THE COMMUNITY RESOURCE
DIRECTORY
Small decisions.
Small decisions.
Endless possibilities.
Find what people build with Jev. Explore tools, libraries, experiments and ideas from across the TypeSafe ecosystem.
11,066unique resources
29clear categories
220focused subcategories
Yourssave a personal reading list
YOUR NEXT BUILD STARTS HERE
Download catalog
Explore the collection
Browse groups
Data, classification & extraction218 of 218 matches
Classification & labeling48 of 48 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| classifier.devgithub.com | 423 | Keyless zero-shot text classification over plain HTTP, a CLI, and an MCP server, answered by Jev, with a smart tier that re-asks a reasoning model when Jev's confidence is below 0.7. | Data, classification & extractionClassification & labeling | |
| classifier.dev — appclassifier.dev | — | Keyless zero-shot text classification over plain HTTP, a CLI, and an MCP server, answered by Jev, with a smart tier that re-asks a reasoning model when Jev's confidence is below 0.7. | Data, classification & extractionClassification & labeling | |
| DocJevgithub.com | 491 | Library, CLI and local app that classifies PDF, DOCX or PPTX files against natural-language category rules or splits a packet into its component documents, using LiteParse text and Jev predictions. | Data, classification & extractionClassification & labeling | |
| DocJev — demox.com | — | Library, CLI and local app that classifies PDF, DOCX or PPTX files against natural-language category rules or splits a packet into its component documents, using LiteParse text and Jev predictions. | Data, classification & extractionClassification & labeling | |
| evaluatorq classify judgesgithub.com | — | Classify judges in orq's evaluatorq Python eval framework that seat typesafe/jev-latest beside prompted LLM judges in an LLM-as-a-jury, answering yes/no, label, scale or pairwise questions through the Orq router's classify endpoint. | Data, classification & extractionClassification & labeling | |
| evaluatorq classify judges — docsorq-ai.github.io | — | Classify judges in orq's evaluatorq Python eval framework that seat typesafe/jev-latest beside prompted LLM judges in an LLM-as-a-jury, answering yes/no, label, scale or pairwise questions through the Orq router's classify endpoint. | Data, classification & extractionClassification & labeling | |
| evaluatorq classify judges — repogithub.com | — | Classify judges in orq's evaluatorq Python eval framework that seat typesafe/jev-latest beside prompted LLM judges in an LLM-as-a-jury, answering yes/no, label, scale or pairwise questions through the Orq router's classify endpoint. | Data, classification & extractionClassification & labeling | |
| firehose-judgegithub.com | — | Watches the live Bluesky firehose and asks Jev eight questions per sampled post, from intent and sarcasm to bot-ness and screen safety, routing uncertain calls to a human lane. | Data, classification & extractionClassification & labeling | |
| firehose-judge — appcloutmetrics.ai | — | Watches the live Bluesky firehose and asks Jev eight questions per sampled post, from intent and sarcasm to bot-ness and screen safety, routing uncertain calls to a human lane. | Data, classification & extractionClassification & labeling | |
| Hard vs soft constraint classifierx.com | — | Small test using Jev to classify whether a written requirement is a hard or soft constraint, answering in under a second. | Data, classification & extractionClassification & labeling | |
| hfjev — apphemanth.github.io | — | Python and JavaScript tool that loads any Hugging Face dataset, adapts an evaluation rubric to its domain, and classifies each row with calibrated probabilities, with streaming, a CLI, and a web studio. | Data, classification & extractionClassification & labeling | |
| JEV Book Tagsgithub.com | — | Library cataloguing: a calibre plugin asks Jev Noul questions about book genres and subjects, applies configurable per-tag probability thresholds, and preserves existing tags while leaving uncertain results for review. | Data, classification & extractionClassification & labeling | |
| Jev Classifierjevclassifier.vercel.app | — | Browser tool that loads a Telegram channel JSON export and labels every post by type, quality, sentiment and reaction tone, with an option to duel Jev against an LLM. | Data, classification & extractionClassification & labeling | |
| Jev Column Race — appjev-column-race.vercel.app | — | Live race labeling 1,000 app reviews in four columns: Jev finished in 4.6 seconds for $0.023 versus 18.8 seconds and $0.158 for Gemini 3.8 Flash, with similar star-rating agreement. | Data, classification & extractionClassification & labeling | |
| Jev Column Race — demox.com | — | Live race labeling 1,000 app reviews in four columns: Jev finished in 4.6 seconds for $0.023 versus 18.8 seconds and $0.158 for Gemini 3.8 Flash, with similar star-rating agreement. | Data, classification & extractionClassification & labeling | |
| jev-align — demox.com | — | Turns human labels into reusable, calibrated judgment functions with active learning and prompt optimization. | Data, classification & extractionClassification & labeling | |
| jev-align — discussionnews.ycombinator.com | — | Turns human labels into reusable, calibrated judgment functions with active learning and prompt optimization. | Data, classification & extractionClassification & labeling | |
| jev-align — pypipypi.org | — | Turns human labels into reusable, calibrated judgment functions with active learning and prompt optimization. | Data, classification & extractionClassification & labeling | |
| jev-classifyx.com | — | Document classification and routing pipeline that pushes 39,700 documents through Jev in under 4 minutes for $1.43 at 96.38% accuracy, about 180 docs/sec aggregate and 485ms p95. | Data, classification & extractionClassification & labeling | |
| jev-classify — repogithub.com | — | Document classification and routing pipeline that pushes 39,700 documents through Jev in under 4 minutes for $1.43 at 96.38% accuracy, about 180 docs/sec aggregate and 485ms p95. | Data, classification & extractionClassification & labeling | |
| jev-sheetsgithub.com | 1 | Google Sheets custom functions (JEV_IF, JEV_PROB, JEV_CHOICE, JEV_SCORE) that classify, tag, and score text per cell with Jev, returning UNSURE when confidence falls below a set minimum. | Data, classification & extractionClassification & labeling | |
| jev-triagegithub.com | — | Active-learning labeling pipeline that accepts high-confidence Jev judgments, queues uncertain rows for a frontier teacher or humans, and logs soft labels for local distillation. | Data, classification & extractionClassification & labeling | |
| jev-ultralightspeedgithub.com | 12 | Packs 32 items into one request for bulk classification and calibrates the confidence cut that sends the least-sure rows to a person, reporting 533 items a second at 89.2 percent agreement with human labels. | Data, classification & extractionClassification & labeling | |
| jevframe — demox.com | — | Python library for pandas and Polars that labels, scores, and classifies DataFrame rows with natural-language questions, returning full probability distributions as ordinary Series with bounded async concurrency. | Data, classification & extractionClassification & labeling | |
| jevframe — pypipypi.org | — | Python library for pandas and Polars that labels, scores, and classifies DataFrame rows with natural-language questions, returning full probability distributions as ordinary Series with bounded async concurrency. | Data, classification & extractionClassification & labeling | |
| LangWatch Instant Evals on Jevgithub.com | — | LangWatch's Instant Evals classifier runs boolean, score and category evaluators over traced text with Jev, one request per text carrying every question, at about 250 ms a call. | Data, classification & extractionClassification & labeling | |
