Jev radar

Useful for the stack

New 2026-09-20

• jev-rerankingBenchmarks Jev zero-shot reranking against monoBERT and BM25 on TREC.evalragsearchgithub★ 742026-09-20
• clean-code-reviewJudges PR files against Clean Code principles using Jev, summarized by Luna.coding-agentevalgithub★ 742026-09-20
• any-autoAuto-approves coding agent tool calls using Jev as a risk-reviewing backend.coding-agentguardrailsgithub★ 742026-09-20
• super-jevExperimental harness connecting evidence, Jev judgments, and verified actions.guardrailssdkgithub★ 722026-09-20
• SemDecideUnix CLI wrapping Jev for classification, scoring, filtering, and guarding.cliclassificationguardrailsx★ 742026-09-20
• TypeSafe on NeonRoutes requests to Grok or GPT based on Jev's task classification.routingclassificationx★ 442026-09-20
• neo4jevScores graph edges with Jev and beam-searches knowledge graph paths.ragdatax★ 4732026-09-20
• PrismJev judges liquidity toxic flow and market stress in shadow trading mode.financeclassificationx★ 5022026-09-20
• NeuroLinkOne interface for 40 LLM providers, adds Jev-backed decide calls.sdkroutingmcpgithub★ 13442026-09-20
• DocJevClassifies and splits PDFs/DOCX into categorized documents using Jev.classificationragdatagithub★ 4542026-09-20
• arc-cuaExecutes bounded desktop UI subtasks click by click using Jev.browsercoding-agentgithub★ 4432026-09-20
• @receptron/layaRuns the open Jev-compatible Laya decision model locally via ONNX.sdkclassificationgithub★ 2242026-09-20
• JevGrepFinds code by behavior using Jev-scored search, returns exact excerpts.coding-agentsearchmcpgithub★ 2042026-09-20
• Jev ExplainedInteractive playground demonstrating Jev's three typed decision primitives live.educationdemogithub★ 1222026-09-20
• System One HarnessTurns any System One model into a confidence-gated agent loop.coding-agentclassificationgithub★ 1332026-09-20
• ReadAloud (dasheng)Scores English read-aloud pronunciation errors locally using ASR plus Jev.educationmediagithub★ 1232026-09-20
• arbiterSelf-hosts typed-decision models like Laya behind a Jev-compatible API.sdkclassificationgithub★ 1132026-09-20
• OpenJev (SiliconLabAI)Approximates Jev by scoring each answer option in parallel calls.classificationgithub★ 1222026-09-20
• dehydratorClient-side tool search for LLM APIs, using BM25 or Jev.mcpsearchclassificationgithub★ 1042026-09-20
• jev-superpowersAdds Jev typed decisions and package vetting to coding-agent workflows.coding-agentguardrailsgithub★ 932026-09-20
• Jev Security ScanReviews Skills and MCP code for injection risk using Jev.securityguardrailsmcpgithub★ 942026-09-20
• JevvyAuto-approves harmless shell commands in coding agents via Jev.guardrailscoding-agentgithub★ 922026-09-20
• JCR (Jev Capability Resolver)Finds deterministic commands in a capability tree using Jev search.coding-agentmcpgithub★ 932026-09-20
• AzdajaKeeps source local and recurses into LLM calls, with optional Jev.coding-agentsdkgithub★ 922026-09-20
• SiftLabels every X post's substance and hides unwanted ones via Jev.classificationmediagithub★ 822026-09-20
• jevQLAdds semantic jev() functions to plain SQL queries over Postgres.dataclassificationgithub★ 842026-09-20
• OpenThai-SystemOneOpen Thai and English decision model mirroring Jev's API contract.classificationsdkgithub★ 822026-09-20
• jev-sentinelScreens agent tool calls and replies for injection using Jev.guardrailssecuritycoding-agentgithub★ 842026-09-20
