that asks Jev whether the main agent is stopping with requested work still undone, and blocks the stop when the evidence is strong.Coding & developer toolsMore projects & source code
jev-swap — READMEgithub.com1jev-swap finds decision-like LLM calls in TypeScript/JavaScript or Python, generates a Jev module with a confidence-gated fallback, and replays recorded traffic for comparison; savings estimates are not guaranteed.Coding & developer toolsMore projects & source code
jev-switchgithub.com—CLI & Pipelines: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details.Coding & developer toolsMore projects & source code
jev-tab-order — READMEgithub.com1A Chrome extension uses Jev to organize tabs and groups by meaning and user-defined sorting rules.Coding & developer toolsMore projects & source code
jev-tab-organizergithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-tab-organizer — READMEgithub.com0Chrome tab organizer offers reviewable local groups and optional consented Jev semantic classification through a user-controlled gateway; its current tidy-up feature suggests, but does not close or save tabs.Coding & developer toolsMore projects & source code
jev-table — READMEgithub.com0jev-table adds TypeSafe Jev-backed typed-answer and probability columns plus a review queue to CSV workflows.Coding & developer toolsMore projects & source code
jev-test-filter — READMEgithub.com30A test-filter CLI gives TypeSafe Jev the Git diff as shared state and asks one question per test.Coding & developer toolsMore projects & source code
jev-test-impact — READMEgithub.com0Test-impact selector combines dependency/ownership rules with Jev judgments for ambiguous change-test pairs; uncertain API results fall back to the full suite.Coding & developer toolsMore projects & source code
jev-timemachinegithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-timemachine — READMEgithub.com0jev-timemachine stores Jev judgments by model, question, and item fingerprints so policies can be replayed over historical data.Coding & developer toolsMore projects & source code
jev-tmmluplus-evalgithub.com—CLI & Pipelines: Jev is not a chat model. You hand it a \state\ plus a map of typed questions, and it returns one typed answer each — with calibrated probabilities and \\no generated text\\.Coding & developer toolsMore projects & source code
jev-totogithub.com—CLI & Pipelines: Counting and comparison happen in Rust, because Jev is documented as unreliable at arithmetic. Jev receives per-number facts plus plain-English labels and answers two questions per number in one request: a yes/no probability for "drawn in the next draw" and a five-level cold-to-hot score with confidence. Module layout and commands are in \CLAUDE.md\.Coding & developer toolsMore projects & source code
jev-tracegithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-trace — READMEgithub.com0This alpha proposes a requirements-to-code-and-test traceability matrix with Jev-assisted matching; accepted links still need human review and do not prove correctness.Coding & developer toolsMore projects & source code
jev-triage — READMEgithub.com4A public GitHub repository triage dashboard uses TypeSafe System One to answer fixed typed questions about issues and pull requests.Coding & developer toolsMore projects & source code
jev-ts-replgithub.com2TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jev-ts-repl — READMEgithub.com2jev is an interactive terminal REPL for composing TypeSafe System One requests; without TYPESAFE_API_KEY, its README says answers are simulated locally.Coding & developer toolsMore projects & source code
jev-webmcp-extension — READMEgithub.com122The WebMCP extension uses TypeSafe Jev to select and populate page-exposed tool calls from their schemas, returning typed probabilities.Coding & developer toolsMore projects & source code
jev-yaba-wechatgithub.com—CLI & Pipelines: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details.Coding & developer toolsMore projects & source code
jev-zorkgithub.com1Jev plays Zork I: on every turn Jericho supplies the valid actions and Jev picks one command with a single Choice, with its confidence and an anti-loop policy shown on a French replay dashboard.Coding & developer toolsMore projects & source code
jev.nvimgithub.com8Neovim plugin that asks a plain-language question of every Treesitter function in the buffer, has Jev score each one, and lists the hits in quickfix ranked by probability, with the scores shown as virtual text.Coding & developer toolsMore projects & source code
jev.nvim — READMEgithub.com8jev.nvim sends one question per function to TypeSafe Jev; the README says its demo clip uses a test fake, so its displayed probabilities are fixtures.Coding & developer toolsMore projects & source code
jev2ui — READMEgithub.com3Apparite uses TypeSafe Jev for structured screen-design decisions, Gemini for copy, and code to assemble a nonfunctional UI mock; the README says an apparition is not an app.Coding & developer toolsMore projects & source code
JEV_COMMAND_GUARDgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
JEV_COMMAND_GUARD — READMEgithub.com0Jev Command Guard uses typed Jev judgments in a command guard while keeping enforcement decisions in code.Coding & developer toolsMore projects & source code
Jev_Onco_Statistical_Hierarchygithub.com—Jev\_Onco\_Statistical\_Hierarchy - CLI & Pipelines: New prepares use \molecular-origin-v6\. The Choice, the two gated checks, and five boundary checks share the evidence and cannot read each other's answers. Boundary answers are reported with the result. Gates still use the leading score, its margin, sufficiency, and conflict. The parent of the leading class is a local lookup on the taxonomy. A request may use Jev's window: 64k tokens for the call, and 32k for the evidence plus the longest question.Coding & developer toolsMore projects & source code
jevassert — READMEgithub.com0jevassert provides regression tests for Jev question packs: it records live predictions, then checks saved predictions offline and deterministically without a model call during the check.Coding & developer toolsMore projects & source code
jevc — READMEgithub.com8jevc turns natural-language agent rules into narrow typed Jev questions and ordinary-code reducers, making rule gates testable without handing over control flow.Coding & developer toolsMore projects & source code
jevcheckgithub.com2Behavioral contracts for TypeSafe Jev — pin production expectations, eval model upgrades, catch flips and confidence regressions.Coding & developer toolsMore projects & source code
jevcodegithub.com1CLI & Pipelines: This repository provides an Astro-based multilingual documentation site explaining Jev's Choice, Score, and Noul decision primitives with architecture patterns and usage examples.Coding & developer toolsMore projects & source code
JevCompact — READMEgithub.com0JevCompact asks Jev for keep/drop decisions on individual tool calls instead of rewriting session text.Coding & developer toolsMore projects & source code
jevctlgithub.com0CLI for typed AI judgments via TypeSafe or OpenRouterCoding & developer toolsMore projects & source code
jevctl — READMEgithub.com0jevctl is a clean-room CLI that sends typed state/questions to TypeSafe or OpenRouter’s decisions endpoint and returns structured answers.Coding & developer toolsMore projects & source code
jevcumbergithub.com1Cucumber runner that needs only the .feature file: for each Gherkin step Jev picks the action, control and literal value from what is on the page, Playwright performs it, and it refuses to guess when unsure.Coding & developer toolsMore projects & source code
jevcutgithub.com—CLI & Pipelines: Auto-clipper that turns long videos (podcasts, talks, essays, comedy) into short standalone clips for Shorts, Reels and TikTok. Code lists every possible cut; an AI judge picks where each clip starts and ends. Benchmarked on 38 hand-labelled videos.Coding & developer toolsMore projects & source code
Jevestgithub.com1Automated PR review pipeline using Jev (TypeSafe AI) as a millisecond decision layer over an LLM reviewerCoding & developer toolsMore projects & source code
Jevest — READMEgithub.com1Jevest puts Jev as the decision layer around an LLM-based pull-request reviewer; README says Jev never judges source code.Coding & developer toolsMore projects & source code
jevgategithub.com9Code-review gate for CI and coding agents: asks TypeSafe Jev small typed questions about functions, files, tests and docs, and reports findings with locations and probabilitiesCoding & developer toolsMore projects & source code
jevgategithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jevgate — READMEgithub.com9A code-review gate that builds local evidence units and asks TypeSafe Jev typed questions about functions, tests and documentation. Code combines answers into located findings for CI or agents; published accuracy figures are author measurements, not reproduced by this review.Coding & developer toolsMore projects & source code
jevgate — READMEgithub.com0Claude Code pre-operation plugin asks Jev about shell/file effects, then code routes actions to allow, ask or refuse.Coding & developer toolsMore projects & source code
jevgrepgithub.com1Streaming grep-by-meaning CLI for logs and other text that asks Jev one yes/no question per line and prints matches about 0.7 s after they are written, around $0.004 per 1,000 lines, with a benchmark against Claude.Coding & developer toolsMore projects & source code
jevgrepgithub.com1TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jevgrep — READMEgithub.com1A streaming semantic grep CLI sends one yes/no question per line to TypeSafe Jev and prints only matching lines.Coding & developer toolsMore projects & source code
jevgrep — READMEgithub.com92JevGrep asks Jev to score eligible source fragments so coding agents can search code by behavior rather than only by literal text.Coding & developer toolsMore projects & source code
jevgrep — READMEgithub.com1jevgrep is a natural-language code search CLI whose README names TypeSafe Jev as its default backend, with other backends available.Coding & developer toolsMore projects & source code
JevGuardgithub.com1Claude Code and Codex CLI plugin that turns CLAUDE.md or AGENTS.md into project rules and checks them with Jev in PreToolUse and Stop hooks, instead of trusting the agent to remember them.Coding & developer toolsMore projects & source code
JevGuardgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
JevGuard — READMEgithub.com1A Claude Code/Codex plugin checks project rules with Jev and can block a tool call or response through hooks.Coding & developer toolsMore projects & source code
JevGuard — READMEgithub.com0JevGuard is an in-development semantic policy checker that asks Jev a narrow question per rule and then applies local gates; the pinned README says the current release is observe-only and does not block tasks.Coding & developer toolsMore projects & source code
jevidencegithub.com0Measure an agent or a prompt with a panel of typed questions, judged by JevCoding & developer toolsMore projects & source code
jevidencegithub.com0Let Jev judge. Let your code decide. Python issue-routing sandbox with labeled replay, an offline demo, and Kev support.Coding & developer toolsMore projects & source code
jevidence — READMEgithub.com0jevidence wraps an LLM step and asks Jev a fixed panel of typed questions.Coding & developer toolsMore projects & source code
jevidence — READMEgithub.com0Jevidence is a Python sandbox for turning TypeSafe Jev judgments into testable issue-routing suggestions, leaving the final policy to application code.Coding & developer toolsMore projects & source code
Jeviesgithub.com7Python helpers that use Jev to design and review Jev setups: suggesting Noul, Score or Choice, reviewing instructions and options, checking score rubrics and testing inputs against preset rules.Coding & developer toolsMore projects & source code
jeview — READMEgithub.com61Jeview is a local visualization gateway that forwards client calls to TypeSafe Jev and records them for inspection; the README says it does not make decisions itself.Coding & developer toolsMore projects & source code
jevifygithub.com0Jev-style Choice, Score and Noul decisions from ordinary models. An optional DSH plugin.Coding & developer toolsMore projects & source code
jevifygithub.com0grep for meaning: find the error in a 10,000-line log, a commit by description, the command for a task. A Unix CLI, no key needed. Powered by Jev from TypeSafe AI.Coding & developer toolsMore projects & source code
jevify — READMEgithub.com0Jevify is an independent DSH plugin that wraps configured third-party models in Jev-style Choice, Score, and Noul interfaces; it is not TypeSafe Jev itself.Coding & developer toolsMore projects & source code
jevify — READMEgithub.com0jevify uses Jev’s typed choices to find a described error in an existing build rather than generating text.Coding & developer toolsMore projects & source code
jevkeep — READMEgithub.com1JevKeep asks TypeSafe API Jev which Codex conversation excerpts remain relevant before compaction.Coding & developer toolsMore projects & source code
jevkit — READMEgithub.com3A CLI for TypeSafe Jev questions returning probabilities, labels and scores, with offline question-set linting before live requests.Coding & developer toolsMore projects & source code
jevkit-js — READMEgithub.com1jevkit is an unofficial set of tools for building around TypeSafe Jev, explicitly distinct from TypeSafe’s own SDK.Coding & developer toolsMore projects & source code
JevLensgithub.com1Jev decision visualisation and debugging panel for coding agents. MCP server + local web timeline, confidence alerts and trace export.Coding & developer toolsMore projects & source code
JevLens — READMEgithub.com1JevLens is a local MCP server and panel that wraps existing systemOne calls and records state, questions, answers, probabilities, confidence, and latency to append-only JSONL; it observes calls rather than making decisions or routing tools.Coding & developer toolsMore projects & source code
jevlintgithub.com2A linting tool combining static analysis and Jev queries to help improve... Jev queriesCoding & developer toolsMore projects & source code
JevLintgithub.com11Semantic linter that checks source files against coding conventions written in plain English and reports the probability that each file violates a rule, for agent write-check-fix loops or CI.Coding & developer toolsMore projects & source code
jevlint — READMEgithub.com2A Jev request linter inspecting state/questions for documented query pitfalls before sending System One requests. Static linting and optional live checks are distinct.Coding & developer toolsMore projects & source code
JevLint — READMEgithub.com11A semantic lint workflow lets users write coding conventions in plain English, with Jev identifying meaning-level violations for an agent to fix.Coding & developer toolsMore projects & source code
jevlogs — READMEgithub.com16Jev Logs uses TypeSafe Jev for diagnostic-value, priority, and routing annotations in its logs workflow.Coding & developer toolsMore projects & source code
jevmem — READMEgithub.com104jevmem asks Jev to check memory lines not written on the current machine before showing them to Claude.Coding & developer toolsMore projects & source code
JevMetergithub.com102Put a live Jev (TypeSafe) meter on any video: every sentence scored, rendered as a 16:9 edit.Coding & developer toolsMore projects & source code
jevmetricsgithub.com4Experimental OpenTelemetry Collector metrics processor that infers the operational value of metric instruments from their metadata and annotates or filters them before they reach a backend.Coding & developer toolsMore projects & source code
jevmetrics — READMEgithub.com4jevmetrics calls TypeSafe Jev to assess metric instruments for retention; its README says filtering and cost-saving effects need evaluation on your telemetry.Coding & developer toolsMore projects & source code
jevoptgithub.com4C/C++ compiler driver that pairs Clang's optimizer with Jev, asking at every discretionary call site whether to inline based on the LLVM IR, original source, and build context.Coding & developer toolsMore projects & source code
jevort — READMEgithub.com0Pi extension asks Jev to recommend a reasoning-effort tier for each prompt; it only notifies by default and applies changes only when explicitly enabled.Coding & developer toolsMore projects & source code