| LangWatch Instant Evals on Jev — applangwatch.ai | — | LangWatch's Instant Evals classifier runs boolean, score and category evaluators over traced text with Jev, one request per text carrying every question, at about 250 ms a call. | Data, classification & extractionClassification & labeling | |
| Latitude Jev preclassifiergithub.com | — | Opt-in Jev preclassifier in Latitude, an observability platform for AI agents, that judges which conversation checks (flaggers) apply to a session and records models, thresholds, latency and selections. | Data, classification & extractionClassification & labeling | |
| Latitude Jev preclassifier — applatitude.so | — | Opt-in Jev preclassifier in Latitude, an observability platform for AI agents, that judges which conversation checks (flaggers) apply to a session and records models, thresholds, latency and selections. | Data, classification & extractionClassification & labeling | |
| NiceEval TypeSafe judgegithub.com | — | Local-first agent eval tool that adds TypeSafe as an explicit judge provider, mapping Jev probabilities into weighted scores and batch classifications. | Data, classification & extractionClassification & labeling | |
| NiceEval TypeSafe judge — appwww.niceeval.com | — | Local-first agent eval tool that adds TypeSafe as an explicit judge provider, mapping Jev probabilities into weighted scores and batch classifications. | Data, classification & extractionClassification & labeling | |
| NiceEval TypeSafe judge — docsgithub.com | — | Local-first agent eval tool that adds TypeSafe as an explicit judge provider, mapping Jev probabilities into weighted scores and batch classifications. | Data, classification & extractionClassification & labeling | |
| OpenTextShield Jev label auditgithub.com | — | Label-audit script for OpenTextShield, an open-source SMS spam and phishing classifier, that asks Jev to flag training rows whose ham/spam/phishing label looks wrong for a person to decide, with answers saved in the repo. | Data, classification & extractionClassification & labeling | |
| OpenTextShield Jev label audit — appots.telecomsxchange.com | — | Label-audit script for OpenTextShield, an open-source SMS spam and phishing classifier, that asks Jev to flag training rows whose ham/spam/phishing label looks wrong for a person to decide, with answers saved in the repo. | Data, classification & extractionClassification & labeling | |
| OpenTextShield Jev label audit — repogithub.com | — | Label-audit script for OpenTextShield, an open-source SMS spam and phishing classifier, that asks Jev to flag training rows whose ham/spam/phishing label looks wrong for a person to decide, with answers saved in the repo. | Data, classification & extractionClassification & labeling | |
| Overmind Jev decision judgesgithub.com | — | Decision layer in Overmind, a platform that turns production agent traces into fine-tuned models, using Jev choices as eval judges (pass/fail/insufficient, claim supported/contradicted) and for semantic dataset checks. | Data, classification & extractionClassification & labeling | |
| Overmind Jev decision judges — repogithub.com | — | Decision layer in Overmind, a platform that turns production agent traces into fine-tuned models, using Jev choices as eval judges (pass/fail/insufficient, claim supported/contradicted) and for semantic dataset checks. | Data, classification & extractionClassification & labeling | |
| Releases.sh marketing classifiergithub.com | — | Filter in the releases.sh changelog registry that runs a Jev Choice on each freshly parsed feed item to separate real product news from case studies, newsletters, event recaps and other marketing, suppressing the latter. | Data, classification & extractionClassification & labeling | |
| Releases.sh marketing classifier — appreleases.sh | — | Filter in the releases.sh changelog registry that runs a Jev Choice on each freshly parsed feed item to separate real product news from case studies, newsletters, event recaps and other marketing, suppressing the latter. | Data, classification & extractionClassification & labeling | |
| Releases.sh marketing classifier — repogithub.com | — | Filter in the releases.sh changelog registry that runs a Jev Choice on each freshly parsed feed item to separate real product news from case studies, newsletters, event recaps and other marketing, suppressing the latter. | Data, classification & extractionClassification & labeling | |
| Sifa SDK Jev organisation intelligencegithub.com | — | Jev subpath of the Sifa SDK for an AT Protocol professional network, defining question builders, taxonomies and thresholds for duplicate-organisation detection and firmographic classification. | Data, classification & extractionClassification & labeling | |
| Sifa SDK Jev organisation intelligence — repogithub.com | — | Jev subpath of the Sifa SDK for an AT Protocol professional network, defining question builders, taxonomies and thresholds for duplicate-organisation detection and firmographic classification. | Data, classification & extractionClassification & labeling | |
| SignalChaingithub.com | — | Data-cleaning framework in which Jev or an LLM only classifies each CSV column into a 13-code field type and each dataset into a scene, while local code does the writes; Jev cost 6.6–12.7× the LLM's tokens here. | Data, classification & extractionClassification & labeling | |
| TC39 Proposal Atlasgithub.com | 1 | Interactive explorer of ECMAScript TC39 proposals enriched with a semantic taxonomy such as adoption path and 7 intent archetypes, refreshed automatically as proposals change. | Data, classification & extractionClassification & labeling | |
| WeChat article classifierx.com | — | Chinese demo that scrapes 148 long-form WeChat official-account articles and has Jev classify them by scenario in under 2 minutes, as a low-cost labeling run. | Data, classification & extractionClassification & labeling | |
| World Monitor Jev headline classifiergithub.com | 87,060 | Headline classifier in a real-time geopolitical news dashboard that asks jev-1.13.0 for a five-level severity and one of 14 topic categories per headline, validating answers and falling back on failure. | Data, classification & extractionClassification & labeling | |
| World Monitor Jev headline classifier — appworldmonitor.app | — | Headline classifier in a real-time geopolitical news dashboard that asks jev-1.13.0 for a five-level severity and one of 14 topic categories per headline, validating answers and falling back on failure. | Data, classification & extractionClassification & labeling | |
| World Monitor Jev headline classifier — repogithub.com | 87,060 | Headline classifier in a real-time geopolitical news dashboard that asks jev-1.13.0 for a five-level severity and one of 14 topic categories per headline, validating answers and falling back on failure. | Data, classification & extractionClassification & labeling |
Structured extraction5 of 5 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| codearia-sievegithub.com | 5 | MCP server and TypeScript library that parses web pages into the state a decision model like Jev needs: dates as dates, numbers with units, chunks sized to fit; in its test Jev verified 11 of 11 extracted facts. | Data, classification & extractionStructured extraction | |
| Extract without guessingopenrouter.ai | — | OpenRouter Labs recipe that extracts fields from invoices, leases, offer letters, and agreements by having Jev pick each value from candidates found in the text, so it cannot invent one: 12 fields in 0.6 s. | Data, classification & extractionStructured extraction | |
| jev-scraper-chrome-extensionx.com | — | Chrome extension that tries to turn web pages into JSON matching a schema using Jev; the author calls it an interesting experiment but ultimately a failure, too ambitious for a classifier. | Data, classification & extractionStructured extraction | |