• Jev vs MLBenchmarks Jev against 11 classical ML pipelines on 8 datasets.evalclassificationresearchgithub★ 942026-09-20
• jsortSorts text by meaning via pairwise Jev comparisons, reports reliability.classificationevaldatagithub★ 842026-09-20
• cascade-searchParses queries locally, escalates only uncertain words to Jev.searchclassificationroutinggithub★ 842026-09-20
• jev-cli (shaharia-lab)CLI turns Jev questions into exit codes for shell and CI.cliclassificationopsgithub★ 732026-09-20
• snapjudgeReads typed-decision probabilities from local Qwen model logits, TypeSafe-compatible.classificationsdkgithub★ 732026-09-20
• jev_jsonschemaConverts a JSON Schema into Jev questions and back into JSON.sdkclassificationgithub★ 742026-09-20
• Jev CookbookFifteen tested Jev recipes for triage, tagging, PII, and reranking.classificationragdatagithub★ 742026-09-20
• JekhovUses Jev only to pick Playwright targets, keeping flow deterministic.browsercoding-agentgithub★ 732026-09-20
• SLO RouterRoutes LLM requests to cheapest backend meeting a latency SLO.routingevalgithub★ 832026-09-20
• JevalsGrades LLM/agent output with Jev, attaching a confidence to each.evalclassificationgithub★ 742026-09-20
• S18ShareMulti-agent runtime with optional Jev-backed routing and skill matching.routingcoding-agentgithub★ 722026-09-20
• av (Agentic Video Intelligence)Video search CLI that uses Jev to filter and rank scenes.ragmediaclassificationgithub★ 732026-09-20
• elons-jobChrome extension hides explicit X replies using Jev classification.browserclassificationgithub★ 822026-09-20
• typesafe-sdk-goOfficial-style Go SDK wrapping the TypeSafe Jev API.sdkgithub★ 822026-09-20
• awesome-jev-usecasesSourced patterns and pitfalls collection for building with Jev.researchgithub★ 1032026-09-20
• mcp-ui-pocJev decides UI widget shape, LLM only writes when uncertain.mcpgithub★ 832026-09-20
• pi-quiet-askJev-backed rule packs gate destructive commands and secret leaks in pi.guardrailssecuritycoding-agentgithub★ 942026-09-20
• pi-jev-routerJev picks model and reasoning effort once per session via AI Gateway.routingcoding-agentgithub★ 942026-09-20
• jevlogsScores OpenTelemetry logs with Jev before routing to expensive LLM analysis.opsroutingdatagithub★ 842026-09-20
• typesafe-skill-routerJev names the one relevant skill to load before the model call.routingcoding-agentgithub★ 842026-09-20
• what-is-jevIndependent research site cataloging 947 rubric-scored Jev repositories.researchgithub★ 122026-09-20
• jev-review (MaxIvanyshen)Code review filter tells coding agents where to look first.coding-agentgithub★ 132026-09-20
• AIStockMulti-agent trading platform offers Jev as a decision backend.financegithub★ 33022026-09-20
• jevmailGmail triage sorts 1,000 emails a minute for 3 cents.classificationgithub★ 2542026-09-20
• JevBenchBenchmark scoring Jev-class models on intelligence, calibration, speed, cost.evalgithub★ 2532026-09-20
• TypeSafe AI PlaygroundCommunity playground with 110 editable Jev classification and routing examples.classificationroutingeducationgithub★ 1532026-09-20
• OpenJev-VisionOpen research toolkit answering multiple typed questions from one image encode.researchmediagithub★ 1722026-09-20
• Jev DirectoryDirectory of 1,300+ Jev builds with an MCP server and index.mcpgithub★ 1322026-09-20
• jev-skill-suggesterRecommends which installed Claude or Codex skill fits a task.coding-agentroutinggithub★ 2232026-09-20
• invalidateFlags stale agent memories when new evidence supersedes them.coding-agentgithub★ 1332026-09-20