jevotron — READMEgithub.com6A Jev-powered field-anomaly CLI scanning structured/text data and databases, then exporting a focused queue for human review.Coding & developer toolsMore projects & source code
jevplay — READMEgithub.com0JevPlay is a playground for building custom Choice/Score/Noul questions and viewing their probability distributions.Coding & developer toolsMore projects & source code
JevPRgithub.com10GitHub App that sends each opened or updated pull request to Jev for a LOW, NORMAL, or SPECIALIST risk verdict, which a deterministic policy maps to auto-approval, a review request, or a check run.Coding & developer toolsMore projects & source code
JevPR — READMEgithub.com10A GitHub App that sends pull-request context to TypeSafe Jev for typed review-routing decisions such as LOW, NORMAL and SPECIALIST. Deterministic policy maps the decision to review/check actions; inclusion does not certify safe automatic approvals.Coding & developer toolsMore projects & source code
JevPromptCoachgithub.com3Claude Code plugin that scores the prompts you write to a coding agent and shows whether your habits improve, with no added latency; backfilling a year of local history costs about six cents.Coding & developer toolsMore projects & source code
JevPromptCoach — READMEgithub.com3An unofficial coding-agent prompt coach using TypeSafe Jev to score prompts and track writing habits, not to generate agent replies.Coding & developer toolsMore projects & source code
jevprune — READMEgithub.com2jevprune filters long command output for coding agents using task context; only Jev-selected runs send the task, command, and candidate lines to TypeSafe.Coding & developer toolsMore projects & source code
jevqgithub.com0Four linters in this ecosystem are powered by Jev. None lints the Jev questions themselves. Nine rules from TypeSafe's own documented failure modes, no API key, no labelled data, 0.03s.Coding & developer toolsMore projects & source code
jevq — READMEgithub.com0jevq locally lints Jev questions against documented failure modes; the README says it needs no API key or model call.Coding & developer toolsMore projects & source code
jevqagithub.com4Jev-guided monkey tester: $0.35 and 5 minutes to find missing features, validation gaps and dead controls before QACoding & developer toolsMore projects & source code
jevqa — READMEgithub.com4JevQA explores a deployed web app against its README specification and uses Jev for per-screen action and oracle judgments; Claude prepares the checklist.Coding & developer toolsMore projects & source code
jevrailgithub.com3Jev powered pre-execution guard for terminal coding agentsCoding & developer toolsMore projects & source code
jevrailgithub.com1Auditable Jev decisions through OpenRouter. A macOS-first TypeScript CLI with Keychain credentials, recorded costs, and conservative retries.Coding & developer toolsMore projects & source code
jevrail — READMEgithub.com1An independent CLI calls Jev through OpenRouter and keeps local request/response, cost and hash records for inspecting batch decisions and interrupted jobs.Coding & developer toolsMore projects & source code
jevrail — READMEgithub.com3Jevrail describes a pre-execution shell-command guard for terminal agents and lists a Jev API key as a prerequisite.Coding & developer toolsMore projects & source code
jevscan-evmgithub.com25Proof-of-concept scanner that builds a heat map of likely bugs in EVM smart-contract code, asking Jev two general, 14 category, and 343 detector questions sourced from audit checklists.Coding & developer toolsMore projects & source code
JevScopegithub.com0Local-first workbench and regression testbench for Jev projects: edit JSON state and choice, score or noul questions, inspect probability distributions, run JSONL cases and compare two project definitions.Coding & developer toolsMore projects & source code
jevscope — READMEgithub.com0JevScope is a local workbench and regression testbench for TypeSafe Jev; README says tests use local fixtures and do not require paid provider calls.Coding & developer toolsMore projects & source code
JevSweepgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
JevSweep — READMEgithub.com0JevSweep uses Jev to prune low-value captured tool output before it reaches the coding model, with removed results recoverable from receipts.Coding & developer toolsMore projects & source code
JevSwitchgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
JevSwitch — READMEgithub.com0JevSwitch is a local migration tool for identifying generative-model calls that may be bounded decisions, then testing candidates on the developer’s own data.Coding & developer toolsMore projects & source code
JevToolsgithub.com4A lightweight collection of developer utilities and scripts designed to streamline Jev development process.Coding & developer toolsMore projects & source code
JevTools — READMEgithub.com4A toolkit and reference implementation provides a web cockpit and utilities for building Jev applications through TypeSafe or OpenRouter.Coding & developer toolsMore projects & source code
jevtracegithub.com0One DevTool for any Jev-powered application. Inspect every state, question, probability, decision and action. Replay decisions locally.Coding & developer toolsMore projects & source code
jevtrace — READMEgithub.com0JevTrace captures decision flows from Jev-powered apps and supports replaying historical Jev decisions.Coding & developer toolsMore projects & source code
jevtriagegithub.com1GitHub Action and Python CLI that asks Jev one Choice per pull request (ready, needs_review or risky) and only passes ready when confidence is at least 0.8, failing closed otherwise.Coding & developer toolsMore projects & source code
jevtriage — READMEgithub.com1jevtriage is a pull-request triage Action/CLI that asks TypeSafe Jev one Choice question; default CI replays fixtures without calling TypeSafe.Coding & developer toolsMore projects & source code
jevudiogithub.com0Editor de codigo open source.Coding & developer toolsMore projects & source code
jevvygithub.com0Use jev model to augment csv files with inferred classification, scoring, or probability scoresCoding & developer toolsMore projects & source code
jevvy — READMEgithub.com0jevvy is a CLI that turns CSV rows into JSON state, asks YAML-defined questions through the TypeSafe Jev API, and appends answers as CSV columns.Coding & developer toolsMore projects & source code
jevxgithub.com2Jev-powered Chrome extension that categorizes X posts and classifies replies in context.Coding & developer toolsMore projects & source code
jevx — READMEgithub.com2A Chrome extension uses TypeSafe Jev to classify X timeline posts and score their apparent AI-generated style.Coding & developer toolsMore projects & source code
jevy-vetgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jevy-vet — READMEgithub.com0OpenCode editing plugin uses Jev judgments to screen weakened tests and checks before writes, and warn about questionable edits or unsupported completion claims.Coding & developer toolsMore projects & source code
jfind — READMEgithub.com2jfind adds a semantic --like predicate to file search, with TypeSafe Jev answering typed yes/no matches.Coding & developer toolsMore projects & source code
jgrepgithub.com131Command-line grep where the pattern is a plain-English description: each line, record, function or diff hunk becomes one Jev yes/no question, judged concurrently as input streams in, including behind tail -f.Coding & developer toolsMore projects & source code
jgrep — READMEgithub.com131jgrep turns each line into a typed yes/no judgment using Jev and offers a streaming semantic search CLI; its local-model backends are separate optional choices.Coding & developer toolsMore projects & source code
jgrep — READMEgithub.com59A semantic code-search CLI asks Jev typed questions about code behavior and returns matching ranges instead of relying only on symbol names.Coding & developer toolsMore projects & source code
joxidegithub.com0Jump to projects by description, powered by zoxide and TypeSafe’s JevCoding & developer toolsMore projects & source code
joxide — READMEgithub.com0A shell directory-jump tool uses Jev to rank known folders when name matching is insufficient; exact-path and name-based queries work locally without a key.Coding & developer toolsMore projects & source code
jsort — READMEgithub.com25Text-sorting CLI uses Jev pairwise comparisons and a Bradley–Terry scale to rank texts along a user-described dimension.Coding & developer toolsMore projects & source code
juraigithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
jurai — READMEgithub.com0Skill-instruction comparison tool uses Jev to assess working-tree versus main requirements, without executing the skills; paired reversed presentations examine ordering bias.Coding & developer toolsMore projects & source code
jury.nvimgithub.com1Calibrated picks for Neovim, judged by TypeSafe Jev. First source: effect-error-pretty.nvimCoding & developer toolsMore projects & source code
jury.nvim — READMEgithub.com1A Neovim plugin asks Jev to choose among precomputed code-fix candidates and shows the suggestion only when its confidence passes a threshold.Coding & developer toolsMore projects & source code
kamchatka Jev supportgithub.com—Jev integration in kamchatka, the terminal agent of the nachalnik Rust runtime: an example of content-aware context compaction and an --advise mode that gives each proposed shell command a color-coded safety rating.Coding & developer toolsMore projects & source code
kamchatka Jev support — repogithub.com—Jev integration in kamchatka, the terminal agent of the nachalnik Rust runtime: an example of content-aware context compaction and an --advise mode that gives each proposed shell command a color-coded safety rating.Coding & developer toolsMore projects & source code
Kiro Crew Jev decision pointsgithub.com—Decision points in the Kiro Crew development workspace where Jev Choices pick skills, route models, rate tool risk, steer messages, recall memory and decide what to keep during compaction.Coding & developer toolsMore projects & source code
Kit eval toolgithub.com—Experimental eval tool in Kit, the Rust coding-agent runtime with a single compose tool: an agent program can send one state with several named Noul, Choice or Score questions to TypeSafe and get named answers back.Coding & developer toolsMore projects & source code
Kit eval tool — repogithub.com—Experimental eval tool in Kit, the Rust coding-agent runtime with a single compose tool: an agent program can send one state with several named Noul, Choice or Score questions to TypeSafe and get named answers back.Coding & developer toolsMore projects & source code
KnoxCoder Jev layergithub.com—Jev module in the built-in coding agent of KnoxCoder, a VS Code-based editor: it judges each agent turn, gates tool calls and context passages, detects semantic doom loops, rescores messages for compaction and scores agent traces.Coding & developer toolsMore projects & source code
KnoxCoder Jev layer — repogithub.com—Jev module in the built-in coding agent of KnoxCoder, a VS Code-based editor: it judges each agent turn, gates tool calls and context passages, detects semantic doom loops, rescores messages for compaction and scores agent traces.Coding & developer toolsMore projects & source code
Koru invariants gategithub.com—Pre-commit gate for the Koru event continuation language: deterministic checks run where declared, and judgment-class invariants are judged by Jev against the staged diff, advisory unless GATE_BLOCK=1.Coding & developer toolsMore projects & source code
Koru invariants gate — repogithub.com—Pre-commit gate for the Koru event continuation language: deterministic checks run where declared, and judgment-class invariants are judged by Jev against the staged diff, advisory unless GATE_BLOCK=1.Coding & developer toolsMore projects & source code
kpatch Jev filtergithub.com—Commit filter in Purdue PurSec Lab's kernel CVE analysis tools that asks Jev security-relevant questions about each Linux/Android kernel patch, such as changes to refcounting, locking, destructors, permission or bounds checks.Coding & developer toolsMore projects & source code
kpatch Jev filter — repogithub.com—Commit filter in Purdue PurSec Lab's kernel CVE analysis tools that asks Jev security-relevant questions about each Linux/Android kernel patch, such as changes to refcounting, locking, destructors, permission or bounds checks.Coding & developer toolsMore projects & source code
ksenxx/kiss_aigithub.com≈554Agent framework whose optional Jev task classifier decides whether a request needs development work, with an LLM fallback.Coding & developer toolsMore projects & source code
KubeDojo Jev epic triagegithub.com—Pre-dispatch triage script for the AI agent fleet that builds a free Kubernetes course: Jev reviews PR and epic state and the script recommends dispatch, parallelism and reviewer family.Coding & developer toolsMore projects & source code
KubeDojo Jev epic triage — repogithub.com—Pre-dispatch triage script for the AI agent fleet that builds a free Kubernetes course: Jev reviews PR and epic state and the script recommends dispatch, parallelism and reviewer family.Coding & developer toolsMore projects & source code
Lean Refactor Arena TypeSafe NCAgithub.com—Harness for the Lean Refactor Arena where Jev acts as a gate, not a generator: Choice, Score and Noul questions judge candidate Lean refactors, such as whether they introduce sorry, while lake compile stays the oracle.Coding & developer toolsMore projects & source code
Lean Refactor Arena TypeSafe NCA — repogithub.com—Harness for the Lean Refactor Arena where Jev acts as a gate, not a generator: Choice, Score and Noul questions judge candidate Lean refactors, such as whether they introduce sorry, while lake compile stays the oracle.Coding & developer toolsMore projects & source code
LeanKG judgegithub.com—Code knowledge graph for AI coding agents that adds a System One judgment layer, calling Jev or a self-hosted Laya server, including a conversation classification path.Coding & developer toolsMore projects & source code
Learn Ukrainian TypeSafe toolinggithub.com—Agent-fleet tooling in a free A1-C2 Ukrainian course repo that uses Jev for pre-dispatch readiness triage of issues, line-level semantic find and skill suggestions.Coding & developer toolsMore projects & source code
Learn Ukrainian TypeSafe tooling — repogithub.com—Agent-fleet tooling in a free A1-C2 Ukrainian course repo that uses Jev for pre-dispatch readiness triage of issues, line-level semantic find and skill suggestions.Coding & developer toolsMore projects & source code
leashgithub.com0Jev-powered guardrail that keeps a coding agent's output quality on a short leash: judges every turn against your un-lintable project rules and tells it which one it broke. Ratcheted, advisory, never a gate.Coding & developer toolsMore projects & source code
leash — READMEgithub.com0Early CLI/library checks project rules with Jev and baselines accepted debt; agent hooks are planned, and without a key it performs no judgment.Coding & developer toolsMore projects & source code
lgtmgithub.com3Prove that your tests actually test something. Powered by Jev and your own API Key.Coding & developer toolsMore projects & source code
lgtm — READMEgithub.com3lgtm asks Jev or another selected provider to compare test blocks with their implementation and identify tests that do not exercise behavior.Coding & developer toolsMore projects & source code
Light Skills Jev integrationgithub.com—Shared Jev integration contract and evals for the Light Skills agent-skill collection, where the project-init, ask-light and agent-config skills use Jev Score, Noul and Choice judgments while code keeps authority.Coding & developer toolsMore projects & source code
Light Skills Jev integration — repogithub.com—Shared Jev integration contract and evals for the Light Skills agent-skill collection, where the project-init, ask-light and agent-config skills use Jev Score, Noul and Choice judgments while code keeps authority.Coding & developer toolsMore projects & source code
Lintusgithub.com9Ruby linter whose rules are plain-language questions in a YAML file; each rule is asked of each file as a Jev Noul, on the whole tree, changed files, or staged files in a pre-commit hook.Coding & developer toolsMore projects & source code
lisagithub.com1LISA: Leak, Injection & Simplicity Auditor. GitHub Action that uses TypeSafe Jev (a system one model) to flag secrets, security vulnerabilities, and unneeded complexity in pull requestsCoding & developer toolsMore projects & source code