| jeveryword — appjeveryword.vercel.app | — | Dependency-free JavaScript library for field extraction, PII detection, and exact quotes that numbers the words of a text so Jev can pick them, returning original substrings with offsets and probabilities. | Data, classification & extractionStructured extraction | |
| JevSpangithub.com | — | Information extraction: zero-shot named entity recognition that splits text at punctuation, asks Jev one Choice over every candidate window per entity type, verifies each nominee with a second Choice (the type, none, mixed or partial) and settles its boundary with a third, averaging 73.7 strict F1 across 12 Chinese and English NER benchmarks against 72.1 for direct extraction with Qwen3.8-27B. | Data, classification & extractionStructured extraction |
SQL & databases8 of 8 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| AILIKE for MySQL — discussionnews.ycombinator.com | — | Native MySQL plugin that adds an AILIKE operator and ailike() functions for filtering rows and comparing text columns with natural-language conditions, each judged by Jev. | Data, classification & extractionSQL & databases | |
| Databend Jev UDFsgithub.com | — | Example UDF server for the Databend data warehouse that adds jev, jev_prob, jev_choice, jev_score and jev_eval SQL functions, sending rows to Jev in batches. | Data, classification & extractionSQL & databases | |
| Databend Jev UDFs — repogithub.com | — | Example UDF server for the Databend data warehouse that adds jev, jev_prob, jev_choice, jev_score and jev_eval SQL functions, sending rows to Jev in batches. | Data, classification & extractionSQL & databases | |
| Jev query plannerx.com | — | Postgres query planner built with Jev that, after some tuning, sped up queries on the Join Order Benchmark by 12%. | Data, classification & extractionSQL & databases | |
| jevQL — appjevql.fly.dev | — | Semantic SQL for vanilla Postgres: a psql-style CLI plus Go, TypeScript and Python SDKs that add jev() conditions to queries, run the plain SQL on the server and judge surviving rows with Jev in batches. | Data, classification & extractionSQL & databases | |
| pg-jevgithub.com | 389 | PostgreSQL extension for WHERE jev(t, '...') queries that batches 20 rows per request and reports how accuracy falls with larger batches. | Data, classification & extractionSQL & databases | |
| pg-jev — demox.com | — | PostgreSQL extension for WHERE jev(t, '...') queries that batches 20 rows per request and reports how accuracy falls with larger batches. | Data, classification & extractionSQL & databases | |
| pg-jev — demo 2x.com | — | PostgreSQL extension for WHERE jev(t, '...') queries that batches 20 rows per request and reports how accuracy falls with larger batches. | Data, classification & extractionSQL & databases |
More projects & source code102 of 102 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| Agent to Trust Jev judge cross-checkgithub.com | — | Cross-check in the Agent to Trust agent-credit lab that grades gold-labelled samples with a deterministic grader, an LLM judge and Jev as an independent third judge, reporting accuracy, latency and cost live. | Data, classification & extractionMore projects & source code | |
| Agent to Trust Jev judge cross-check — repogithub.com | — | Cross-check in the Agent to Trust agent-credit lab that grades gold-labelled samples with a deterministic grader, an LLM judge and Jev as an independent third judge, reporting accuracy, latency and cost live. | Data, classification & extractionMore projects & source code | |
| AgentEval decision evalsgithub.com | — | Evaluation toolkit for .NET AI agents that adds Jev as a third evaluator kind, for example a groundedness Noul over query, response and context, via TypeSafe or OpenRouter. | Data, classification & extractionMore projects & source code | |
| AgentEval decision evals — evidencegithub.com | — | Evaluation toolkit for .NET AI agents that adds Jev as a third evaluator kind, for example a groundedness Noul over query, response and context, via TypeSafe or OpenRouter. | Data, classification & extractionMore projects & source code | |
| Airflow LLM branchinggithub.com | ≈46,900 | Picks the next task with a Jev Choice and hands low-confidence cases to a person. | Data, classification & extractionMore projects & source code | |
| Airflow LLM branching — repogithub.com | ≈46,900 | Picks the next task with a Jev Choice and hands low-confidence cases to a person. | Data, classification & extractionMore projects & source code | |
| Aludel Jev typed judgesgithub.com | — | Jev-backed typed_judge assertions in Aludel, a Phoenix-native LLM evaluation workbench for Elixir: generated outputs are judged by Jev and pass/fail and normalized scores come from the typed answer, with a seeded safety-boundary demo. | Data, classification & extractionMore projects & source code | |
| Aludel Jev typed judges — repogithub.com | — | Jev-backed typed_judge assertions in Aludel, a Phoenix-native LLM evaluation workbench for Elixir: generated outputs are judged by Jev and pass/fail and normalized scores come from the typed answer, with a seeded safety-boundary demo. | Data, classification & extractionMore projects & source code | |
| ASIMOV CLI Jev filtergithub.com | — | Jev-backed semantic filter in the ASIMOV command-line interface for the ASIMOV OSINT and AI platform that streams input records through Jev and keeps only those a plain-language rubric describes. | Data, classification & extractionMore projects & source code | |
| ASIMOV CLI Jev filter — repogithub.com | — | Jev-backed semantic filter in the ASIMOV command-line interface for the ASIMOV OSINT and AI platform that streams input records through Jev and keeps only those a plain-language rubric describes. | Data, classification & extractionMore projects & source code | |
| Avalanche classifier stepsgithub.com | 19 | Agentic ETL framework that mixes deterministic Python steps and agent steps in one DAG, with TypeSafe-backed @ava.classifier_step nodes that answer Choice, Noul and Score questions and are inspectable in the operator UI. | Data, classification & extractionMore projects & source code | |
| Avalanche classifier steps — examplegithub.com | 19 | Agentic ETL framework that mixes deterministic Python steps and agent steps in one DAG, with TypeSafe-backed @ava.classifier_step nodes that answer Choice, Noul and Score questions and are inspectable in the operator UI. | Data, classification & extractionMore projects & source code | |
| ChainForge Jev judgegithub.com | — | ChainForge's prompt-evaluation environment adds Jev as a decision judge, querying it separately from text judges and building reliability tables from its stated probabilities against labels. | Data, classification & extractionMore projects & source code | |
| chat2jevgithub.com | — | Model Routing: Convert OpenAI-compatible Chat Completions requests into TypeSafe System One (Jev) \\State / Questions\\, compare generated text with structured judgments, and publish reusable question sets as proxy routes. | Data, classification & extractionMore projects & source code | |
| codex-jev-native-routergithub.com | — | Model Routing: Experimental native Codex Desktop and CLI model routing with Jev and a configurable allowlist | Data, classification & extractionMore projects & source code | |
| cultivar TypeSafe gradergithub.com | 41 | Optional grading backend in Pinecone's agent-skill testing CLI that scores sandboxed agent runs against task criteria with Jev instead of Claude, reported as about 30x cheaper and aimed at CI gates. | Data, classification & extractionMore projects & source code | |
| cultivar TypeSafe grader — docsgithub.com | 41 | Optional grading backend in Pinecone's agent-skill testing CLI that scores sandboxed agent runs against task criteria with Jev instead of Claude, reported as about 30x cheaper and aimed at CI gates. | Data, classification & extractionMore projects & source code | |