• fast-browser-useLocal open-weight browser automation reproducing Jev's fast decision pattern.browsergithub★ 1622026-09-20
• jev-radar (everyinfra)Tracks the Jev ecosystem, rescanning 220+ documented builds every 3 hours.researchdatagithub★ 1622026-09-20
• jevifyServes any logprob-capable LLM through Jev's typed-question API shape.sdkgithub★ 1332026-09-20
• jevwireMCP server and Claude Code plugin gating tool calls with Jev.mcpguardrailscoding-agentgithub★ 1332026-09-20
• jev-linkmapRebuilds a 566-page site's internal link map in 6 seconds.datagithub★ 1532026-09-20
• openvonsDecision layer answering finite choices from text, image, or voice.classificationmediagithub★ 1232026-09-20
• patdownLints a codebase against fuzzy markdown rules using Jev as judge.coding-agentevalgithub★ 1032026-09-20
• jev-rag-benchmarkMeasures whether Jev reranking actually improves RAG quality and cost.ragevalgithub★ 1342026-09-20
• jev (BorisLeMeec)Claude Code plugin answers codebase questions without loading files into context.coding-agentgithub★ 1132026-09-20
• fable-jevSub-100ms Jev reflex layer for routing and context compaction.routingcoding-agentgithub★ 922026-09-20
• jev-mail-classifierConfig-driven inbox classifier tags, moves and flags mail via Jev.classificationgithub★ 1032026-09-20
• findmeFinds files by natural-language memory using beam search and Jev.searchcoding-agentgithub★ 922026-09-20
• hermes-jev-approvalsJev reviews shell command approvals, 8.7x faster on 153 commands.guardrailscoding-agentsecuritygithub★ 932026-09-20
• OpenSourceJevTurns a local Qwen model into a sub-100ms decision engine.researchgithub★ 1622026-09-20
• jev-fn (aaazzam)Decorator turns a Python function signature into a typed Jev query.sdkgithub★ 932026-09-20
• jevcalCalibrates confidence thresholds and flags drift for typed decision models.evalgithub★ 842026-09-20
• jev-recallfilters AI-assistant memories by one calibrated yes/no per item, not embeddingscoding-agentgithub★ 2532026-09-20
• Pi-Jev-GuidePi coding-agent guardrail plugin with configurable rules, redaction, stall detectionguardrailscoding-agentgithub★ 1932026-09-20
• jev-arena (NanmiCoder)CSV-driven side-by-side Jev vs DeepSeek comment-labeling comparison with saved reportsevalclassificationgithub★ 2722026-09-20
• sarvam-jevgeneration-free typed decisions on Sarvam's Indic tokenizer instead of Qwenclassificationresearchgithub★ 3622026-09-20
• jev-dsh-decisionJev decision plugin for DeepSeek Harness, OpenCode and Codex Harnesscoding-agentgithub★ 3222026-09-20
• TypeLLMbrings Jev-style typed decoding to open autoregressive models via constrained decodingclassificationresearchgithub★ 1832026-09-20
• siftrankLLM needle-in-haystack ranker adds a Jev provider optionsearchgithub★ 20032026-09-20
• third-handmacOS menu bar assistant reads Accessibility tree, types and clicks via Jevroboticsgithub★ 23322026-09-20
• hermes-jev-skillsJev handles model routing, memory, compaction and skill selection for agentscoding-agentroutinggithub★ 24732026-09-20
• openJev-verdict-2.0open 150M model claims higher accuracy and better calibration than Jevevalresearchgithub★ 17732026-09-20
• jev-skill (wuyoscar)90 catalogued Jev use cases and installable agent skillscoding-agentgithub★ 14422026-09-20
• awesome-jev-gallerycurated papers, open reproductions and independent Jev evaluationsresearchevalgithub★ 9522026-09-20
• jev-use (savka777)voice-controlled Mac computer use via Accessibility tree, no vision modelroboticsmediagithub★ 6732026-09-20