lisa — READMEgithub.com1GitHub Action asks Jev typed questions about PR diffs to flag secrets, security issues, complexity and prompt injection for review.Coding & developer toolsMore projects & source code
LLM Debuggergithub.com359VS Code extension that hunts bugs by driving the real debugger, with Jev answering each step's "what next?" and a generation model woken only for hypotheses, breakpoints, expressions and the fix.Coding & developer toolsMore projects & source code
LLM Debugger — demogithub.com—VS Code extension that hunts bugs by driving the real debugger, with Jev answering each step's "what next?" and a generation model woken only for hypotheses, breakpoints, expressions and the fix.Coding & developer toolsMore projects & source code
llm-to-jev — READMEgithub.com5A browser-local converter finds bounded choices, scores, and yes/no decisions in LLM prompts and proposes Jev primitives; live API testing is optional.Coding & developer toolsMore projects & source code
Lockstepgithub.com10Shared decision record for a team's coding agents that checks whether changed code may contradict accepted decisions, using Jev for the recall and recheck stages of its funnel.Coding & developer toolsMore projects & source code
Lockstep — codegithub.com10Shared decision record for a team's coding agents that checks whether changed code may contradict accepted decisions, using Jev for the recall and recheck stages of its funnel.Coding & developer toolsMore projects & source code
Loom Jev merge-risk scoregithub.com—Shadow-mode pre-score for automatic PR merging in an agent orchestration daemon: Jev rates diff composition, blast radius, review depth and revertability before the merge decision.Coding & developer toolsMore projects & source code
Loom Jev merge-risk score — repogithub.com—Shadow-mode pre-score for automatic PR merging in an agent orchestration daemon: Jev rates diff composition, blast radius, review depth and revertability before the merge decision.Coding & developer toolsMore projects & source code
lorenzinigithub.com17Claude Code skills that poll CodeRabbit, Copilot and Codex PR reviewers and decide whether their verdict really permits a merge; an optional Jev shadow classifier flags collapsed review sections that describe unaddressed work.Coding & developer toolsMore projects & source code
lossless-compact — READMEgithub.com0Lossless Compact is a context-compaction plugin whose Jev classifier calls System One when configured; without a key it uses local rules.Coding & developer toolsMore projects & source code
lossless-compact — request.tsgithub.com0Lossless Compact is a context-compaction plugin whose Jev classifier calls System One when configured; without a key it uses local rules.Coding & developer toolsMore projects & source code
LUCID Jev judgment backendgithub.com18Jev support in LUCID, a security-first desktop coding harness around oh-my-pi: typed judgments, the agent's judge() calls and a browser_run policy ported from jev-ultrafast are answered by Jev and traced under each reply.Coding & developer toolsMore projects & source code
Magic Composegithub.com—BB plugin that adds a wand to the new-thread composer: as you type, Jev picks the project, machine, model, reasoning level and environment for the coding-agent thread, moving the pickers before you send.Coding & developer toolsMore projects & source code
Magic Compose — marketplacegithub.com—BB plugin that adds a wand to the new-thread composer: as you type, Jev picks the project, machine, model, reasoning level and environment for the coding-agent thread, moving the pickers before you send.Coding & developer toolsMore projects & source code
Magic Compose — repogithub.com—BB plugin that adds a wand to the new-thread composer: as you type, Jev picks the project, machine, model, reasoning level and environment for the coding-agent thread, moving the pickers before you send.Coding & developer toolsMore projects & source code
Magic Jevgithub.com1Chrome extension that turns a pull request into a Magic 8 Ball: click it and Jev picks one of the 20 classic phrases to answer "should I approve this?" from nine PR signals in about 200 ms.Coding & developer toolsMore projects & source code
Marionette Jev turn judgmentgithub.com—Opt-in Jev turn judgment in Marionette, a desktop AI coding harness built on the Puppetmaster kernel, where one OpenRouter Decisions call picks which skill body to load instead of token-overlap retrieval.Coding & developer toolsMore projects & source code
Marionette Jev turn judgment — repogithub.com—Opt-in Jev turn judgment in Marionette, a desktop AI coding harness built on the Puppetmaster kernel, where one OpenRouter Decisions call picks which skill body to load instead of token-overlap retrieval.Coding & developer toolsMore projects & source code
Math-To-Manimgithub.com—Math and physics animation pipeline that uses TypeSafe Jev to review staged explanation, equation, and code checkpoints.Coding & developer toolsMore projects & source code
MCode enhancement packgithub.com8Patches for the MiniMax Code CLI on DGX Spark with a local GLM model: context meter, continuous context renewal, a compaction fix, and a jev-ultrafast browser agent wired to a local /v1/systemone-compatible policy.Coding & developer toolsMore projects & source code
meaning-assertgithub.com0Test what an AI output means. Natural-language assertions powered by Jev, with pass, fail, and inconclusive results.Coding & developer toolsMore projects & source code
meaning-assert — READMEgithub.com0Semantic assertion library and CLI use Jev to compare output with plain-language requirements and supplied evidence, retaining inconclusive results when evidence or certainty is insufficient.Coding & developer toolsMore projects & source code
meaning-aware-diff — READMEgithub.com0Diff-review library and CLI use Jev to flag weakened assertions, hidden failures and placeholder substitutions next to exact patch hunks, with a limited local-rule alternative.Coding & developer toolsMore projects & source code
meko-jev-code-review-agent — READMEgithub.com0Code-review demonstration asks Jev whether diff hunks violate written rules and stores human rulings in Meko so later reviews can recognize approved exceptions.Coding & developer toolsMore projects & source code
mergeCraftgithub.com11Self-hosted, bring-your-own-key GitHub Action for AI PR review that adds a Jev screening gate to rank and filter diff units before the generative reviewer and verifier run.Coding & developer toolsMore projects & source code
mergeCraft — docsgithub.com11Self-hosted, bring-your-own-key GitHub Action for AI PR review that adds a Jev screening gate to rank and filter diff units before the generative reviewer and verifier run.Coding & developer toolsMore projects & source code
mikan Jev harnessgithub.com—Self-hosted Slack coding agent where call sites needing fast classification, routing or guardrail decisions call Jev directly instead of treating it as a chat provider.Coding & developer toolsMore projects & source code
mikan Jev harness — repogithub.com—Self-hosted Slack coding agent where call sites needing fast classification, routing or guardrail decisions call Jev directly instead of treating it as a chat provider.Coding & developer toolsMore projects & source code
Mjolnir Jev turn verdictsgithub.com—Meta-harness for Codex, Claude Code and other ACP coding agents that uses Jev to judge whether a turn needs user input or is background work, and to rank help-search entries.Coding & developer toolsMore projects & source code
Mjolnir Jev turn verdicts — docsgithub.com—Meta-harness for Codex, Claude Code and other ACP coding agents that uses Jev to judge whether a turn needs user input or is background work, and to rank help-search entries.Coding & developer toolsMore projects & source code
model-picker — READMEgithub.com0A Codex hook periodically asks a configured Jev endpoint to recommend a model and effort level; the README describes suggestions, not automatic switching.Coding & developer toolsMore projects & source code
Moongategithub.com0MoonBit GitHub Action that checks a PR diff against semantic rules stored as JSON, with Jev answering violation, compliant or insufficient evidence per rule and your thresholds deciding the annotations and exit code.Coding & developer toolsMore projects & source code
moongate — READMEgithub.com0Moongate checks pull-request diffs against repository-defined semantic rules using Jev and reports results as PR annotations; this does not imply repository-code execution.Coding & developer toolsMore projects & source code
mustsgithub.com9Validation loop for coding agents that keeps a task open until every declared check passes, including uses: jev checks that ask Jev whether changed files break a prose rule.Coding & developer toolsMore projects & source code
musts — examplegithub.com9Validation loop for coding agents that keeps a task open until every declared check passes, including uses: jev checks that ask Jev whether changed files break a prose rule.Coding & developer toolsMore projects & source code
Navigatorgithub.com232Claude Code context-engineering plugin that adds a typed TypeSafe prompt judge behind its loop, complexity and ambiguity scorers.Coding & developer toolsMore projects & source code
Navigator — docsgithub.com232Claude Code context-engineering plugin that adds a typed TypeSafe prompt judge behind its loop, complexity and ambiguity scorers.Coding & developer toolsMore projects & source code
nestjs-hexagonal Jev semantic rulesgithub.com—Semantic architecture rules in a Claude Code plugin for NestJS hexagonal/DDD code, where Jev checks files for issues like thin controllers, anemic entities or infrastructure leaking into ports, with calibration data.Coding & developer toolsMore projects & source code
nestjs-hexagonal Jev semantic rules — repogithub.com—Semantic architecture rules in a Claude Code plugin for NestJS hexagonal/DDD code, where Jev checks files for issues like thin controllers, anemic entities or infrastructure leaking into ports, with calibration data.Coding & developer toolsMore projects & source code
neuralint — READMEgithub.com1neuralint’s alpha CLI uses Jev decision matrices to check changed code against repository rules; its README identifies check as read-only.Coding & developer toolsMore projects & source code
nl-lintgithub.com0Near-instant natural-language lint for your code.Coding & developer toolsMore projects & source code
nl-lint — READMEgithub.com0nl-lint uses Jev as its current text classifier, returning structured decisions for application-defined rules; this describes the package’s integration, not an independent classifier evaluation.Coding & developer toolsMore projects & source code
nvim-jev-plugingithub.com2TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
nvim-jev-plugin — READMEgithub.com2A Neovim plugin asks Jev to map plain-language requests to a fixed catalog of editor actions.Coding & developer toolsMore projects & source code
o8 judgment refereegithub.com—Judgment layer in o8, a control room for a fleet of coding agents, that asks Jev locked yes/no questions about each diff (docs-only, touches auth, mock data, adds tests) with per-question calibration notes.Coding & developer toolsMore projects & source code
o8 judgment referee — repogithub.com—Judgment layer in o8, a control room for a fleet of coding agents, that asks Jev locked yes/no questions about each diff (docs-only, touches auth, mock data, adds tests) with per-question calibration notes.Coding & developer toolsMore projects & source code
oc-auto-permsgithub.com—OpenCode plugin that checks each proposed tool action against natural-language permission rules with Jev, then allows, denies or asks the user.Coding & developer toolsMore projects & source code
oc-auto-perms — repogithub.com—OpenCode plugin that checks each proposed tool action against natural-language permission rules with Jev, then allows, denies or asks the user.Coding & developer toolsMore projects & source code
Oh My Pi Jev judgmentsgithub.com32,042Native Jev judgment backend in the Oh My Pi coding agent: a judge() call in its eval kernel and a jevify keyword that bulk-classifies diffs, logs or search hits so the agent reads only flagged items.Coding & developer toolsMore projects & source code
Oh My Pi Jev judgments — repogithub.com32,042Native Jev judgment backend in the Oh My Pi coding agent: a judge() call in its eval kernel and a jevify keyword that bulk-classifies diffs, logs or search hits so the agent reads only flagged items.Coding & developer toolsMore projects & source code
oh-my-claudecode Jev judgment pointsgithub.com—Judgment points in the oh-my-claudecode orchestration hooks where Jev decides skill triggers, loop continuation, model tier for delegated tasks, context staleness and task size.Coding & developer toolsMore projects & source code
oh-my-harnessgithub.com17CLI that turns a natural-language description into enforced guardrails (CLAUDE.md, hooks, settings) for Claude Code, Codex and Pi; with a TypeSafe key, Jev picks which guard blocks to enable and how strict to be in one call.Coding & developer toolsMore projects & source code
omg.dev Jev e2e testsgithub.com—End-to-end test layer for the omg.dev mobile app where Jev reads the screen's accessibility tree and answers whether a step is done, whether it hit a dead end and which element to tap next.Coding & developer toolsMore projects & source code
omp-classifiergithub.com0Model-judged permission gate for OMP: classifies bash commands and spawn-bearing eval payloads before they run, prompts on risk, fails closedCoding & developer toolsMore projects & source code
omp-classifier — READMEgithub.com0The README describes an OMP permission gate that sends otherwise-unmatched bash commands and process-spawning eval payloads to Jev; unavailable judgments raise a permission request.Coding & developer toolsMore projects & source code
omp-fabric typed judgmentgithub.com—Typed judgment lane in Fabric, a programmable tool and agent runtime for Oh My Pi, where judgment.ask sends choice, bool and score questions to TypeSafe System One and four gates read the answers.Coding & developer toolsMore projects & source code
omp-fabric typed judgment — repogithub.com—Typed judgment lane in Fabric, a programmable tool and agent runtime for Oh My Pi, where judgment.ask sends choice, bool and score questions to TypeSafe System One and four gates read the answers.Coding & developer toolsMore projects & source code
omp-jevgithub.com7Oh My Pi plugin that adds Jev routing judgments to the coding agent plus an opt-in checkpoint orchestrator, configured through editable XDG files and /jev commands and disabled by default.Coding & developer toolsMore projects & source code
omp-typesafe — READMEgithub.com1omp-typesafe adds advisory Jev review to the omp coding agent and exposes a typesafe_ask tool for typed questions.Coding & developer toolsMore projects & source code
OpenByteInc/QuantDingergithub.com11,782Open-source AI Trading OS, agent trading, and vibe trading, with Jev System One integration. Research, build Python strategies, backtest, and paper/live trade across crypto, stocks, and forex. Launch your own multi-tenant trading SaaS with built-in user management, billing, payments, and settlement.Coding & developer toolsMore projects & source code
opencode-context-pruner — READMEgithub.com4This OpenCode hook prunes stale tool calls/results using Jev judgments while keeping retained messages verbatim; failures are fail-open and leave the request untouched.Coding & developer toolsMore projects & source code
opencode-jev-compaction — jev-compaction.tsgithub.com1Before OpenCode requests, Jev judges whether older tool calls/results remain useful; stale results may be dropped or truncated, while retained calls/results stay verbatim.Coding & developer toolsMore projects & source code
opencode-jev-compaction — READMEgithub.com1Before OpenCode requests, Jev judges whether older tool calls/results remain useful; stale results may be dropped or truncated, while retained calls/results stay verbatim.Coding & developer toolsMore projects & source code
opencode-jev-guardgithub.com2OpenCode 2 plugin that sends every shell command (local or via FarHand) to TypeSafe's Jev and asks you first when it leaves files outside the project, installs software globally, changes global settings, is harmful or exposes private dataCoding & developer toolsMore projects & source code
opencode-jev-guard — READMEgithub.com2An OpenCode plugin submitting local and FarHand remote shell commands to TypeSafe Jev for six typed risk questions before policy applies execution decisions.Coding & developer toolsMore projects & source code