| Dagu decision.evaluategithub.com | — | Built-in decision.evaluate action for Dagu workflows that asks Jev choice, score or yes/no questions about shared context and routes the DAG on the typed answers. | Data, classification & extractionMore projects & source code | |
| Dagu decision.evaluate — repogithub.com | — | Built-in decision.evaluate action for Dagu workflows that asks Jev choice, score or yes/no questions about shared context and routes the DAG on the typed answers. | Data, classification & extractionMore projects & source code | |
| dejevugithub.com | — | Model Routing: Jev? Déjà vu. Browser agents that run on instinct, no Jev needed. One look at the page, one call to any open model, one action. Faster than the Jev demo on Google Flights. | Data, classification & extractionMore projects & source code | |
| dinostompgithub.com | 6 | Local-first verification layer for AI evaluations that audits datasets, scorers, runs and claims; for Jev users it tests a question like an if-statement, reporting accuracy, the p(yes) cutoff and calibration. | Data, classification & extractionMore projects & source code | |
| doc-routergithub.com | 26 | Rust library and CLI with Python bindings that decides page by page which PDF pages need OCR, using Jev as the page judge; on 155 pages it billed 87 and ran 1.74x cheaper than OCRing everything. | Data, classification & extractionMore projects & source code | |
| docker-paperless-ai Jev evaluatorgithub.com | — | Evaluation module in docker-paperless-ai, an AI OCR and metadata pipeline for Paperless-ngx archives, that has Jev grade extracted date, correspondent, title and summary for each document. | Data, classification & extractionMore projects & source code | |
| docker-paperless-ai Jev evaluator — repogithub.com | — | Evaluation module in docker-paperless-ai, an AI OCR and metadata pipeline for Paperless-ngx archives, that has Jev grade extracted date, correspondent, title and summary for each document. | Data, classification & extractionMore projects & source code | |
| duckdb-aigithub.com | 12 | DuckDB extension that calls LLMs from SQL for summarizing, classifying, extraction and embeddings, with a TypeSafe Jev provider for native choices, scores and yes/no probabilities. | Data, classification & extractionMore projects & source code | |
| duckdb-ai — docsgithub.com | 12 | DuckDB extension that calls LLMs from SQL for summarizing, classifying, extraction and embeddings, with a TypeSafe Jev provider for native choices, scores and yes/no probabilities. | Data, classification & extractionMore projects & source code | |
| duckdb-jevgithub.com | 28 | DuckDB extension that asks a Jev question about every row in SQL and returns the answer as a real SQL type such as ENUM, numeric, or STRUCT. | Data, classification & extractionMore projects & source code | |
| eval-genius decision-model judge lanegithub.com | — | Lane in the eval-genius agent skill that suggests Jev as a judge for binary or closed-label eval dimensions, with scripts for calibration, confidence routing and cascade cost, kept discovery-only until it clears a kappa-0.8 floor. | Data, classification & extractionMore projects & source code | |
| eval-genius decision-model judge lane — repogithub.com | — | Lane in the eval-genius agent skill that suggests Jev as a judge for binary or closed-label eval dimensions, with scripts for calibration, confidence routing and cascade cost, kept discovery-only until it clears a kappa-0.8 floor. | Data, classification & extractionMore projects & source code | |
| Expanso × Jevgithub.com | — | Log-pipeline demos where Expanso Edge archives routine lines without a model call and Jev answers four questions on the rest (actionable, severity, owning team, recurrence) to page, notify, review or archive. | Data, classification & extractionMore projects & source code | |
| FleetQ decision-model eval harnessgithub.com | 70 | Self-hosted agent orchestration platform with a System One decision driver and a jev:eval harness that scores Jev or LLMs on JSONL datasets for accuracy, calibration, coverage, latency, cost and determinism. | Data, classification & extractionMore projects & source code | |
| Flyte System One examplegithub.com | — | Alternates Jev and an LLM, splitting each task into 11 to 16 atomic questions and routing results to auto, review, or escalate. | Data, classification & extractionMore projects & source code | |
| Ground Zerogithub.com | 1 | Alpha Python eval library that uses Jev to judge model outputs for hallucinations, correctness, and instruction-following drift, and warns users to validate results manually. | Data, classification & extractionMore projects & source code | |
| Hacker News Judgegithub.com | 0 | Hacker News-styled site where Jev reads every comment of the most-discussed threads one by one and reduces each thread to a single verdict, served from Neon Postgres with BM25 full-text search. | Data, classification & extractionMore projects & source code | |
| Harbor rewardkit Jev judgegithub.com | — | Jev judge option in rewardkit, the grading package of the Terminal-Bench team's Harbor eval framework, scoring agent output against binary and rubric criteria with no reasoning text. | Data, classification & extractionMore projects & source code | |
| harness-evals decision metricsgithub.com | — | Optional decision extra for Harness's open-source eval framework for LLM agents that adds TypeSafe Choice, Score and Noul metrics behind a provider abstraction, alongside its correctness, groundedness and safety metrics. | Data, classification & extractionMore projects & source code | |
| harness-evals decision metrics — repogithub.com | — | Optional decision extra for Harness's open-source eval framework for LLM agents that adds TypeSafe Choice, Score and Noul metrics behind a provider abstraction, alongside its correctness, groundedness and safety metrics. | Data, classification & extractionMore projects & source code | |
| hermes-jev-routergithub.com | — | Model Routing: Hermes Agent plugin: TypeSafe Jev model routing + trim-then-compress | Data, classification & extractionMore projects & source code | |
| Jev Arenagithub.com | 79 | Side-by-side arena that labels the same comments with Jev and DeepSeek or another chat model; on 10,000 comments Jev took 203.2 s and $0.84 versus 823.5 s and $1.50, at slightly lower accuracy. | Data, classification & extractionMore projects & source code | |
| Jev CSV Workbenchgithub.com | 0 | Browser-first workbench that parses a CSV in memory and turns each selected row into a Jev state for your own questions, forwarded through a Cloudflare Worker that stores nothing. | Data, classification & extractionMore projects & source code | |
| Jev Data Analysisgithub.com | 0 | Web app where you upload a CSV or paste a public CSV URL, the UI inspects its schema and proposes a dashboard of 2-4 charts, and Jev fills in the insight values. | Data, classification & extractionMore projects & source code | |
| jev demogithub.com | — | Customer-service chatbot routed by Jev that asks every level of its routing tree in one call per turn and hands off to a human on request, frustration or low confidence. | Data, classification & extractionMore projects & source code | |
| Jev Wrappedgithub.com | 3 | Web app that has Jev judge a public Telegram channel's posts from the last twelve months and returns a shareable card on its post mix, ads, clickbait and emotional pressure, about $0.10 per 1,500 posts. | Data, classification & extractionMore projects & source code | |
| jev-cc-codex-routergithub.com | — | Model Routing: Per-turn model routing proxy for Codex: asks Jev which tier each task needs, rewrites the model, retries flaky upstream errors. | Data, classification & extractionMore projects & source code | |
| jev-claude-routergithub.com | — | Model Routing: Model router for Claude Code using Jev | Data, classification & extractionMore projects & source code | |