• LLM2Jevadapts local LLMs into Jev-style Choice, Score and Noul outputssdkgithub★ 8232026-09-20
• pi-bifrostPi model router with an optional Jev tier-selection backendroutingcoding-agentgithub★ 5232026-09-20
• jegrepsemantic grep with calibrated per-path yes/no probabilities, no indexsearchcligithub★ 5332026-09-20
• grok-bot-jevJev gates Grok Bot retries, caching and subagent spawn decisionsguardrailscoding-agentgithub★ 6022026-09-20
• Astraagent runtime with native Jev judgments and EXPLAIN ANALYZE for contextcoding-agentevalgithub★ 3232026-09-20
• jev-desktopJev picks the action inside Codex Computer Use sessionsroboticscoding-agentgithub★ 4222026-09-20
• awesome-jev-zhChinese curated index with an independent-evaluation caveats sectionresearchevalgithub★ 2822026-09-20
• jev-lintflags function names, comments and tests that no longer match codecoding-agentevalgithub★ 3532026-09-20
• jev-as-a-judgebenchmarks Jev as an agent-eval judge against three LLM judgesevalgithub★ 2832026-09-20
• jev-mcp (burnigtm)MCP server for coding-loop routing, review, verify and screen toolsmcpcoding-agentgithub★ 3132026-09-20
• typesafe-mcp (PyModel)Go MCP server exposing Jev's Choice, Score and Noul as one toolmcpsdkgithub★ 2322026-09-20
• JevbridgeACP and MCP adapter bridges Jev with Claude, Codex, Grok and OpenCodemcpgithub★ 3122026-09-20
• jev-botturns an idea into a Jev integration design from a research libraryresearchgithub★ 122026-09-20
• jevproxyreverse proxy claims 65% cost cut and 90% latency cut for agentsopscoding-agentdiscord22026-09-20
• FakeCatchChrome extension scores review trust as real, generic, ad-like or shortbrowserclassificationx22026-09-20
• Vettlycontent moderation API for text, image and video built on Jevsecuritymediaclassificationdiscord22026-09-20
• eslint-plugin-jevESLint rules are plain-English questions Jev answers with a probabilitycoding-agentgithub★ 232026-09-20
• awesome-jev-toolsanother curated index of public projects and patterns built on Jevresearchgithub★ 54322026-09-20
• Jeveloper100 example Jev use cases: routing, scoring, agents, support workflowseducationcoding-agentdiscord22026-09-20
• Jev on Venice APIJev added to Venice API in beta, no JSON parsing neededsdkx32026-09-20
• mefi-studiodesktop AI workspace, Jev routes models by performance, speed and costroutinggithub★ 422026-09-20
• jevfanity-apihosted profanity detector using Jev, no API key neededclassificationsecuritygithub★ 022026-09-20
• jev-codex-pilotCodex web overlay picks model and reasoning depth, automated Kanban task sequencingcoding-agentroutinggithub★ 122026-09-20
• hunchRuby gem wraps Jev's three question types as chance, pick and ratesdkgithub★ 1332026-09-20
• pi-jev-wikiPi package, agent-maintained project wiki holds only reasoning, not code-derivable factscoding-agentgithub★ 032026-09-20
• okolocal search plus Jev reranking cuts what a coding agent has to readsearchcoding-agentraggithub★ 242026-09-20
• jev-prune-kitcapability-aware context-pruning installer for agent harnesses, experimentalcoding-agentgithub★ 122026-09-20
• Jev Moderation BotDiscord moderation with 4-stage escalation, audit logging, one-click pardon learningsecurityclassificationgithub★ 3932026-09-20
• jev-logtriageJev decides whether a batch of logs is worth acting on, confidence-gatedopsdatagithub★ 122026-09-20
• jevcache.shmemoizes Jev-class decisions so repeats are free and deterministicopsdiscord22026-09-20