openevals — READMEgithub.com4A service grades agent traces and steps with Jev and can write scores to Langfuse; without a TypeSafe key, only code checks run.Coding & developer toolsMore projects & source code
OpenGantry Jev preflightgithub.com—Experimental preflight in the OpenGantry agent-governance CLI that asks Jev which manifest skill owns a coding agent's intent and which allowed directories are relevant, falling back to heuristics on low confidence.Coding & developer toolsMore projects & source code
OpenGantry Jev preflight — repogithub.com—Experimental preflight in the OpenGantry agent-governance CLI that asks Jev which manifest skill owns a coding agent's intent and which allowed directories are relevant, falling back to heuristics on low confidence.Coding & developer toolsMore projects & source code
openjevgithub.com3Turn any local LLM into a Jev-style System One decision engine: type-safe answers with raw softmax probabilities. 100% offline, zero API cost.Coding & developer toolsMore projects & source code
OpenWork Jev verification — repogithub.com—Evaluator in the OpenWork desktop app's eval testkit that sends a test intent and a dictionary of UI checks to Jev via Vercel AI Gateway, which picks the checks to run and whether they cover the intent in one call.Coding & developer toolsMore projects & source code
ora answer auditgithub.com—Audit script in the ora Claude Code research plugin that checks logged answers against their evidence and, with a key, asks Jev whether versions and dates are carried and whether the conclusion contradicts the evidence.Coding & developer toolsMore projects & source code
ora answer audit — repogithub.com—Audit script in the ora Claude Code research plugin that checks logged answers against their evidence and, with a key, asks Jev whether versions and dates are carried and whether the conclusion contradicts the evidence.Coding & developer toolsMore projects & source code
OrchestKit Jev session categoriesgithub.com≈285Held-out eval for OrchestKit's opt-in Jev session classifier for Claude Code: on 150 sessions Jev matched Haiku at 80.0% vs 79.3% with the same criteria, about 58x faster with about 34x fewer input tokens.Coding & developer toolsMore projects & source code
OrchestKit Jev session categories — repogithub.com≈285Held-out eval for OrchestKit's opt-in Jev session classifier for Claude Code: on 150 sessions Jev matched Haiku at 80.0% vs 79.3% with the same criteria, about 58x faster with about 34x fewer input tokens.Coding & developer toolsMore projects & source code
Ortus Jev pre-turn gategithub.com—Optional pre-turn gate in Ortus, which closes a backlog of bd issues with Claude Code, Codex, Grok or local models, where Jev decides per task whether to run, skip or hand to a human and how risky the action is.Coding & developer toolsMore projects & source code
Ortus Jev pre-turn gate — repogithub.com—Optional pre-turn gate in Ortus, which closes a backlog of bd issues with Claude Code, Codex, Grok or local models, where Jev decides per task whether to run, skip or hand to a human and how risky the action is.Coding & developer toolsMore projects & source code
overseer-judgegithub.com0Typed verdicts for overseer judgments (pane state, task-file lint, review typing) via TypeSafe's JevCoding & developer toolsMore projects & source code
overseer-judge — READMEgithub.com0The Go CLI sends judgments to TypeSafe System One (Jev) and emits typed JSON verdicts.Coding & developer toolsMore projects & source code
Overwatchgithub.com5TypeSafe AI Overwatch — System One / Jev related tooling from the official org.Coding & developer toolsMore projects & source code
oxlint-plugin-jevgithub.com65Experimental Oxlint plugin where rules are plain-English yes/no questions about functions, calls, JSX elements or files, and matches are reported when Jev's yes-probability clears your cutoff.Coding & developer toolsMore projects & source code
oxlint-plugin-jev — READMEgithub.com65An Oxlint plugin sends English yes/no code-rule questions to Jev and reports a lint error when the yes-probability exceeds the configured cutoff.Coding & developer toolsMore projects & source code
Paper Radargithub.com—Python CLI that fetches recent arXiv papers, scores their relevance to a configurable research profile, and writes daily Markdown and JSON rankings.Coding & developer toolsMore projects & source code
Paseo Jev evaluator plugingithub.com3Paseo plugin that exposes Jev through Vercel AI Gateway as an MCP tool, jev_evaluate, to the Codex and Claude agents Paseo launches, including subagents created with its create_agent.Coding & developer toolsMore projects & source code
patdowngithub.com14Fuzzy linter that judges files or changes against rules written in one Markdown file, packaged as a CLI, GitHub Action, Pi write-steering package and Claude Code Write/Edit hooks, with a swappable Jev judge.Coding & developer toolsMore projects & source code
patdown — READMEgithub.com14patdown has a swappable decision provider; its README names TypeSafe as the default backend.Coding & developer toolsMore projects & source code
Pattern MCPgithub.com8MCP server that checks UI component libraries against product requirements before a coding agent builds, with an experimental Jev scorer that rates each Figma component page against the checklist.Coding & developer toolsMore projects & source code
perchgithub.com339Semantic code scanner that parses methods and their call graph, asks Jev typed questions about each method in scope and ranks likely defects and security issues, with custom YAML rules, agent skills and CI gating.Coding & developer toolsMore projects & source code
perch — READMEgithub.com339A semantic code-linting tool using Jev. Its setup explicitly points requests to TypeSafe System One and supports project-specific rules in perch.yaml; inclusion reflects that documented decision-model integration, not verified bug-detection accuracy.Coding & developer toolsMore projects & source code
Percussion CMS issue prescreengithub.com—Maintainer script for the Percussion CMS repository that prescreens a GitHub issue inventory with Jev, recommending which issues to skip, which are PR-sized and which are close candidates, before rule-based filtering.Coding & developer toolsMore projects & source code
Percussion CMS issue prescreen — repogithub.com—Maintainer script for the Percussion CMS repository that prescreens a GitHub issue inventory with Jev, recommending which issues to skip, which are PR-sized and which are close candidates, before rule-based filtering.Coding & developer toolsMore projects & source code
Pi Interactive Shellgithub.com587Pi coding-agent extension that autonomously drives interactive CLIs in an observable overlay, with optional Jev semantic supervision of terminal output.Coding & developer toolsMore projects & source code
Pi Jev extensiongithub.com—Extension in a personal Pi coding-agent configuration that gives the agent advisory Choice, Score and Noul classifiers over text state, backed by TypeSafe or Cloudflare Workers AI.Coding & developer toolsMore projects & source code
Pi Jev extension — repogithub.com—Extension in a personal Pi coding-agent configuration that gives the agent advisory Choice, Score and Noul classifiers over text state, backed by TypeSafe or Cloudflare Workers AI.Coding & developer toolsMore projects & source code
Pi MCP Adapter Jev searchgithub.com1,570Optional Jev layer in a token-efficient MCP adapter for the Pi coding agent that ranks MCP tools on explicit semantic searches across enabled servers and can evaluate tool results behind an allowlist.Coding & developer toolsMore projects & source code
Pi MCP Adapter Jev search — repogithub.com1,570Optional Jev layer in a token-efficient MCP adapter for the Pi coding agent that ranks MCP tools on explicit semantic searches across enabled servers and can evaluate tool results behind an allowlist.Coding & developer toolsMore projects & source code
pi-automode — prgithub.com—Adds a Jev classifier backend to pi-automode, a Claude Code-style auto mode for the Pi coding agent that decides which actions can run without asking.Coding & developer toolsMore projects & source code
pi-automode — repogithub.com—Adds a Jev classifier backend to pi-automode, a Claude Code-style auto mode for the Pi coding agent that decides which actions can run without asking.Coding & developer toolsMore projects & source code
pi-codemcpgithub.com7Sandboxed Code Mode for MCP servers in the Pi coding agent; optional Jev routing replaces search by scoring enabled tools against the subtask intent and picking at most eight with composition guidance.Coding & developer toolsMore projects & source code
pi-ext-model-router — Readmegithub.com0This Pi extension uses TypeSafe Jev to choose a model from the configured model list.Coding & developer toolsMore projects & source code
pi-follow-throughgithub.com1Pi extension that sends a settled run's request, tool calls and final answer to Jev and sends a continuation prompt only when Jev cites an exact evidence line showing the work is incomplete above a threshold.Coding & developer toolsMore projects & source code
pi-guardrailsgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
pi-guardrails — READMEgithub.com0Pi workflow-extension package optionally integrates Jev to review comments, decide when code reviews are needed and detect test-assertion gaps.Coding & developer toolsMore projects & source code
pi-jevgithub.com0Claude-style auto-mode classifier for Pi, powered by TypeSafe JevCoding & developer toolsMore projects & source code
pi-jev-anti-slopgithub.com1TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
pi-jev-anti-slop — READMEgithub.com1This pi extension uses Jev to assess code-review dimensions such as bugs, security, and quality; its verdict is a tool output, not a verified quality guarantee.Coding & developer toolsMore projects & source code
pi-jev-checkpointsgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
pi-jev-checkpoints — READMEgithub.com0The Pi extension checks request clarity before the agent runs; Jev failures fail open.Coding & developer toolsMore projects & source code
pi-jev-codegithub.com0Jev coprocessor for Pi coding sessions that keeps your Pi model as the only coding agent and adds semantic gates, baseline-to-current diff review and append-only telemetry.Coding & developer toolsMore projects & source code
pi-jev-code — READMEgithub.com0This Pi coding-session coprocessor exposes Jev tools for ranking, evidence checks, diffs, and decisions; the normal Pi model remains the coding agent.Coding & developer toolsMore projects & source code
pi-jev-compact — READMEgithub.com0A Pi extension uses Jev to choose which old tool calls and results can be removed from outgoing context; its linked offline contract evaluation uses fixtures and explicitly does not measure Jev accuracy.Coding & developer toolsMore projects & source code
pi-jev-compaction — READMEgithub.com4This Pi extension clears safely superseded reads/test runs directly and asks TypeSafe Jev to score other older tool outputs for compaction.Coding & developer toolsMore projects & source code
pi-jev-compaction-litegithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
pi-jev-compaction-lite — READMEgithub.com0Experimental Pi extension uses Jev relevance judgments to decide which tool-call text and exact results to retain during context compaction.Coding & developer toolsMore projects & source code
pi-jev-context-curator — READMEgithub.com5A pi extension using TypeSafe Jev to keep or remove context units before LLM calls, preserving tool-call/result pairs and grading only new units after a checkpoint.Coding & developer toolsMore projects & source code
pi-jev-eyegithub.com1Ultra-lean supervisor for Pi coding agent using local regex filters, verification state tracking, and TypeSafe Jev semantic gateCoding & developer toolsMore projects & source code
pi-jev-eye — READMEgithub.com1A Pi supervisor adds TypeSafe Jev semantic checks for larger diffs alongside deterministic checks and model routing; its documented provider failures are fail-open.Coding & developer toolsMore projects & source code
pi-jev-harnessgithub.com—Pi extension where Jev routes each turn, picks which files to pre-fetch, trims long tool results, detects loops and guards tool calls, so the main model spends tokens only on generation.Coding & developer toolsMore projects & source code
pi-jev-routergithub.com—Pi coding-agent extension that asks Jev, before each prompt enters the session, whether to continue, fork, or start a new session so tangents stay out of a long context.Coding & developer toolsMore projects & source code
pi-jev-router — repogithub.com—Pi coding-agent extension that asks Jev, before each prompt enters the session, whether to continue, fork, or start a new session so tangents stay out of a long context.Coding & developer toolsMore projects & source code
pi-jev-scoregithub.com1TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
pi-jev-score — READMEgithub.com1A Pi extension sends each completed AI answer for a Jev score and displays the result in the footer; the README describes the data sent and excluded.Coding & developer toolsMore projects & source code
pi-jev-todo-auditgithub.com3Pi extension that audits rpiv-todo board drift every 10 agent loops using TypeSafe's jev model, injecting corrective nudges when the agent wanders off-taskCoding & developer toolsMore projects & source code
pi-jev-todo-audit — READMEgithub.com3An advisory Pi todo-board auditor using batched Jev Choice judgments against conversation evidence, proposing supported task-state corrections without replacing the worker.Coding & developer toolsMore projects & source code
pi-maestro-flow JEV classifiergithub.com—Multi-agent orchestration for the Pi coding agent with a unified JEV classifier that sorts provider failures into retry kinds and tags handoff files as required, conditional or skip.Coding & developer toolsMore projects & source code
pi-prompt-enhancergithub.com0pi extension that rates every prompt with TypeSafe's jev decision model and injects the missing pieces as hints.Coding & developer toolsMore projects & source code
pi-prompt-enhancer — READMEgithub.com0This Pi extension evaluates missing information while a prompt is being drafted, with jev-latest listed as its model alias; the README describes its behavior, not a measured quality guarantee.Coding & developer toolsMore projects & source code
pi-saver — READMEgithub.com0Pi-Saver groups conversation history into turns and asks Jev which chunks matter for the latest prompt; without a usable key, Jev review is skipped.Coding & developer toolsMore projects & source code
pi-typesafegithub.com0Jev sidecar for the Pi coding agent that runs four hazard checks before bash, write and edit, and four diff-quality checks after edits, in advisory, blocking or shadow mode.Coding & developer toolsMore projects & source code
pi-typesafe — READMEgithub.com0A Pi coding-agent extension uses Jev as an advisory decision sidecar and offers an optional bounded startup selector; advisory mode is the default.Coding & developer toolsMore projects & source code
pi-typesafe-approvegithub.com2Pi extension: auto-approve routine Bash commands with a System One (Jev) decision model, escalate the rest to a humanCoding & developer toolsMore projects & source code
pi-typesafe-approve — READMEgithub.com2A Pi extension using System One/Jev tiers and probabilities to approve routine Bash commands or escalate to a human. It documents TypeSafe’s direct API; approval safety is not certified.Coding & developer toolsMore projects & source code
pkg-gategithub.com0Pre-install security gate for npm that evaluates preinstall, install, and postinstall scripts in parallel for intent, threat severity, secret access, and remote execution before they run.Coding & developer toolsMore projects & source code
pkg-gate — READMEgithub.com0An npm pre-install gate evaluates lifecycle scripts with TypeSafe System One; without an API key, it uses an offline simulator instead.Coding & developer toolsMore projects & source code
plan-driftgithub.com0Plan vs. implementation drift report powered by TypeSafe Jev.Coding & developer toolsMore projects & source code
plan-drift — READMEgithub.com0Claude Code plugin compares implementation diffs against approved plan items using Jev, reporting plan drift and recording human triage without blocking the Stop hook.Coding & developer toolsMore projects & source code