| jev-curategithub.com | 93 | Rust streaming CLI and Python API that filters synthetic and pretraining datasets in Parquet or JSONL with local prefilters and typed Jev Choice, Score and Noul judgments, keeping or rejecting each record by threshold. | Data, classification & extractionMore projects & source code | |
| jev-for-allgithub.com | — | Model Routing: Jev for every agentic development workflow — the System One decision model wired into whatever harness an agent codes in: OpenCode today, Claude Code and Hermes adapters next. | Data, classification & extractionMore projects & source code | |
| jev-gatewaygithub.com | — | Model Routing: Session-aware OpenAI-compatible model-routing gateway powered by JEV | Data, classification & extractionMore projects & source code | |
| jev-labgithub.com | 1 | Open lab: Jev (TypeSafe System One) routing in front of Claude Code - measured bugs, patch, and a hard fallback with alerts | Data, classification & extractionMore projects & source code | |
| Jev-Model-Router-Claude-Codegithub.com | — | Model Routing: Begleitmaterial zum Video „Jev + Claude Code: 3 Use Cases". | Data, classification & extractionMore projects & source code | |
| jev-opusgithub.com | — | Model Routing: Claude Opus 5.5 with the effort level re-decided every step by the TypeSafe Jev reflex — without breaking the prompt cache. CLI + Claude Code plugin. | Data, classification & extractionMore projects & source code | |
| jev-pilotgithub.com | — | Model Routing: Let Jev steer Claude Code: the right reasoning effort, subagent model and skill for every prompt. A Claude Code plugin powered by TypeSafe's Jev (OpenRouter / TypeSafe). | Data, classification & extractionMore projects & source code | |
| jev-seo by AgriciDanielgithub.com | 409 | Crawls a site, checks it against 52 SEO rules, has Jev judge every page, and writes PDF, spreadsheet, and Markdown reports. | Data, classification & extractionMore projects & source code | |
| jevalgithub.com | 21 | Python CLI that measures how well a classifier's confidence, including Jev's, matches its accuracy on labeled data and sets the cost-optimal human hand-off threshold, outputting an offline HTML report and a YAML config. | Data, classification & extractionMore projects & source code | |
| Jevalsgithub.com | 7 | Research-preview eval framework that grades LLM and agent outputs, including full traces, against rules written in code, using Jev so each verdict carries a confidence; 32 answers cost about $0.0002. | Data, classification & extractionMore projects & source code | |
| jevbusgithub.com | 2 | Rust streaming event bus whose routing, subscription and consumption are decided by a probabilistic judge, with Jev as the reference judge, plain-language subscriptions and replayable, IO-free decision policies. | Data, classification & extractionMore projects & source code | |
| jevernetesgithub.com | 9 | Live Kubernetes log analysis in the terminal or a local dashboard: ask questions to surface matching logs, inspect context, and hand selected evidence to a coding agent. | Data, classification & extractionMore projects & source code | |
| JevLensgithub.com | — | Toolkit that runs Jev questions from YAML over labeled CSV or JSONL, keeps full probability distributions for offline replay, suggests thresholds and ships a Streamlit dashboard and a CI accuracy gate. | Data, classification & extractionMore projects & source code | |
| jlinkgithub.com | 6 | Links records across two datasets from a match rule written in plain English, from Python, the shell, Stata, or R, and reports F1 0.73 against 0.69 for tuned string matching on NBER patent assignees to Compustat. | Data, classification & extractionMore projects & source code | |
| judge-auditgithub.com | 8 | Shadow-mode calibration audits of AI judges against human decisions; on 200 partly adversarial emails Jev scored 95.5% versus 96.5% for Claude Sonnet 4.5, but auto-approved 73% before its first error versus 2%. | Data, classification & extractionMore projects & source code | |
| JudgeJevgithub.com | 1 | A hands-on educational lab for DeepEval + Jev: inspect real recorded judgments, explore release gates and drift playback, and run fresh evaluations on your own workloads. | Data, classification & extractionMore projects & source code | |
| Kitaru TypeSafe evaluatorgithub.com | — | Opt-in evaluator package for a replay-based agent eval platform that judges recorded sessions with Jev, sending one request per session and storing one result per question. | Data, classification & extractionMore projects & source code | |
| laya-jev-labgithub.com | 2 | Independent measurements of Jev versus the open-weight Laya on an M4 Max, where Jev scored 78% and Laya 57% on 40 Chinese support tickets, plus a local-first cascade matching Jev's accuracy at about 1.8x the speed. | Data, classification & extractionMore projects & source code | |
| Lightdash AI decisionsgithub.com | — | Typed Jev decisions inside Lightdash's BI agent for catalog ranking, date-range checks, chart quality, error classification, answer-claim evidence and field recovery. | Data, classification & extractionMore projects & source code | |
| Maple incident triagegithub.com | — | Pre-LLM incident triage in the Maple OpenTelemetry observability platform: one Jev decision gates whether an incident is worth spending a full model investigation on. | Data, classification & extractionMore projects & source code | |
| MonsterMQ topic decisionsgithub.com | — | Industrial IoT MQTT broker that adds topic-triggered decisions: Jev via OpenRouter evaluates current and historical topic values and publishes the answer back to MQTT. | Data, classification & extractionMore projects & source code | |
| MonsterMQ topic decisions — plangithub.com | — | Industrial IoT MQTT broker that adds topic-triggered decisions: Jev via OpenRouter evaluates current and historical topic values and publishes the answer back to MQTT. | Data, classification & extractionMore projects & source code | |
| muse-jev-playbookgithub.com | — | Model Routing: Jev decision layer for Muse: a fast, cheap TypeSafe AI gate before expensive agent work — confidence policy, recipes, reference router, honest measurement. | Data, classification & extractionMore projects & source code | |
| naiasno.bg Jev chat routinggithub.com | — | Bulgarian open-data platform on elections, parliament and budgets whose chat uses Jev to route each question to the right data lane, falling back to Gemini on any failure. | Data, classification & extractionMore projects & source code | |
| naiasno.bg Jev chat routing — docsgithub.com | — | Bulgarian open-data platform on elections, parliament and budgets whose chat uses Jev to route each question to the right data lane, falling back to Gemini on any failure. | Data, classification & extractionMore projects & source code | |
| naiasno.bg Jev chat routing — repogithub.com | — | Bulgarian open-data platform on elections, parliament and budgets whose chat uses Jev to route each question to the right data lane, falling back to Gemini on any failure. | Data, classification & extractionMore projects & source code | |
| Neuronpedia Jev autointerp scorergithub.com | — | Neuronpedia, an open interpretability platform, scores neuron-explanation quality with Jev via detection, fuzzing and a 5-level rating, one request per explanation. | Data, classification & extractionMore projects & source code | |
| NimBus TypeSafe failure intelligencegithub.com | — | Integration platform for .NET on Azure Service Bus whose failure-intelligence extension sends one bounded TypeSafe request per failed message to classify the failure for operators. | Data, classification & extractionMore projects & source code | |
| NimBus TypeSafe failure intelligence — repogithub.com | — | Integration platform for .NET on Azure Service Bus whose failure-intelligence extension sends one bounded TypeSafe request per failed message to classify the failure for operators. | Data, classification & extractionMore projects & source code | |