• jev-javaidiomatic Java SDK for the Jev decision enginesdkgithub★ 322026-09-20
• jev-builderbrowser form builds Jev requests from 34 templates, no JSON by handbrowserdiscord22026-09-20
• AlloyJev routes tasks across Claude, Codex, Gemini and Antigravity by capabilityroutingcoding-agentgithub★ 232026-09-20
• opencode-plugin-variantizerJev picks the cheapest capable OpenCode model and reasoning variant per taskroutingcoding-agentgithub★ 032026-09-20
• enzymecompiles Markdown-wiki reading rules into Jev queries, claims 350x cheaper than frontierclassificationopsgithub★ 7932026-09-20
• jev-studioMCP tools for Choice, Noul and Score plus prompt libraries and commandsmcpgithub★ 1222026-09-20
• typesafe4stype-safe Scala 3 SDK wrapping the Jev APIsdkgithub★ 222026-09-20
• NolaTypeScript superset adds Jev as a typed-inference provider alongside plain TS typessdkdiscord22026-09-20
• Jev 1.13 jaggedness (TypeSafe docs)TypeSafe's own list of known Jev weak spots, math firstevalresearchdiscord32026-09-20
• jarviscore-frameworkmulti-agent framework, Jev routes subagents and filters RAG passages by conflict/injectionroutingragguardrailsgithub★ 1332026-09-20
• pi-jev-sentinelPi agent extension screens tool calls and output for injection, scrubs secrets, pins taskguardrailssecuritycoding-agentgithub★ 832026-09-20
• jev-alignCLI calibrates Jev to your own judgment criteria using GEPAclievalgithub★ 23532026-09-20
• jev-usehands no-text-output agent steps to Jev, p50 ~230ms, ~$0.02 per 1,000 judgmentscoding-agentclassificationgithub★ 1032026-09-20
• JevSwiftSDKunofficial Swift SDK, async/await, batching, zero dependencies, iOS/macOS/Linuxsdkgithub★ 732026-09-20
• revolve-agentcoding agent uses Jev to gate risky commands, set permissions and flag instruction drift, still WIPcoding-agentguardrailsdiscord22026-09-20
• JevFindsemantic code search CLI, returns files, line ranges and confidence scores, author calls it a democlisearchgithub★ 422026-09-20
• jevrouter-tsTS router picks cheapest model tier per query using Jev classificationroutingclassificationgithub★ 522026-09-20
• JevNQLnatural-language database queries via Jev instead of SQLdatasearchgithub★ 522026-09-20
• Jev CLI (vectorz)terminal CLI with skills and plugins for agent workflowsclicoding-agentdiscord22026-09-20
• Jev in production (iambraun)self-updating list of 18 products running Jev in production, scored 1-5researchdiscord32026-09-20
• jev-frontier-bench200-decision benchmark of Jev vs five frontier models, calibration and cost dataevalgithub★ 042026-09-20
• Pi-SaverJev prunes Pi Coder conversation history per turn, claims 75%+ token cutcoding-agentopsgithub★ 032026-09-20
• JevRouter.nvimNeovim plugin routes prompts to file, edit or terminal handlers via Jev, under 200msroutingcoding-agentgithub★ 132026-09-20
• hushGitHub Action triages issues with Jev, stays silent below a confidence thresholdopsclassificationgithub★ 132026-09-20
• jev-ushermodel routing plus recoverable context filtering for Claude Code, local compare UIroutingcoding-agentgithub★ 132026-09-20
• openpoke-meets-jevmoves OpenPoke's email screening, tool guardrail and reranking to Jevguardrailsclassificationgithub★ 132026-09-20
• JevArenaopen BYOK arena blind-votes Jev against other models acting as judgeevalgithub★ 332026-09-20
• Jev For Dummieswraps Jev primitives as plain HTTP GET endpoints for newcomerssdkeducationgithub★ 122026-09-20
• jev4kKotlin DSL and client for the Jev APIsdkgithub★ 122026-09-20