PLang Typesafe decidergithub.com—Natural-language programming language whose builder can let Jev make the module, method and parameter decisions for each step, batching several questions per request.Coding & developer toolsMore projects & source code
plzgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
plz — READMEgithub.com0A Rust daemon checks interactive shell commands using TypeSafe’s jev-latest model.Coding & developer toolsMore projects & source code
pr-scout — READMEgithub.com1PR triage tool uses Jev typed questions to score open pull requests and present take, consider or skip recommendations.Coding & developer toolsMore projects & source code
pr-sievegithub.com1GitHub Action that treats .jev.yml as a semantic linter for pull requests, compiling each rule into a Jev question and failing, commenting, or passing the check from the numbers.Coding & developer toolsMore projects & source code
prompt-oscilloscope — READMEgithub.com1A terminal client and VS Code extension analyze prompts with Jev before forwarding the original text unchanged to GitHub Copilot; ambiguous or risky prompts require confirmation.Coding & developer toolsMore projects & source code
prompt-oscilloscope — repogithub.com1Two experiments: a terminal client that checks prompts with Jev before they reach the GitHub Copilot CLI, and a PR reviewer that uses Jev to trim diffs before a generative review.Coding & developer toolsMore projects & source code
prompt-oscilloscope — repo2github.com1Two experiments: a terminal client that checks prompts with Jev before they reach the GitHub Copilot CLI, and a PR reviewer that uses Jev to trim diffs before a generative review.Coding & developer toolsMore projects & source code
prompt2jevgithub.com6Agent skill and CLI that turn a plain-language requirement, an LLM prompt, or the code that runs one into a Jev decision: typed state, Choice, Score, and Noul questions, and a runnable script.Coding & developer toolsMore projects & source code
prompt2jev — READMEgithub.com6prompt2jev describes converting one of three input types into a Jev request and leaving the decision to application code.Coding & developer toolsMore projects & source code
pytest-jev — READMEgithub.com2pytest-jev adds semantic assertions for an LLM app's output and judges meaning with Jev's probability-returning model rather than matching exact words.Coding & developer toolsMore projects & source code
PZ_Optimization Jev harnessgithub.com—Project Zomboid performance patch project whose benchmark harness uses Jev to judge run uplift and visual parity, and drives the game's Workshop UI via OCR plus Jev in about 1.5 s per step.Coding & developer toolsMore projects & source code
quoin-jevgithub.com—Rust crate in Quoin, a spec-driven development kit for Claude Code, that runs a criterion-strength lens over each functional requirement with one batched Jev request and turns the answers into typed spec-review findings.Coding & developer toolsMore projects & source code
quoin-jev — repogithub.com—Rust crate in Quoin, a spec-driven development kit for Claude Code, that runs a criterion-strength lens over each functional requirement with one batched Jev request and turns the answers into typed spec-review findings.Coding & developer toolsMore projects & source code
R2R + Jevgithub.com1Rust integration that admits two Jev checks per agent tool call, beyond scope and destructive, as evidence into persistent R2R relation state, so trust and authorization degrade until a human override repairs them.Coding & developer toolsMore projects & source code
Radiant decisionsgithub.com—Local Mac coding harness that uses Jev via OpenRouter to pick which MCP servers a message needs, after one Linear server was found adding 69 tool schemas and 16.6k tokens to every call.Coding & developer toolsMore projects & source code
ReadAloud (dasheng)github.com144English read-aloud trainer where the R2T2 streaming ASR transcribes your speech, Jev judges whether each mismatched word was misread and how, and code turns those judgments into a score.Coding & developer toolsMore projects & source code
review-gategithub.com1TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
review-gate — READMEgithub.com1Review Gate asks Jev to decide which kind of reviewer a pull request needs from the diff, returning typed yes/no probabilities.Coding & developer toolsMore projects & source code
RFxCheckgithub.com0Open source Chrome extension for semantic RFx requirement checking, powered by TypeSafe Jev.Coding & developer toolsMore projects & source code
RFxCheck — READMEgithub.com0RFxCheck's README describes using Jev to evaluate meaningful text blocks against seven bounded yes/no questions for RFx review.Coding & developer toolsMore projects & source code
roblox-jev-studio — READMEgithub.com0Roblox Studio plugin asks Jev finite questions about selected Instance metadata and links findings to existing Instance IDs; it does not edit the DataModel.Coding & developer toolsMore projects & source code
Roomote judgment backendgithub.com—Self-hostable cloud coding agent with an optional Jev judgment backend for bounded helper decisions, plus a shadow mode that measures a self-run judgment model's agreement with Jev.Coding & developer toolsMore projects & source code
Roomote judgment backend — docsgithub.com—Self-hostable cloud coding agent with an optional Jev judgment backend for bounded helper decisions, plus a shadow mode that measures a self-run judgment model's agreement with Jev.Coding & developer toolsMore projects & source code
s1s — READMEgithub.com0System One Search combines local code indexing and reference graphs with Jev judgments to find, map, and trace relevant repository files.Coding & developer toolsMore projects & source code
sai jev-audit plugingithub.com—Lua plugin for the sai terminal AI assistant that reviews pending tool actions with a Jev Choice instead of a chat-model reviewer in auto-approval mode.Coding & developer toolsMore projects & source code
sai jev-audit plugin — repogithub.com—Lua plugin for the sai terminal AI assistant that reviews pending tool actions with a Jev Choice instead of a chat-model reviewer in auto-approval mode.Coding & developer toolsMore projects & source code
sathariels_jevtriagegithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
sathariels_jevtriage — READMEgithub.com0A GitHub Action and Python CLI ask Jev a typed Choice question to triage pull requests; a ready verdict exits successfully only above the confidence threshold.Coding & developer toolsMore projects & source code
schema-truthgithub.com0Checks whether the values in a page's JSON-LD (prices, ratings, reviews, authors, dates, FAQ answers, addresses) are shown in the text visitors see, using TypeSafe's Jev model.Coding & developer toolsMore projects & source code
schema-truth — READMEgithub.com0Structured-data audit tool extracts JSON-LD claims and visible page text in code, then asks Jev whether each value is supported, routing uncertain checks for review.Coding & developer toolsMore projects & source code
scrapy-jevgithub.com0A scraped item quality monitor using jev, that samples scraped itemsCoding & developer toolsMore projects & source code
scrapy-jev — READMEgithub.com0A Scrapy pipeline samples selected fields, asks Jev to score their plausibility, and stops a crawl when too few pass; README says it stays inactive until configured.Coding & developer toolsMore projects & source code
security-scangithub.com13Agent skill for read-only security scans (dependency audits, Semgrep, secrets, IaC, GitHub Actions) with an optional Jev step that suggests verdicts for unreviewed findings.Coding & developer toolsMore projects & source code
Seedit Jev helpersgithub.com—Optional Jev developer helpers in the Seedit reddit-alternative repo for browser benchmarking, code review and translation review, run outside the shipped app.Coding & developer toolsMore projects & source code
semantic-assert — READMEgithub.com4semantic-assert defaults to TypeSafe Jev as the judge for plain-English assertions about captured UI or page state.Coding & developer toolsMore projects & source code
semantic-lintgithub.com0TypeSafe Jev ecosystem repository.Coding & developer toolsMore projects & source code
semantic-lint — READMEgithub.com0A policy linter asks Jev one yes/no question per rule and diff hunk, returning probabilities; dry-run mode shows the questions without a key or API call.Coding & developer toolsMore projects & source code
semdecide — READMEgithub.com75SemDecide adds typed TypeSafe Jev decisions to Unix pipelines and CI, wrapping results with predictable input, output, limits, retries, schemas, and exit behavior.Coding & developer toolsMore projects & source code
sensibilitygithub.com0Give Claude Code a second opinion it can ask for in half a second. Reusable "battery" checks plus optional risk and finish gates, powered by TypeSafe's Jev.Coding & developer toolsMore projects & source code
sensibility — READMEgithub.com0Claude Code plugin delegates small review questions about diffs, messages, shell commands and completion claims to Jev for a second opinion.Coding & developer toolsMore projects & source code
Shepherd Jev judgegithub.com—Vendor-neutral Judge seam in Shepherd, self-hosted mission control for interactive Claude Code, backed by the TypeSafe SDK so a fast Jev choice or noul runs on every agent stop instead of spawning a 120 s claude call.Coding & developer toolsMore projects & source code
Shepherd Jev judge — repogithub.com—Vendor-neutral Judge seam in Shepherd, self-hosted mission control for interactive Claude Code, backed by the TypeSafe SDK so a fast Jev choice or noul runs on every agent stop instead of spawning a 120 s claude call.Coding & developer toolsMore projects & source code
shipit-gate — READMEgithub.com0A deploy gate asks Jev whether it is safe to ship before a push; the README says its no-key quick demo uses mocked Jev verdicts.Coding & developer toolsMore projects & source code
should-i-jevgithub.com0Describe a feature, get a verdict on what should actually power it: plain code, Jev, an LLM, Jev + LLM, or classical ML. An executable reading of TypeSafe's own jaggedness list.Coding & developer toolsMore projects & source code
should-i-jev — READMEgithub.com0A feature-fit recommender compares Jev with plain code, LLMs, hybrid Jev-plus-LLM, and classical ML, and can answer that there is not enough to judge.Coding & developer toolsMore projects & source code
Should-I-Jev — READMEgithub.com1should-i-jev scans LLM logs and source for decision-shaped calls that may fit Jev; its optional Jev self-check uses an offline stand-in unless a real endpoint is configured.Coding & developer toolsMore projects & source code
siftrgithub.com2CLI and MCP server giving coding agents four read-only Jev tools for semantic search, focused reads, test-file picking and log filtering, benchmarked on SWE-bench Lite.Coding & developer toolsMore projects & source code
Sim TypeSafe Jev Providergithub.com—Workflow builder with Jev evaluation models for structured agent decisions.Coding & developer toolsMore projects & source code
similarity-ts-jevgithub.com2similarity-ts and fallow duplicate detection for TypeScript, filtered by TypeSafe's Jev down to the pairs worth refactoringCoding & developer toolsMore projects & source code
similarity-ts-jev — READMEgithub.com2similarity-ts-jev uses TypeSafe Jev to judge candidate code-similarity pairs found by similarity-ts and fallow.Coding & developer toolsMore projects & source code
SiYuan TypeSafe Decision Toolgithub.com—Knowledge management system with an agent decision tool backed by TypeSafe System One.Coding & developer toolsMore projects & source code
Skill Dashgithub.com1Local dashboard that has Jev judge each Claude Code or Codex skill for usefulness, redundancy, clarity and safety with a keep, rewrite, merge or delete verdict; also the pipeline behind a census of 18,041 skills.Coding & developer toolsMore projects & source code
skill-dash — READMEgithub.com1A local dashboard uses Jev to judge Claude Code and Codex skills one row at a time.Coding & developer toolsMore projects & source code
slop-gradergithub.com31CLI and agent skill that grades text and Markdown files line by line against rulesets for AI slop, grammar and technical-doc quality with Jev, then hands the flagged violations to an AI agent to fix.Coding & developer toolsMore projects & source code
slop-grader — READMEgithub.com31Slop Grader combines rule-based text-file checks with Jev judgments.Coding & developer toolsMore projects & source code
slop-guard — READMEgithub.com0Slop Guard is a code-review bot where code selects what to inspect, Jev judges focused guideline questions, and explicit rules determine what to report; README examples are not recorded Jev results.Coding & developer toolsMore projects & source code
slopcheckgithub.com—Prose linter, run as a Claude Code Stop hook, that reports which AI writing tells a passage contains with the line quoted; regex settles 18 tells and Jev judges the other 15.Coding & developer toolsMore projects & source code
Snow App decision modelgithub.com—Developer desktop app with AI chat, terminal and SSH that uses Jev to filter search results in its agent-review loop and to decide whether a sensitive command may run without confirmation.Coding & developer toolsMore projects & source code
snow-cli decision modelgithub.com—Adds a TypeSafe Jev decision-model client to snow-cli, a terminal coding agent, used during codebase agent review to judge whether each search result is relevant to the query.Coding & developer toolsMore projects & source code
softlintgithub.com6GitHub Action and CLI that checks each changed piece of a pull request against plain-English rules with Jev and marks the lines that break them; it caught 10/10 violations in the example PR at about $0.001 per PR.Coding & developer toolsMore projects & source code
Specflow TypeSafe advisorygithub.com—Advisory capability in Specflow, a spec-to-contract toolkit for coding agents, that asks Jev whether tests actually cover acceptance criteria and whether evidence supports completion claims during duo-build review.Coding & developer toolsMore projects & source code
Specflow TypeSafe advisory — repogithub.com—Advisory capability in Specflow, a spec-to-contract toolkit for coding agents, that asks Jev whether tests actually cover acceptance criteria and whether evidence supports completion claims during duo-build review.Coding & developer toolsMore projects & source code
SpecPi jev-advisorgithub.com—Opt-in Jev advisor extension in SpecPi, a harness for the Pi coding agent, that judges capabilities, progress, retention, sources and untrusted content, with per-feature switches to apply or only record advice.Coding & developer toolsMore projects & source code
specpi-jev-guardgithub.com9Pi extension that checks risky shell and file commands before the agent runs them: local rules settle obvious cases in 0 ms and Jev scores the rest, blocking high scores and asking on the middle band.Coding & developer toolsMore projects & source code
SpecWeave sw:jevgithub.com—Spec-first AI development toolkit for Claude Code, Codex and Cursor that ships an sw:jev skill for closed-set decisions in about 250 ms via Jev.Coding & developer toolsMore projects & source code
SREGym/SREGymgithub.com≈305Benchmark for incident-resolution agents with optional Jev review of Codex diagnostic tests and submissions, disabled by default.Coding & developer toolsMore projects & source code
Stanleygithub.com118Coding CLI for checking code changes where Jev routes each plain-language request to one deterministic workflow that asks Jev fixed-choice questions about the evidence it gathers, falling back to Pi when none fits.Coding & developer toolsMore projects & source code
stil-lintgithub.com0Style and quality linter for Norwegian and English text - MCP server for agents, with TypeSafe's Jev as the judgment layer. Reports findings, never authorship verdicts.Coding & developer toolsMore projects & source code
stil-lint — READMEgithub.com0The style linter uses Jev as its judgment layer alongside separate checks.Coding & developer toolsMore projects & source code
stingraygithub.com6A Stop hook for Claude Code and Codex, for the turn that ends half-done — nothing done, an announced action never carried out, a promise to watch CI with nothing running — or in the wrong language. Every judgement is made by TypeSafe Jev.Coding & developer toolsMore projects & source code
stingray — READMEgithub.com6Stingray asks Jev to judge certain prematurely ended coding-agent turns, including no-action and broken-promise cases; some checks remain local.Coding & developer toolsMore projects & source code