| OASISgithub.com | 29 | Open-source CLI that benchmarks AI models on offensive-security CTF challenges with MITRE ATT&CK mapping; an opt-in TypeSafe judge re-scores each step's success after the run instead of the default substring regex. | Data, classification & extractionMore projects & source code | |
| omp-plugin-jev-routergithub.com | — | Model Routing: Route Oh My Pi prompts between simple and advanced models with TypeSafe AI's Jev classifier. | Data, classification & extractionMore projects & source code | |
| OpenWebTrack Jev insightsgithub.com | — | Insights feature in OpenWebTrack, a self-hosted open-source web analytics platform, that asks Jev for the dominant traffic trend, the top driver and traffic quality of a period from aggregated stats. | Data, classification & extractionMore projects & source code | |
| OpenWebTrack Jev insights — repogithub.com | — | Insights feature in OpenWebTrack, a self-hosted open-source web analytics platform, that asks Jev for the dominant traffic trend, the top driver and traffic quality of a period from aggregated stats. | Data, classification & extractionMore projects & source code | |
| OpenWork Jev verificationgithub.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. | Data, classification & extractionMore projects & source code | |
| pg_typesafegithub.com | 87 | Pre-alpha PostgreSQL C extension that calls Jev from SQL, returning Choice, Noul, and Score answers for classification and scoring inside queries. | Data, classification & extractionMore projects & source code | |
| Polar Llama TypeSafe supportgithub.com | — | Native Rust TypeSafe layer in the Polars LLM plugin that answers Noul, Choice, and Score questions per row, or applies a Pydantic contract to every line of a document, as typed dataframe columns. | Data, classification & extractionMore projects & source code | |
| SchemeWeaver TypeSafe auto-mappergithub.com | — | Optional companion package for the SchemeWeaver Umbraco JSON-LD plugin that replaces name-matching with Jev judgments when auto-mapping CMS content properties to Schema.org properties, in the UI and via MCP. | Data, classification & extractionMore projects & source code | |
| SchemeWeaver TypeSafe auto-mapper — repogithub.com | — | Optional companion package for the SchemeWeaver Umbraco JSON-LD plugin that replaces name-matching with Jev judgments when auto-mapping CMS content properties to Schema.org properties, in the UI and via MCP. | Data, classification & extractionMore projects & source code | |
| Secondlayer Jev ops gatesgithub.com | — | Ops scripts in the self-hosted Secondlayer Stacks indexer: a Slack gate that pages only when Jev rates page_now at 0.8 or more with severity 3+, and a spike testing Jev triage of decoder failures. | Data, classification & extractionMore projects & source code | |
| Secondlayer Jev ops gates — repogithub.com | — | Ops scripts in the self-hosted Secondlayer Stacks indexer: a Slack gate that pages only when Jev rates page_now at 0.8 or more with severity 3+, and a spike testing Jev triage of decoder failures. | Data, classification & extractionMore projects & source code | |
| Stratumgithub.com | 51 | Log intelligence system for semantic search, anomaly detection and root-cause analysis whose optional Jev step classifies each query's intent, service and severity with confidence. | Data, classification & extractionMore projects & source code | |
| stuntdoublegithub.com | 1 | Drop-in /v1/systemone proxy that shadows Jev with local decision models (Kev, Laya) and reports whether you can swap | Data, classification & extractionMore projects & source code | |
| tocsingithub.com | 2 | Rust log triage at ingest that masks and groups log lines into Drain patterns, asks Jev about each new pattern once, and routes lines to page, ticket or log under a plain-English policy. | Data, classification & extractionMore projects & source code | |
| typesafe (Zig CLI)github.com | — | Zig document-review CLI that sends each text or Markdown file to Jev with configured Score questions and writes JSONL with scores, level probabilities, usage and a needs_review flag. | Data, classification & extractionMore projects & source code | |
| typesafe (Zig CLI) — repogithub.com | — | Zig document-review CLI that sends each text or Markdown file to Jev with configured Score questions and writes JSONL with scores, level probabilities, usage and a needs_review flag. | Data, classification & extractionMore projects & source code | |
| Umwelten Jev judgment backendgithub.com | — | Model evaluation and agent-habitat toolkit with a judgment backend that runs typed questions on Jev through OpenRouter, TypeSafe or the Mycel decisions endpoint, recording each attempt and its cost. | Data, classification & extractionMore projects & source code | |
| Umwelten Jev judgment backend — repogithub.com | — | Model evaluation and agent-habitat toolkit with a judgment backend that runs typed questions on Jev through OpenRouter, TypeSafe or the Mycel decisions endpoint, recording each attempt and its cost. | Data, classification & extractionMore projects & source code | |
| Valcraft Jev gradergithub.com | — | Eval grading script for Valcraft's spec-driven agent skills that asks Jev one Noul per graded assertion over a run's transcript and outputs, writing a jev-grading.json beside the LLM grader's result. | Data, classification & extractionMore projects & source code | |
| Valcraft Jev grader — repogithub.com | — | Eval grading script for Valcraft's spec-driven agent skills that asks Jev one Noul per graded assertion over a run's transcript and outputs, writing a jev-grading.json beside the LLM grader's result. | Data, classification & extractionMore projects & source code | |
| Vane Jev judgmentsgithub.com | — | Multimodal data engine built on a DuckDB fork that adds Jev judgments over Vane expressions, batching rows through the async TypeSafe SDK. | Data, classification & extractionMore projects & source code | |
| Vane Jev judgments — examplegithub.com | — | Multimodal data engine built on a DuckDB fork that adds Jev judgments over Vane expressions, batching rows through the async TypeSafe SDK. | Data, classification & extractionMore projects & source code | |
| vgi-typesafegithub.com | 4 | DuckDB worker, loaded through the VGI extension, that exposes Jev Choice, Noul and Score as SQL table functions you LATERAL join against a table, returning typed columns with confidence and probabilities. | Data, classification & extractionMore projects & source code | |
| waza typesafe-judgegithub.com | — | Grader script in a personal dotfiles repo for the waza skill-eval tool that asks Jev yes/no questions about an agent's answer, optionally against a source text, and passes only when every judgment clears the threshold. | Data, classification & extractionMore projects & source code | |
| waza typesafe-judge — repogithub.com | — | Grader script in a personal dotfiles repo for the waza skill-eval tool that asks Jev yes/no questions about an agent's answer, optionally against a source text, and passes only when every judgment clears the threshold. | Data, classification & extractionMore projects & source code | |
| whileai Jev judge backendgithub.com | — | Agent post-training and eval SDK with a typesafe: judge backend that runs its pointwise, pairwise, rubric and audit judges on Jev instead of a chat model. | Data, classification & extractionMore projects & source code | |
| whileai Jev judge backend — repogithub.com | — | Agent post-training and eval SDK with a typesafe: judge backend that runs its pointwise, pairwise, rubric and audit judges on Jev instead of a chat model. | Data, classification & extractionMore projects & source code | |