strapi-plugin-jev-reviewgithub.com0Strapi 5 editorial review and publish guard powered by TypeSafe JevCoding & developer toolsMore projects & source code
strapi-plugin-jev-review — READMEgithub.com0Strapi server plugin uses Jev editorial-review decisions to approve, escalate or revise drafts, with an optional confidence-gated publication guard.Coding & developer toolsMore projects & source code
Supercovgithub.com143CLI 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.Coding & developer toolsMore projects & source code
SuperQode Jev decision harnessgithub.com—Harness layer for coding agents over ACP, A2A and MCP that adds a Jev decision harness for tool routing and tool-permission checks, with benchmarks and a deployable Jev backend.Coding & developer toolsMore projects & source code
SuperQode Jev decision harness — docsgithub.com—Harness layer for coding agents over ACP, A2A and MCP that adds a Jev decision harness for tool routing and tool-permission checks, with benchmarks and a deployable Jev backend.Coding & developer toolsMore projects & source code
Switchloomgithub.com1Workflow prompt for persistent Codex tasks that assigns capabilities to different models with optional Jev routing; the author's own benchmark advises against adopting it.Coding & developer toolsMore projects & source code
Synapse Jev guard seamgithub.com—Build-time guard nodes in the development harness of a Houdini AI assistant: Jev routes each mission to a model tier and screens builder receipts as clear, referee or flag.Coding & developer toolsMore projects & source code
Synapse Jev guard seam — repogithub.com—Build-time guard nodes in the development harness of a Houdini AI assistant: Jev routes each mission to a model tier and screens builder receipts as clear, referee or flag.Coding & developer toolsMore projects & source code
sys1github.com0Sys1 gives coding agents tools to review code, check completion claims, and get structured answers from Jev or a local model.Coding & developer toolsMore projects & source code
sys1 — READMEgithub.com0Agent tools use Jev or local models for code review, completion-claim checks and typed decisions; review and completion features are explicitly advisory.Coding & developer toolsMore projects & source code
SztuCode LLM + Jev agent modegithub.com—Local-first coding and office agent for universities whose experimental mode keeps the LLM for reasoning and asks Jev to choose among alternative next actions at meaningful choice points.Coding & developer toolsMore projects & source code
SztuCode LLM + Jev agent mode — docsgithub.com—Local-first coding and office agent for universities whose experimental mode keeps the LLM for reasoning and asks Jev to choose among alternative next actions at meaningful choice points.Coding & developer toolsMore projects & source code
Taskuary Jev routing judge — repogithub.com—Routing judge in Taskuary, a local-first AI task hub for email, Teams and Slack, where Jev decides after each scheduled report run whether to post it, put it on the work rail, alert the owner or send it out.Coding & developer toolsMore projects & source code
Taste Lintgithub.com7CLI that catches AI slop in UI code, copy and agent instructions before shipping, running mechanical checks locally and adding Jev review notes for semantic taste rules through Vercel AI Gateway.Coding & developer toolsMore projects & source code
taste-lint — READMEgithub.com7Taste-lint sends selected ambiguous design and copy checks to Jev; its README says unlabelled checks remain report-only rather than blocking commits.Coding & developer toolsMore projects & source code
tenet — READMEgithub.com5tenet is an agent-code review gate that uses TypeSafe Jev probabilities to assess each declared rule.Coding & developer toolsMore projects & source code
the-jev-enator — READMEgithub.com0A Claude Code hook project uses Jev for lightweight coding-workflow decisions, as described by its README.Coding & developer toolsMore projects & source code
ThumbGate typed questionsgithub.com—Pre-action firewall for AI coding agents that maps Jev's question format onto its PreToolUse hazard checks (atomic hazard Nouls, a hazard-family Choice, severity), with a --live mode that shadows the battery against the Jev API.Coding & developer toolsMore projects & source code
ThumbGate typed questions — repogithub.com—Pre-action firewall for AI coding agents that maps Jev's question format onto its PreToolUse hazard checks (atomic hazard Nouls, a hazard-family Choice, severity), with a --live mode that shadows the battery against the Jev API.Coding & developer toolsMore projects & source code
Tocketgithub.com38File-first project notebook that multiple coding agents read and write; Jev only picks the next step among options and records a tool_gate allow/block/ask decision before bash, deploy or browser actions.Coding & developer toolsMore projects & source code
tracecheck — READMEgithub.com2A local MCP/CLI review tool asks Jev for typed evidence, impact, and missing-evidence judgments; it neither edits nor executes reviewed code.Coding & developer toolsMore projects & source code
TradingAgentsgithub.com—Multi-agent trading framework that uses Jev to screen StockTwits and Reddit posts for relevance and stance before sentiment analysis.Coding & developer toolsMore projects & source code
trueKeep-jevgithub.com0Claude Code plugin: swaps LLM-summarized /compact for Jev's yes/no calls — nothing kept is ever rewritten, ~20–50x cheaper than routing it through an LLM.Coding & developer toolsMore projects & source code
trueKeep-jev — READMEgithub.com0truekeep uses Jev judgments to prune or shorten old tool-call/result pairs during Claude Code compaction, while retaining user and assistant text verbatim. It classifies what to keep; it does not rewrite the surviving transcript.Coding & developer toolsMore projects & source code
Turbodiff Jev complexity classifiergithub.com—Task-complexity classifier in Turbodiff, an open-source software factory that plans, codes, opens and reviews pull requests, where Jev sets the planning tier for each task.Coding & developer toolsMore projects & source code
Turbodiff Jev complexity classifier — repogithub.com—Task-complexity classifier in Turbodiff, an open-source software factory that plans, codes, opens and reviews pull requests, where Jev sets the planning tier for each task.Coding & developer toolsMore projects & source code
typeful-triagegithub.com4Multiplayer triage dashboard for public GitHub repositories where Jev answers typed questions on each issue, such as kind, severity, urgency, duplicate and next step, and every human correction is fed back on later runs.Coding & developer toolsMore projects & source code
typesafe-bash-guardgithub.com—Pi extension in a personal collection that has Jev classify bash tool calls and ! shell commands before execution and gates potentially harmful ones, with support for 1Password op:// key references.Coding & developer toolsMore projects & source code
typesafe-bash-guard — repogithub.com—Pi extension in a personal collection that has Jev classify bash tool calls and ! shell commands before execution and gates potentially harmful ones, with support for 1Password op:// key references.Coding & developer toolsMore projects & source code
typesafe-cli — READMEgithub.com3A small Python CLI and agent skill sends typed Choice, Score, and Noul judgments to TypeSafe System One (Jev).Coding & developer toolsMore projects & source code
typesafe-cli — READMEgithub.com4A shell CLI sends Jev typed Noul, Choice and Score questions and returns numeric judgments instead of generated prose.Coding & developer toolsMore projects & source code
typesafe-commentgithub.com0Python linter that scores code comments on usefulness, readability, accuracy, redundancy and coverage across Python, C, C++, JavaScript and more, exiting non-zero below thresholds so pipelines block.Coding & developer toolsMore projects & source code
typesafe-comment — READMEgithub.com0typesafe-comment is a CLI that scores code comments on usefulness, readability, accuracy, redundancy, and coverage using TypeSafe System One.Coding & developer toolsMore projects & source code
typesafe-jev-calibrate-for-code-reviewgithub.com3About calibrating Jev for code reviewsCoding & developer toolsMore projects & source code
typesafe-jev-plugingithub.com—CLI & Pipelines: Jev from Typesafe.ai is a "System One" AI model that returns \\typed, calibrated judgments\\ instead of generating text. You define what to classify (a Choice), score (a Score), or verify (a Noul), and Jev returns a structured answer with a probability distribution — fast, cheap, and directly consumable by code.Coding & developer toolsMore projects & source code
typesafe-jev-tools — READMEgithub.com1This Claude Code hook screens edits with local signals and can optionally ask Jev; the README says the API judge is off unless explicitly enabled with a key.Coding & developer toolsMore projects & source code
typesafe-jev-uigithub.com0TypeSafe Jev 调试台，支持 Choice、Score、Noul 和批量处理Coding & developer toolsMore projects & source code
typesafe-jev-ui — READMEgithub.com0React/Express debugging UI sends configurable typed questions to Jev through a server-side TypeSafe System One proxy and displays the structured responses.Coding & developer toolsMore projects & source code
typesafe-migration-guard — READMEgithub.com2Typesafe Migration Guard describes Jev-based evaluation of DDL statements; that description is not a verified safety guarantee.Coding & developer toolsMore projects & source code
typesafe-ui — READMEgithub.com0TypeSafe UI includes a Jev Labs guide with local examples and interface patterns; the README identifies Lab outputs as local fixtures.Coding & developer toolsMore projects & source code
typesafeai-cli — READMEgithub.com2This CLI exposes TypeSafe Jev's noul, choice, and score questions over JSON state for shell or agent use.Coding & developer toolsMore projects & source code
typesafeai-review — READMEgithub.com0A standalone worktree-review CLI asks Jev narrow typed questions per diff hunk, then composes verdicts and findings in code rather than using an LLM agent.Coding & developer toolsMore projects & source code
umby-jev-stack — READMEgithub.com1This HTTP skill tree offers individual TypeSafe Jev agent skills for repository review, diff gates, secrets, tests, safety, and browsing.Coding & developer toolsMore projects & source code
uploads.sh file classifiergithub.com—Experimental classifier in uploads.sh, a file host for coding agents' pull-request artifacts, where Jev picks each upload's kind, surface and screen from closed enums after a vision model describes images.Coding & developer toolsMore projects & source code
uploads.sh file classifier — repogithub.com—Experimental classifier in uploads.sh, a file host for coding agents' pull-request artifacts, where Jev picks each upload's kind, surface and screen from closed enums after a vision model describes images.Coding & developer toolsMore projects & source code
Varin Jev fast decisionsgithub.com—Coding and research agent workspace whose Fast Decision capability calls Jev to judge material relevance, pick the next read, and score context before the main model runs.Coding & developer toolsMore projects & source code
Varin Jev fast decisions — repogithub.com—Coding and research agent workspace whose Fast Decision capability calls Jev to judge material relevance, pick the next read, and score context before the main model runs.Coding & developer toolsMore projects & source code
vibecheck-mcpgithub.com0An MCP server that gives AI coding agents an external "done" criterion — judged by TypeSafe AI's Jev decision model. Agents submit work before declaring completion and get specific, evidence-backed fix lists instead of a vague pass/fail.Coding & developer toolsMore projects & source code
vibecheck-mcp — READMEgithub.com0vibecheck-mcp is an external completion gate for coding agents that asks Jev to assess a task against declared quality dimensions; without a key it uses an offline heuristic, not real review.Coding & developer toolsMore projects & source code
VideoAdGuard-Jevgithub.com—CLI & Pipelines: This project integrates Jev to provide structured decisions for its workflow. See the repository for implementation details.Coding & developer toolsMore projects & source code
wagtail-jevgithub.com1Use jev to classify tags for pagesCoding & developer toolsMore projects & source code
wagtail-jev — READMEgithub.com1The Wagtail integration adds an editor action for Jev to suggest page tags.Coding & developer toolsMore projects & source code
wcode Jev agent-context decisionsgithub.com—Optional Jev decision provider in wcode, an MCP control plane for coding agents, that judges from bounded repository state whether an edit can start, more retrieval or navigation is needed, and which risk to check.Coding & developer toolsMore projects & source code
wcode Jev agent-context decisions — repogithub.com—Optional Jev decision provider in wcode, an MCP control plane for coding agents, that judges from bounded repository state whether an edit can start, more retrieval or navigation is needed, and which risk to check.Coding & developer toolsMore projects & source code
webctlgithub.com148Web search CLI for agents that queries three search backends, has Jev score each result set against the query and goal, and dedupes the high-scoring subset so the agent reads far less.Coding & developer toolsMore projects & source code
webctl — READMEgithub.com148A command-line web-search tool for agents uses Jev; token-saving is the README’s pitch, not an independently measured result.Coding & developer toolsMore projects & source code
wellposed — READMEgithub.com2wellposed has offline structural checks for Jev requests and separate Jev-on-Jev checks for issues that structure alone cannot decide.Coding & developer toolsMore projects & source code
wp-debuglog-triagegithub.com0Group a WordPress debug.log by message, attribute each group to core, a plugin or a theme, and rank the groups by kind and urgency with Jev.Coding & developer toolsMore projects & source code
wp-debuglog-triage — READMEgithub.com0WordPress debug-log CLI groups repeated entries and attributes files in code, then uses Jev to classify problems and urgency, retaining uncertain groups for review.Coding & developer toolsMore projects & source code
wporg-forum-triagegithub.com0Sorts a WordPress.org plugin's support threads into bugs, how-to questions, feature requests, conflicts, praise and spam, and lists the ones waiting for the author's reply, using TypeSafe's Jev model.Coding & developer toolsMore projects & source code
wporg-forum-triage — READMEgithub.com0WordPress.org support-forum triage reads thread titles and first posts, using Jev to classify topics and list those needing replies with uncertain cases separated for review.Coding & developer toolsMore projects & source code
WrongStack Jev decisionsgithub.com≈342Jev integration in the WrongStack coding agent: an agent-callable jev tool for bounded judgments, Jev-backed dispatch among specialist agents, skill suggestion and memory recall, configurable per feature.Coding & developer toolsMore projects & source code
WrongStack Jev decisions — repogithub.com≈342Jev integration in the WrongStack coding agent: an agent-callable jev tool for bounded judgments, Jev-backed dispatch among specialist agents, skill suggestion and memory recall, configurable per feature.Coding & developer toolsMore projects & source code
x-reply-shieldgithub.com0Chrome / Edge Manifest V3 扩展，使用 TypeSafe 判断 X（推特）回复中的色情内容和色情引流，命中后折叠，支持点击恢复。Coding & developer toolsMore projects & source code
x-reply-shield — READMEgithub.com0This Manifest V3 extension uses TypeSafe for X reply content checks; its README says Jev is text-only, so this is not visual moderation.Coding & developer toolsMore projects & source code
x-scannergithub.com15Chrome extension that sends each X post nearing the viewport to Jev with six typed questions and shows behavioral labels in a chip under the post in about 150 ms, with a live cost counter (about 80 posts for $0.0027).Coding & developer toolsMore projects & source code
Xal TypeSafe plugingithub.com—Built-in TypeSafe decision provider for the Xal terminal coding harness; one switch turns on Jev-driven context compaction, Jev read-ahead prefetching and a general-purpose classify tool.Coding & developer toolsMore projects & source code