| zevals Jev judgegithub.com | 11 | TypeScript library for end-to-end AI agent tests whose assertions can use Jev as the judge, reporting a calibrated probability per assertion at about 0.5 s and $0.00005 per call. | Data, classification & extractionMore projects & source code |
More posts & discussions18 of 18 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| AI news filteringx.com | — | Screened nearly 2,700 AI news items from the past 7 days one by one with Jev in about 2 minutes for $0.21, to pick content topics. | Data, classification & extractionMore posts & discussions | |
| Bannerbear field mappingx.com | — | Live Bannerbear feature that maps template fields to differently named data-source fields (photo to avatar, company_name to business) in one click. | Data, classification & extractionMore posts & discussions | |
| Decision layer for BI data agentsx.com | — | Decision layer that data agents act on for automated BI analysis; in a policy-bound 120-case run Jev took 953 ms with 100% recall versus Luna's 3,383 ms, 90.3% recall and 14 unsafe actions. | Data, classification & extractionMore posts & discussions | |
| Doc-OCR routerx.com | — | Router that looks at a PDF page by page, has Jev decide which pages actually need OCR, and extracts the rest locally, cutting OCR cost and time. | Data, classification & extractionMore posts & discussions | |
| duckdb-jev — demonews.ycombinator.com | — | DuckDB extension that asks a Jev question about every row in SQL and returns the answer as a real SQL type such as ENUM, numeric, or STRUCT. | Data, classification & extractionMore posts & discussions | |
| Flowsery session replay triagex.com | — | Ran Jev over 3 million session-replay events: in 40 seconds it reviewed 3,247 sessions, caught 132 rage clicks, 116 dead clicks and 95 JavaScript errors, and opened 213 draft fix PRs for $2.17. | Data, classification & extractionMore posts & discussions | |
| jeq — discussionwww.reddit.com | — | Go 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. | Data, classification & extractionMore posts & discussions | |
| Jev for proactive chart monitoringx.com | — | Analytics experiment using Jev to flag which charts deserve deeper analysis: on a synthetic benchmark it cost about 1/3 as much and ran 5x faster than the strongest cheap hosted baseline, first in recall but last in precision. | Data, classification & extractionMore posts & discussions | |
| Jev Logs — discussionwww.reddit.com | — | TypeScript library and OpenTelemetry exporter wrapper that scores each log record's diagnostic value, priority and routing with Jev before any expensive LLM analysis, keeping every record in the archive. | Data, classification & extractionMore posts & discussions | |
| jev-cli — demonews.ycombinator.com | — | Command-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. | Data, classification & extractionMore posts & discussions | |
| jevals — demonews.ycombinator.com | — | Framework-agnostic evals and guardrails for agents that send all of a trace's checks to Jev in one request, fast enough for the agent loop, with Kev or Laya locally or a chat-model fallback. | Data, classification & extractionMore posts & discussions | |
| jevals — demox.com | — | Local browser workbench for authoring Jev Noul, Choice and Score questions with example cases and expected answers, running them and comparing saved results. | Data, classification & extractionMore posts & discussions | |
| jevcal — demox.com | — | Toolkit that measures a typed decision model like Jev on your own labeled data against an LLM teacher, picks the confidence threshold for a target accuracy, reports how much traffic still needs an LLM, and fails CI on drift. | Data, classification & extractionMore posts & discussions | |
| jevmetrics — demonews.ycombinator.com | — | Experimental 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. | Data, classification & extractionMore posts & discussions | |
| MotherDuck prompt_jev()x.com | — | SQL function in MotherDuck that runs Jev text classification inside queries, including meaning-based filters in a WHERE clause, reported at 50x the speed and 1% the cost of comparable frontier models. | Data, classification & extractionMore posts & discussions | |
| Physical-AI action label QAx.com | — | Quality checks on egocentric training data for physical AI, where Jev QA'd 58,643 action labels in under 3 minutes for 90 cents. | Data, classification & extractionMore posts & discussions | |
| Real-time decision dashboardx.com | — | Next.js dashboard that processes 50 events per second with Jev, flags what matters, tracks attention areas and surfaces the next best action. | Data, classification & extractionMore posts & discussions | |
| typed_evals — discussionwww.reddit.com | — | Python library and CLI that evaluates LLM responses, RAG datasets, and recorded agent runs with Jev as the judge, guards tools before they execute, and can calibrate metrics against human pass/fail labels. | Data, classification & extractionMore posts & discussions |
More packages & releases3 of 3 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| jev-curate — cratecrates.io | — | Rust streaming CLI and Python API that filters synthetic and pretraining datasets in Parquet or JSONL with local prefilters and typed Jev Choice, Score and Noul judgments, keeping or rejecting each record by threshold. | Data, classification & extractionMore packages & releases | |
| jev-logtriage — pypipypi.org | — | CLI 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. | Data, classification & extractionMore packages & releases | |
| pytest-jev — pypipypi.org | — | Plugin 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. | Data, classification & extractionMore packages & releases |
More guides & websites34 of 34 matches
| Resource | Stars | Description | Category | Save |
|---|---|---|---|---|
| Agent to Trust Jev judge cross-check — appsealit.cc | — | Cross-check in the Agent to Trust agent-credit lab that grades gold-labelled samples with a deterministic grader, an LLM judge and Jev as an independent third judge, reporting accuracy, latency and cost live. | Data, classification & extractionMore guides & websites | |
| AgentEval decision evals — appagenteval.dev | — | Evaluation toolkit for .NET AI agents that adds Jev as a third evaluator kind, for example a groundedness Noul over query, response and context, via TypeSafe or OpenRouter. | Data, classification & extractionMore guides & websites | |
| Aludel Jev typed judges — docshexdocs.pm | — | Jev-backed typed_judge assertions in Aludel, a Phoenix-native LLM evaluation workbench for Elixir: generated outputs are judged by Jev and pass/fail and normalized scores come from the typed answer, with a seeded safety-boundary demo. | Data, classification & extractionMore guides & websites | |
| Calling Jev from Aurora PostgreSQLqiita.com | — | Japanese experiment that calls Jev from Aurora PostgreSQL via Lambda to judge 1,000 Amazon product reviews, measuring time and accuracy when batching 1, 10, 50, or 100 rows per request. | Data, classification & extractionMore guides & websites | |
| ChainForge Jev judge — docschainforge.ai | — | ChainForge's prompt-evaluation environment adds Jev as a decision judge, querying it separately from text judges and building reliability tables from its stated probabilities against labels. | Data, classification & extractionMore guides & websites | |
| dinostomp — appcollapseindex.org | — | Local-first verification layer for AI evaluations that audits datasets, scorers, runs and claims; for Jev users it tests a question like an if-statement, reporting accuracy, the p(yes) cutoff and calibration. | Data, classification & extractionMore guides & websites | |
| duckdb-ai — appleonardovida.github.io | — | DuckDB extension that calls LLMs from SQL for summarizing, classifying, extraction and embeddings, with a TypeSafe Jev provider for native choices, scores and yes/no probabilities. | Data, classification & extractionMore guides & websites | |