Xal TypeSafe plugin — repogithub.com—Built-in TypeSafe decision provider for the Xal terminal coding harness; one switch turns on Jev-driven context compaction, Jev read-ahead prefetching and a general-purpose classify tool.Coding & developer toolsMore projects & source code
XM8M acceptance advisorgithub.com—Acceptance ledger for the XM8M PC-8801 emulator where Jev classifies an operator's test report as consistent, problem, not run or insufficient evidence; the advice never decides acceptance.Coding & developer toolsMore projects & source code
XM8M acceptance advisor — repogithub.com—Acceptance ledger for the XM8M PC-8801 emulator where Jev classifies an operator's test report as consistent, problem, not run or insufficient evidence; the advice never decides acceptance.Coding & developer toolsMore projects & source code
XYZ Forge jev_triagegithub.com—Triage classifier in XYZ Forge, an operations system for running coding agents as a workforce, that asks Jev for the status, severity and failure category of agent test runs from exit codes and stderr tails.Coding & developer toolsMore projects & source code
XYZ Forge jev_triage — repogithub.com—Triage classifier in XYZ Forge, an operations system for running coding agents as a workforce, that asks Jev for the status, severity and failure category of agent test runs from exit codes and stderr tails.Coding & developer toolsMore projects & source code
Yao Agentsgithub.com—Yao Agents connects its agent runtime to TypeSafe Jev through Tao for typed classification and scoring.Coding & developer toolsMore projects & source code
zcode-gatekeepergithub.com7External tool-call approver for the ZCode coding agent, modeled on Claude Code auto mode: task-aware review of shell, file and MCP actions, a deterministic blacklist for destructive commands, and an optional Jev backend.Coding & developer toolsMore projects & source code
zttp invariant advisegithub.com—Option in zttp, a Zig TypeScript runtime and agent compiler, where zttp invariant author --advise asks Jev whether a plain-language statement matches the selected invariant kind; the answer is advisory, never proof.Coding & developer toolsMore projects & source code
zttp invariant advise — repogithub.com—Option in zttp, a Zig TypeScript runtime and agent compiler, where zttp invariant author --advise asks Jev whether a plain-language statement matches the selected invariant kind; the answer is advisory, never proof.Coding & developer toolsMore projects & source code
← Back to Awesome Jevgithub.com24Listed by the source without a separate description.Coding & developer toolsMore projects & source code
More posts & discussions42 of 42 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
Abide — demox.com—Hook plugin for Claude Code, Codex and OpenCode in which Jev checks each agent edit against your AGENTS.md and CLAUDE.md rules, about 300 ms per check, and makes the agent fix violations.Coding & developer toolsMore posts & discussions
agent-dispatcher — demowww.reddit.com—Claude Code and Codex dispatcher that routes each request to one of 27 specialist roles with matching skills, tools and a bounded context packet; an opt-in Jev engine can select roles, skills and tools from the catalog.Coding & developer toolsMore posts & discussions
Aition file filterx.com—Code-review tool Aition puts Jev in front of Claude to filter candidate files: 33% fewer files sent, 25% fewer tokens, 23% lower cost, $0.037 to score 543 files (0.38s/file) and no loss in recall.Coding & developer toolsMore posts & discussions
Android e2e testing with Jevx.com—End-to-end test on a real Android device: from the same prompt and 15 test steps, Jev navigated Wikipedia about 14.8x faster than DeepSeek with vision.Coding & developer toolsMore posts & discussions
Azdaja — demowww.reddit.com—Recursive language model layer for Claude Code, Codex, Gemini CLI, OpenCode, and Jcode that keeps full source in a local Python evaluator, with optional Jev judgments and budgeted batches.Coding & developer toolsMore posts & discussions
Chat Seek — demox.com—VS Code extension that finds past Claude Code, Codex and OpenCode conversations from a plain-language description, reranking local matches with the open Laya decision model.Coding & developer toolsMore posts & discussions
Claude Code prompt scorerx.com—Claude Code plugin that uses Jev to score how well you prompt your coding agent, without adding latency.Coding & developer toolsMore posts & discussions
Codebase complexity classifierx.com—Classifier that has Jev judge code in a codebase for complexity, aimed at catching the overengineered code agents tend to produce.Coding & developer toolsMore posts & discussions
Hey Jev, should I deploy? — discussionwww.reddit.com—Deployment vibe check: describe your deploy in plain English and Jev answers 18 typed questions in one pass, which the site turns into a ship-or-wait verdict.Coding & developer toolsMore posts & discussions
Hey Jev, should I deploy? — redditwww.reddit.com—Deployment vibe check: describe your deploy in plain English and Jev answers 18 typed questions in one pass, which the site turns into a ship-or-wait verdict.Coding & developer toolsMore posts & discussions
Highlighter — demox.com—Syntax highlighter where code is split into tokens and Jev decides the language, what every token means, and which of nine lint rules it breaks.Coding & developer toolsMore posts & discussions
Instant variable generationx.com—Demo of Jev producing variables almost instantly, from a series of Jev developer-tool experiments.Coding & developer toolsMore posts & discussions
jeff — discussionnews.ycombinator.com—Read-only Go CLI that checks files against your own rules, such as hidden side effects or weak error handling, with Jev judging each file semantically, locally or in CI.Coding & developer toolsMore posts & discussions
jev (Claude Code plugin) — discussionwww.reddit.com—Go Claude Code plugin that takes over searches and large file reads, using Jev to find relevant files and answer bounded questions about code so whole files never enter the agent's context.Coding & developer toolsMore posts & discussions
JEV OAS Sentinel — demowww.reddit.com—GitHub Action and CLI that compares two OpenAPI documents with deterministic structural checks plus semantic review of changed descriptions, defaults, retries, pagination, auth, and error behavior.Coding & developer toolsMore posts & discussions
Jev QA testerx.com—Demo of Jev acting as a QA tester, driving an iOS app in the simulator through test steps from a command-line runner.Coding & developer toolsMore posts & discussions
Jev skill suggestion for Claude Code — demox.com—Hides the skill roster and lets Jev attach at most one skill to each prompt.Coding & developer toolsMore posts & discussions
Jev skill suggestion for Claude Code — demo 2x.com—Hides the skill roster and lets Jev attach at most one skill to each prompt.Coding & developer toolsMore posts & discussions
Jev UX linter extensionx.com—Japanese Chrome extension that evaluates a site's UX by sending Jev several parallel questions per component, such as whether the result of clicking it is predictable, as a UX linter.Coding & developer toolsMore posts & discussions
jev-auto-approve — demonews.ycombinator.com—GitHub Action that asks whether a pull request needs a human reviewer and approves it only when every question confidently clears your threshold, naming the question that held back a skip.Coding & developer toolsMore posts & discussions
jev-code — discussionwww.reddit.com—Coding CLI and TypeScript SDK where Jev writes Python or Bash one grammar rule at a time by choosing AST productions, then reads back, fixes, runs and reports the program.Coding & developer toolsMore posts & discussions
jev-leftpad — demonews.ycombinator.com—Joke npm package that left-pads strings by asking Jev to choose between space_0 and space_10 options, one API request per call, so it can add at most 10 spaces and costs more than padStart().Coding & developer toolsMore posts & discussions
jev-lint — discussionnews.ycombinator.com—Linter that checks JavaScript and TypeScript against project conventions written in plain English, inline or in Markdown, with Jev reporting possible violations in compact output built for coding agents.Coding & developer toolsMore posts & discussions
jev-mcp — demox.com—Local stdio MCP server that puts Jev in the coding loop of Cursor, Codex and other MCP clients, with tools that route the next step, select prepared tool calls, review patches, verify claims and screen content.Coding & developer toolsMore posts & discussions
jevopt — discussionnews.ycombinator.com—C/C++ compiler driver that pairs Clang's optimizer with Jev, asking at every discretionary call site whether to inline based on the LLVM IR, original source, and build context.Coding & developer toolsMore posts & discussions
JevRLSx.com—Web app where you paste Supabase Row Level Security policies and watch Jev race GPT and Gemini to flag leaks under the same rubric and decision rule, comparing speed and cost side by side.Coding & developer toolsMore posts & discussions
Jevvy — demowww.reddit.com—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.Coding & developer toolsMore posts & discussions
Jevvy — discussionwww.reddit.com—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.Coding & developer toolsMore posts & discussions
Magic Jev Ballx.com—Joke button for GitHub pull requests that checks CI, diff size and reviews, then lets Jev answer 'should I approve this?' in about 200 ms.Coding & developer toolsMore posts & discussions
Nitro — write-upx.com—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.Coding & developer toolsMore posts & discussions
OpenCode browser use with Jev — demox.com—Preview of fast browser automation for app testing that pairs Jev with OpenCode's browser-use CLI.Coding & developer toolsMore posts & discussions
perch — demox.com—Semantic code scanner that parses methods and their call graph, asks Jev typed questions about each method in scope and ranks likely defects and security issues, with custom YAML rules, agent skills and CI gating.Coding & developer toolsMore posts & discussions
pi-automodewww.reddit.com—Adds a Jev classifier backend to pi-automode, a Claude Code-style auto mode for the Pi coding agent that decides which actions can run without asking.Coding & developer toolsMore posts & discussions
prompt-oscilloscopewww.reddit.com—Two experiments: a terminal client that checks prompts with Jev before they reach the GitHub Copilot CLI, and a PR reviewer that uses Jev to trim diffs before a generative review.Coding & developer toolsMore posts & discussions
Self-filling agent prompt boxx.com—Prompt box that lets Jev pick the coding agent, model, machine and folder for each request, such as Fable with Claude Code for a major rewrite or a Mac for iOS changes.Coding & developer toolsMore posts & discussions
Shipper Website to Appx.com—Website-to-app feature in Shipper where Jev decides how to rebuild a pasted website as a native mobile app, which Shipper then submits to the app stores.Coding & developer toolsMore posts & discussions
skill-router — demox.com—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.Coding & developer toolsMore posts & discussions
specpi-jev-guard — discussionwww.reddit.com—Pi extension that checks risky shell and file commands before the agent runs them: local rules settle obvious cases in 0 ms and Jev scores the rest, blocking high scores and asking on the middle band.Coding & developer toolsMore posts & discussions
Subconscious checks for a coding agentwww.reddit.com—Coding-agent helper where Jev collapses passing test logs, blocks destructive shell commands in about 75 ms, and routes note searches, for under half a cent a day.Coding & developer toolsMore posts & discussions
Supabase RLS linterx.com—Experimental linter that has Jev review Supabase row-level-security policies and flag problems, with a hosted tester to try it.Coding & developer toolsMore posts & discussions
Supercov — discussionwww.reddit.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.Coding & developer toolsMore posts & discussions
Supercov — discussion 2www.reddit.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.Coding & developer toolsMore posts & discussions
More videos & channels1 of 1 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
Jev Realtime Code Checkwww.youtube.com—Short demo of a VS Code and Cursor extension that checks local Git changes against 370 Markdown coding rules with Jev on every save, in under 2 seconds.Coding & developer toolsMore videos & channels
More packages & releases10 of 10 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
jev-axi — npmwww.npmjs.com—Agent-ergonomic CLI that gives coding agents Jev judgments from the shell, including a PreToolUse hook that blocks risky Bash calls, prompt-injection screening of fetched text, log triage and ranking.Coding & developer toolsMore packages & releases
jev-kitwww.npmjs.com—Agent-first CLI for Cursor, Claude Code, Codex and Pi that pipes logs, diffs or files into Jev with Noul, Choice or Score flags and prints JSON for the agent or jq to act on.Coding & developer toolsMore packages & releases
jev-spec — npmwww.npmjs.com—CLI that checks code against the requirements in Markdown specs, asking Jev one question per requirement and failing a pre-commit hook or CI build when the probability falls below your threshold.Coding & developer toolsMore packages & releases
jev-use — npmwww.npmjs.com—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.Coding & developer toolsMore packages & releases
jevgate — npmwww.npmjs.com—Permission hook for Claude Code and Codex that auto-approves routine agent requests in about 300 ms, logs every decision, and passes anything risky or unclear to your usual review.Coding & developer toolsMore packages & releases
jevgrep — pypipypi.org—Streaming grep-by-meaning CLI for logs and other text that asks Jev one yes/no question per line and prints matches about 0.7 s after they are written, around $0.004 per 1,000 lines, with a benchmark against Claude.Coding & developer toolsMore packages & releases
jevtest — npmwww.npmjs.com—Semantic matchers for Vitest and Jest: write the expectation in plain English and Jev's probability for one narrow question decides pass or fail, usually in 150 ms to 1.5 s.Coding & developer toolsMore packages & releases
oh-my-harness — npmwww.npmjs.com—CLI that turns a natural-language description into enforced guardrails (CLAUDE.md, hooks, settings) for Claude Code, Codex and Pi; with a TypeSafe key, Jev picks which guard blocks to enable and how strict to be in one call.Coding & developer toolsMore packages & releases
pi-jev-guardwww.npmjs.com—Pi extension that validates model output with Jev, either on demand through a tool, command and review skill, or automatically via a guarded provider that holds replies until Jev approves.Coding & developer toolsMore packages & releases
pi-jev-router — npmwww.npmjs.com—Pi coding-agent extension that asks Jev, before each prompt enters the session, whether to continue, fork, or start a new session so tangents stay out of a long context.Coding & developer toolsMore packages & releases
More guides & websites66 of 66 matches
Resources, repository stars, descriptions and categories
ResourceStarsDescriptionCategorySave
5chan Jev dev helpers — app5chan.app—Optional Node scripts for a peer-to-peer imageboard that use Jev for browser test planning and benchmarks, blind review handoffs and contrastive translation evaluation.Coding & developer toolsMore guides & websites
agmsg jev tool — demowww.producthunt.com—External tool driver in agmsg, a cross-vendor messaging layer for CLI coding agents, that lets a seat ask Jev typed questions about a state as a decision aid.Coding & developer toolsMore guides & websites
aidevops — appwww.aidevops.sh—OpenCode plugin and AI DevOps framework that ships opt-in Jev typed decisions: marketing decision batches, retrieval and shadow-triage pilots, and private evaluation reports.Coding & developer toolsMore guides & websites
AIOS judgment gate — appcli.rexai.top—Local-first control plane for long-horizon coding agents with an opt-in System One judgment gate, exposed as a CLI and MCP tool, that stays off until explicitly enabled and fails closed.Coding & developer toolsMore guides & websites
Aition file filter — appaition.app—Code-review tool Aition puts Jev in front of Claude to filter candidate files: 33% fewer files sent, 25% fewer tokens, 23% lower cost, $0.037 to score 543 files (0.38s/file) and no loss in recall.Coding & developer toolsMore guides & websites
Albatross — appalbatross.sh—Terminal coding agent that routes bounded questions through Jev before the main LLM, returning fixed direct answers only when confident, with off, shadow and active modes.Coding & developer toolsMore guides & websites
Atomic Jev decision backend — websitebastani.ai—Optional Jev backend in the Atomic coding-agent runtime for schema-validated structured decisions, used by default to pick workflows and subagent models when TypeSafe credentials exist, with chat-model fallback.Coding & developer toolsMore guides & websites
Azdaja — siteazdaja.dev—Recursive language model layer for Claude Code, Codex, Gemini CLI, OpenCode, and Jcode that keeps full source in a local Python evaluator, with optional Jev judgments and budgeted batches.Coding & developer toolsMore guides & websites
Conflux judge command — articledev.to—Judge-command boundary in Conflux, a spec-driven orchestrator running AI coding agents in parallel worktrees, that asks a jev-cli-compatible command whether changes depend on each other as a non-authoritative observer.Coding & developer toolsMore guides & websites
CyberCode Fast Judgment — docswk42worldworld.github.io—CyberCode, an agent combining Claude Code with Hermes self-evolution, adds a Fast Judgment service backed by Jev for quick classification and browser decisions.Coding & developer toolsMore guides & websites
Decapod — docsdecapodlabs.github.io—Repo-native governance kernel for AI coding agents with an optional Jev provider that records whether an agent's trajectory satisfies the stated intent as an observation, never as an approval.Coding & developer toolsMore guides & websites
Dev Command Center decision provider — appwww.devcommandcenter.com—Local-first desktop workbench for coding agents with an opt-in decision provider that asks Jev to judge which retrieved ai-memory candidates are relevant.Coding & developer toolsMore guides & websites
dure jev — appwww.dureai.dev—CLI command in Dure, an open-source workspace coordinating Claude Code, Codex and Pi agents across projects, that evaluates a JSON request of typed Noul, Choice and Score questions against TypeSafe Jev from any machine.Coding & developer toolsMore guides & websites
evot — appevot.ai—Lightweight terminal coding agent whose Jev prune step deletes stale context instead of summarising it, asking of every old tool call whether the task still depends on it.Coding & developer toolsMore guides & websites
Feather Wand — appjmeter.ai—AI agent plugin for Apache JMeter whose optional Jev Smart Routing classifies each request's intent to give Agent Mode a focused tool pack, with a visible route card and full-tool fallback.Coding & developer toolsMore guides & websites
Fleet auto-mode Jev classifier — appkhang859.github.io—Option in Fleet, a desktop terminal multiplexer for running several AI coding agents, to use Jev 1.13 as the auto-mode classifier model that answers yes or no about each agent command instead of a text model.Coding & developer toolsMore guides & websites
Formatho Jev Playgroundwww.formatho.com—Client-side request builder for the System One API that composes a state plus Noul, Choice and Score questions, previews mock distributions and generates typesafe_sdk Python code.Coding & developer toolsMore guides & websites
get-fable Reflex (Fable-Jev) — appget-fable.vercel.app—Reflex layer in get-fable, a skills-and-harness pack for existing coding agents, that uses Jev for semantic task routing with probability margins and for verbatim transcript compaction that drops obsolete tool results.Coding & developer toolsMore guides & websites
Glowbom OSS Jev tool — appglowbom.com—Optional tool in a Project Book coding-agent workflow that lets the agent ask Jev small decisions, such as whether build and test results look healthy or need another fix, via OpenCode.Coding & developer toolsMore guides & websites
Goddard Jev evaluation features — appgoddardai.org—Evaluation-model client in Goddard, a Rust GUI for orchestrating Codex, Claude, OpenCode and other coding agents, that uses Jev for per-turn status markers and experimental next-action prediction chips.Coding & developer toolsMore guides & websites
herdr-mcp — appwhshang.github.io—MCP/OAuth edge that lets ChatGPT and other web planners run code, Git and tests on local machines through Herdr, using Jev routes first for post-turn semantic judgments before LLM and script fallbacks.Coding & developer toolsMore guides & websites
Hey Jev, should I deploy?heyjev.ai—Deployment vibe check: describe your deploy in plain English and Jev answers 18 typed questions in one pass, which the site turns into a ship-or-wait verdict.Coding & developer toolsMore guides & websites
Highlighterlab.saeed.sh—Syntax highlighter where code is split into tokens and Jev decides the language, what every token means, and which of nine lint rules it breaks.Coding & developer toolsMore guides & websites
IssueCrush — appgray-bush-0c5cb190f.2.azurestaticapps.net—Tinder-style app for swiping through your GitHub issues to close or keep them; optional structured triage asks Jev for a recommended action, staleness and effort scores and an actionable probability for each card.Coding & developer toolsMore guides & websites
jcode Jev memory and browser handoff — appjcode.sh—The jcode coding-agent harness uses Jev for memory recall relevance and for a fast browser handoff, where Jev picks each offered action and hands back to the main LLM for code or text.Coding & developer toolsMore guides & websites
Jev Issue Radar — apppatrick-sch03.github.io—Read-only dashboard for GitHub issue triage that retrieves candidate pairs and shows side-by-side evidence for whether two reports are duplicate, related, distinct, or under-documented.Coding & developer toolsMore guides & websites
Jev Logsjevlogs.com—Open-source Jev log triage for OpenTelemetry. Score the signal before expensive LLM analysis.Coding & developer toolsMore guides & websites
Jev Realtime Code Check — articlepatrickdesjardins.com—Short demo of a VS Code and Cursor extension that checks local Git changes against 370 Markdown coding rules with Jev on every save, in under 2 seconds.Coding & developer toolsMore guides & websites
jev-spec — writeupkoborin.ai—CLI that checks code against the requirements in Markdown specs, asking Jev one question per requirement and failing a pre-commit hook or CI build when the probability falls below your threshold.Coding & developer toolsMore guides & websites
jevcumber — appjevcumber.dev—Cucumber runner that needs only the .feature file: for each Gherkin step Jev picks the action, control and literal value from what is on the page, Playwright performs it, and it refuses to guess when unsure.Coding & developer toolsMore guides & websites
jevgatejevgate.dev—Permission hook for Claude Code and Codex that auto-approves routine agent requests in about 300 ms, logs every decision, and passes anything risky or unclear to your usual review.Coding & developer toolsMore guides & websites
jevopt — siteramneet-singh.github.io—C/C++ compiler driver that pairs Clang's optimizer with Jev, asking at every discretionary call site whether to inline based on the LLVM IR, original source, and build context.Coding & developer toolsMore guides & websites
JevRLS — appjevrls.lovable.app—Web app where you paste Supabase Row Level Security policies and watch Jev race GPT and Gemini to flag leaks under the same rubric and decision rule, comparing speed and cost side by side.Coding & developer toolsMore guides & websites
JevTest — appjevtest.dev—TypeScript library and CLI for exploratory browser testing: Jev chooses actions toward a goal and judges results, Playwright executes them in isolated sessions, and exact assertions decide pass or fail.Coding & developer toolsMore guides & websites
Kiro Crew Jev decision points — appkiro.dev—Decision points in the Kiro Crew development workspace where Jev Choices pick skills, route models, rate tool risk, steer messages, recall memory and decide what to keep during compaction.Coding & developer toolsMore guides & websites
KnoxCoder Jev layer — appknox.chat—Jev module in the built-in coding agent of KnoxCoder, a VS Code-based editor: it judges each agent turn, gates tool calls and context passages, detects semantic doom loops, rescores messages for compaction and scores agent traces.Coding & developer toolsMore guides & websites
LeanKG judge — appwww.leankg.com—Code knowledge graph for AI coding agents that adds a System One judgment layer, calling Jev or a self-hosted Laya server, including a conversation classification path.Coding & developer toolsMore guides & websites
Learn Ukrainian TypeSafe tooling — applearn-ukrainian.github.io—Agent-fleet tooling in a free A1-C2 Ukrainian course repo that uses Jev for pre-dispatch readiness triage of issues, line-level semantic find and skill suggestions.Coding & developer toolsMore guides & websites
Lockstep — appwww.getlockstep.dev—Shared decision record for a team's coding agents that checks whether changed code may contradict accepted decisions, using Jev for the recall and recheck stages of its funnel.Coding & developer toolsMore guides & websites
Marionette Jev turn judgment — appprofessorpalmer.github.io—Opt-in Jev turn judgment in Marionette, a desktop AI coding harness built on the Puppetmaster kernel, where one OpenRouter Decisions call picks which skill body to load instead of token-overlap retrieval.Coding & developer toolsMore guides & websites
mikan Jev harness — appmikan.geminixiang.com—Self-hosted Slack coding agent where call sites needing fast classification, routing or guardrail decisions call Jev directly instead of treating it as a chat provider.Coding & developer toolsMore guides & websites
Mjolnir Jev turn verdicts — appmjolnir.brokk.ai—Meta-harness for Codex, Claude Code and other ACP coding agents that uses Jev to judge whether a turn needs user input or is background work, and to rank help-search entries.Coding & developer toolsMore guides & websites
o8 judgment referee — appo8.run—Judgment layer in o8, a control room for a fleet of coding agents, that asks Jev locked yes/no questions about each diff (docs-only, touches auth, mock data, adds tests) with per-question calibration notes.Coding & developer toolsMore guides & websites
Oh My Pi Jev judgments — appomp.sh—Native Jev judgment backend in the Oh My Pi coding agent: a judge() call in its eval kernel and a jevify keyword that bulk-classifies diffs, logs or search hits so the agent reads only flagged items.Coding & developer toolsMore guides & websites
oh-my-claudecode Jev judgment points — appoh-my-claudecode.dev—Judgment points in the oh-my-claudecode orchestration hooks where Jev decides skill triggers, loop continuation, model tier for delegated tasks, context staleness and task size.Coding & developer toolsMore guides & websites
OrchestKit Jev session categories — apporchestkit.yonyon.ai—Held-out eval for OrchestKit's opt-in Jev session classifier for Claude Code: on 150 sessions Jev matched Haiku at 80.0% vs 79.3% with the same criteria, about 58x faster with about 34x fewer input tokens.Coding & developer toolsMore guides & websites
Pattern MCP — appusepattern.sh—MCP server that checks UI component libraries against product requirements before a coding agent builds, with an experimental Jev scorer that rates each Figma component page against the checklist.Coding & developer toolsMore guides & websites
perch — siteperchscan.com—Semantic code scanner that parses methods and their call graph, asks Jev typed questions about each method in scope and ranks likely defects and security issues, with custom YAML rules, agent skills and CI gating.Coding & developer toolsMore guides & websites
pi-maestro-flow JEV classifier — appcatlog22.github.io—Multi-agent orchestration for the Pi coding agent with a unified JEV classifier that sorts provider failures into retry kinds and tags handoff files as required, conditional or skip.Coding & developer toolsMore guides & websites
pkg-gate — docshemanth.github.io—Pre-install security gate for npm that evaluates preinstall, install, and postinstall scripts in parallel for intent, threat severity, secret access, and remote execution before they run.Coding & developer toolsMore guides & websites
Radiant decisions — apptempletongroup.dev—Local Mac coding harness that uses Jev via OpenRouter to pick which MCP servers a message needs, after one Linear server was found adding 69 tool schemas and 16.6k tokens to every call.Coding & developer toolsMore guides & websites
Roomote judgment backend — approomote.dev—Self-hostable cloud coding agent with an optional Jev judgment backend for bounded helper decisions, plus a shadow mode that measures a self-run judgment model's agreement with Jev.Coding & developer toolsMore guides & websites
Seedit Jev helpers — appseedit.app—Optional Jev developer helpers in the Seedit reddit-alternative repo for browser benchmarking, code review and translation review, run outside the shipped app.Coding & developer toolsMore guides & websites
Shepherd Jev judge — appshepherd.run—Vendor-neutral Judge seam in Shepherd, self-hosted mission control for interactive Claude Code, backed by the TypeSafe SDK so a fast Jev choice or noul runs on every agent stop instead of spawning a 120 s claude call.Coding & developer toolsMore guides & websites
Skill Dash — appwhichskills.dev—Local dashboard that has Jev judge each Claude Code or Codex skill for usefulness, redundancy, clarity and safety with a keep, rewrite, merge or delete verdict; also the pipeline behind a census of 18,041 skills.Coding & developer toolsMore guides & websites
SpecPi jev-advisor — apptannermidd.github.io—Opt-in Jev advisor extension in SpecPi, a harness for the Pi coding agent, that judges capabilities, progress, retention, sources and untrusted content, with per-feature switches to apply or only record advice.Coding & developer toolsMore guides & websites
SpecWeave sw:jev — appspec-weave.com—Spec-first AI development toolkit for Claude Code, Codex and Cursor that ships an sw:jev skill for closed-set decisions in about 250 ms via Jev.Coding & developer toolsMore guides & websites
Supabase RLS linter — apprls-jev-tester.vercel.app—Experimental linter that has Jev review Supabase row-level-security policies and flag problems, with a hosted tester to try it.Coding & developer toolsMore guides & websites
SuperQode Jev decision harness — appsuperqode.dev—Harness layer for coding agents over ACP, A2A and MCP that adds a Jev decision harness for tool routing and tool-permission checks, with benchmarks and a deployable Jev backend.Coding & developer toolsMore guides & websites
Switchloom — siteswitchloom.ai—Workflow prompt for persistent Codex tasks that assigns capabilities to different models with optional Jev routing; the author's own benchmark advises against adopting it.Coding & developer toolsMore guides & websites
Taste Lint — docstaste-lint.blode.md—CLI that catches AI slop in UI code, copy and agent instructions before shipping, running mechanical checks locally and adding Jev review notes for semantic taste rules through Vercel AI Gateway.Coding & developer toolsMore guides & websites
Turbodiff Jev complexity classifier — appturbodiff.dev—Task-complexity classifier in Turbodiff, an open-source software factory that plans, codes, opens and reviews pull requests, where Jev sets the planning tier for each task.Coding & developer toolsMore guides & websites
uploads.sh file classifier — appuploads.sh—Experimental classifier in uploads.sh, a file host for coding agents' pull-request artifacts, where Jev picks each upload's kind, surface and screen from closed enums after a vision model describes images.Coding & developer toolsMore guides & websites
wcode Jev agent-context decisions — appwcode.francis.run—Optional Jev decision provider in wcode, an MCP control plane for coding agents, that judges from bounded repository state whether an edit can start, more retrieval or navigation is needed, and which risk to check.Coding & developer toolsMore guides & websites
WrongStack Jev decisions — appwrongstack.com—Jev integration in the WrongStack coding agent: an agent-callable jev tool for bounded judgments, Jev-backed dispatch among specialist agents, skill suggestion and memory recall, configurable per feature.Coding & developer toolsMore guides & websites
Xal TypeSafe plugin — appxal.sh—Built-in TypeSafe decision provider for the Xal terminal coding harness; one switch turns on Jev-driven context compaction, Jev read-ahead prefetching and a general-purpose classify tool.Coding & developer toolsMore guides & websites

Showing 1,328 of 1,328 resources