| FleetQ decision-model eval harness — appfleetq.net | — | Self-hosted agent orchestration platform with a System One decision driver and a jev:eval harness that scores Jev or LLMs on JSONL datasets for accuracy, calibration, coverage, latency, cost and determinism. | Data, classification & extractionMore guides & websites | |
| Flyte System One example — appflyte.org | — | Alternates Jev and an LLM, splitting each task into 11 to 16 atomic questions and routing results to auto, review, or escalate. | Data, classification & extractionMore guides & websites | |
| Hacker News Judge — apphnjudge.vercel.app | — | Hacker News-styled site where Jev reads every comment of the most-discussed threads one by one and reduces each thread to a single verdict, served from Neon Postgres with BM25 full-text search. | Data, classification & extractionMore guides & websites | |
| Harbor rewardkit Jev judge — appharborframework.com | — | Jev judge option in rewardkit, the grading package of the Terminal-Bench team's Harbor eval framework, scoring agent output against binary and rubric criteria with no reasoning text. | Data, classification & extractionMore guides & websites | |
| Jev Arena — sitenanmicoder.github.io | — | Side-by-side arena that labels the same comments with Jev and DeepSeek or another chat model; on 10,000 comments Jev took 203.2 s and $0.84 versus 823.5 s and $1.50, at slightly lower accuracy. | Data, classification & extractionMore guides & websites | |
| Jev Data Analysis — appjev-gamecast.vercel.app | — | Web app where you upload a CSV or paste a public CSV URL, the UI inspects its schema and proposes a dashboard of 2-4 charts, and Jev fills in the insight values. | Data, classification & extractionMore guides & websites | |
| Jev Logs — apphuggingface.co | — | TypeScript library and OpenTelemetry exporter wrapper that scores each log record's diagnostic value, priority and routing with Jev before any expensive LLM analysis, keeping every record in the archive. | Data, classification & extractionMore guides & websites | |
| Jev Logs — modelhuggingface.co | — | TypeScript library and OpenTelemetry exporter wrapper that scores each log record's diagnostic value, priority and routing with Jev before any expensive LLM analysis, keeping every record in the archive. | Data, classification & extractionMore guides & websites | |
| Jev Logs — model 2huggingface.co | — | TypeScript library and OpenTelemetry exporter wrapper that scores each log record's diagnostic value, priority and routing with Jev before any expensive LLM analysis, keeping every record in the archive. | Data, classification & extractionMore guides & websites | |
| Jev Wrapped — appwrapped.ivanhabor.com | — | Web app that has Jev judge a public Telegram channel's posts from the last twelve months and returns a shareable card on its post mix, ads, clickbait and emotional pressure, about $0.10 per 1,500 posts. | Data, classification & extractionMore guides & websites | |
| jev-curate — appjev-curate.vercel.app | — | Rust streaming CLI and Python API that filters synthetic and pretraining datasets in Parquet or JSONL with local prefilters and typed Jev Choice, Score and Noul judgments, keeping or rejecting each record by threshold. | Data, classification & extractionMore guides & websites | |
| JevScope — appjeiel85.github.io | — | Local-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. | Data, classification & extractionMore guides & websites | |
| Kitaru TypeSafe evaluator — appkitaru.ai | — | Opt-in evaluator package for a replay-based agent eval platform that judges recorded sessions with Jev, sending one request per session and storing one result per question. | Data, classification & extractionMore guides & websites | |
| Lightdash AI decisions — applightdash.com | — | Typed Jev decisions inside Lightdash's BI agent for catalog ranking, date-range checks, chart quality, error classification, answer-claim evidence and field recovery. | Data, classification & extractionMore guides & websites | |
| Maple incident triage — appmaple.dev | — | Pre-LLM incident triage in the Maple OpenTelemetry observability platform: one Jev decision gates whether an incident is worth spending a full model investigation on. | Data, classification & extractionMore guides & websites | |
| MonsterMQ topic decisions — appmonstermq.com | — | Industrial IoT MQTT broker that adds topic-triggered decisions: Jev via OpenRouter evaluates current and historical topic values and publishes the answer back to MQTT. | Data, classification & extractionMore guides & websites | |
| MotherDuck prompt_jev() — appmotherduck.com | — | SQL function in MotherDuck that runs Jev text classification inside queries, including meaning-based filters in a WHERE clause, reported at 50x the speed and 1% the cost of comparable frontier models. | Data, classification & extractionMore guides & websites | |
| Neuronpedia Jev autointerp scorer — appneuronpedia.org | — | Neuronpedia, an open interpretability platform, scores neuron-explanation quality with Jev via detection, fuzzing and a 5-level rating, one request per explanation. | Data, classification & extractionMore guides & websites | |
| OASIS — appoasis.kryptsec.com | — | Open-source CLI that benchmarks AI models on offensive-security CTF challenges with MITRE ATT&CK mapping; an opt-in TypeSafe judge re-scores each step's success after the run instead of the default substring regex. | Data, classification & extractionMore guides & websites | |
| OpenWebTrack Jev insights — appopenwebtrack.one | — | Insights feature in OpenWebTrack, a self-hosted open-source web analytics platform, that asks Jev for the dominant traffic trend, the top driver and traffic quality of a period from aggregated stats. | Data, classification & extractionMore guides & websites | |
| Polar Llama TypeSafe support — sitepnthn.ai | — | Native Rust TypeSafe layer in the Polars LLM plugin that answers Noul, Choice, and Score questions per row, or applies a Pydantic contract to every line of a document, as typed dataframe columns. | Data, classification & extractionMore guides & websites | |
| Secondlayer Jev ops gates — appsecondlayer.tools | — | Ops scripts in the self-hosted Secondlayer Stacks indexer: a Slack gate that pages only when Jev rates page_now at 0.8 or more with severity 3+, and a spike testing Jev triage of decoder failures. | Data, classification & extractionMore guides & websites | |
| Umwelten Jev judgment backend — appumwelten.thefocus.ai | — | Model evaluation and agent-habitat toolkit with a judgment backend that runs typed questions on Jev through OpenRouter, TypeSafe or the Mycel decisions endpoint, recording each attempt and its cost. | Data, classification & extractionMore guides & websites | |
| Vane Jev judgments — appvane.astrovela.ai | — | Multimodal data engine built on a DuckDB fork that adds Jev judgments over Vane expressions, batching rows through the async TypeSafe SDK. | Data, classification & extractionMore guides & websites | |
| vgi-typesafe — sitequery.farm | — | DuckDB worker, loaded through the VGI extension, that exposes Jev Choice, Noul and Score as SQL table functions you LATERAL join against a table, returning typed columns with confidence and probabilities. | Data, classification & extractionMore guides & websites | |
| whileai Jev judge backend — appwithwhile.com | — | Agent post-training and eval SDK with a typesafe: judge backend that runs its pointwise, pairwise, rubric and audit judges on Jev instead of a chat model. | Data, classification & extractionMore guides & websites | |
| YOLO + Jev scene filterhuggingface.co | — | Vision pipeline where YOLO-World proposes open-vocabulary boxes and Jev answers a yes/no per box on whether to keep it, producing fewer, better-filtered detections. | Data, classification & extractionMore guides & websites |
No matches this time
Try a broader search or reset your filters to explore the full collection.
Showing 218 of 218 resources