• code-compact-jevJev trims excess comments out of codecoding-agentgithub★ 022026-09-20

New 2026-09-19

• Jev on Vercel AI GatewayJev added to Vercel AI Gateway, one line to call itsdkdiscord32026-09-19
• Jev on Cloudflare AI GatewayJev live on Cloudflare AI Gateway as a typed decision endpointsdkdiscord32026-09-19
• flaviocopes Jev deep divepractitioner walkthrough of the API, primitives and basic usage patternseducationdiscord22026-09-19
• jev-capability-atlasevidence-based map of where Jev's calibration holds up or failsresearchevalgithub★ 2142026-09-19
• jev-reviewersystematic-review data extraction, every answer a verbatim cited quoteresearchdatagithub★ 2342026-09-19
• jev-column-raceJev vs Gemini labelling 1,000 reviews, 4.1x faster, 7x cheaperevalclassificationgithub★ 1842026-09-19
• pi-jev-auto-modeJev auto-approves Pi bash/write/edit calls, fails closed, 171 testsguardrailscoding-agentgithub★ 1842026-09-19
• winnowcalibrated context sieve for Claude Code, stubs blocks Jev judges unneededcoding-agentopsgithub★ 3132026-09-19
• yoshiJev-judged context-pruning proxy for Claude Code and Codex, POCcoding-agentopsgithub★ 1932026-09-19
• pi-dcpdedups and prunes Pi coding-agent context, experimental Jev selectioncoding-agentopsgithub★ 1732026-09-19
• pi-jevsemantic tool router and skill finder for the Pi coding agentroutingcoding-agentgithub★ 2132026-09-19
• pi-typesafebatched Jev eval tool, terminal playground, shared client for Pi extensionsevalclisdkgithub★ 2032026-09-19
• stanley-codeJev routes requests to deterministic coding workflows, falls back to an agentroutingcoding-agentgithub★ 10132026-09-19
• blinkJev-driven file-system walkers search a codebase by natural languagesearchcoding-agentgithub★ 2732026-09-19
• ErisLintRust linter with Jev-judged rules for complexity, naming, comment valuecoding-agentgithub★ 1532026-09-19
• ts-browser-agentLangChain browser agent using Jev as the tool-picking modelbrowserroutinggithub★ 2732026-09-19
• ruby_llm-typesafeTypeSafe provider for RubyLLM, Noul/Choice/Score via structured outputsdkgithub★ 1732026-09-19
• jev (Elixir/OTP)GenServer replies to Jev, pattern-match the typed answersdkgithub★ 1932026-09-19
• effect-questionssemantic judgment as an Effect-TS primitive: is, choose, rank, branchsdkclassificationgithub★ 1532026-09-19
• Verdict (OpenJev)open 151M non-autoregressive decision model, calibrated uncertainty, WebGPU demoresearchdemogithub★ 4232026-09-19
• litjevturns any Qwen checkpoint into a Jev-style typed decision layerresearchgithub★ 2832026-09-19
• jev-on-a-laptopreproduces Jev's parallel constrained decoding on a 1.5B model locallyresearchgithub★ 2232026-09-19
• open-jev (DiffusionGemma)typed JSON via diffusion-model denoising, benchmarked against saved Jev decisionsresearchgithub★ 2032026-09-19
• jev_locallocal Jev-style API on LFM or ModernBERT, JGLUE-benchmarkedresearchevalgithub★ 1832026-09-19
• jevscan-evmheat map of likely bugs across a repo, vibe-coded proof of conceptsecuritycoding-agentgithub★ 2122026-09-19
• ulkaexperimental browser extension, Jev selects actions from the accessibility treebrowserdemogithub★ 2122026-09-19
• ainodeself-hosted GPU appliance with a typed decision endpoint alongside inferencesdkopsgithub★ 1522026-09-19
• awesome-jevanother curated Jev index, links to madewithjev.com use casesresearchgithub★ 6122026-09-19
• phuryn/experiments (jev-decisions-api)hardened invoice test: Jev ties Haiku 4.5, beats Opus 115x on costevalfinancegithub★ 5242026-09-19
• Sandobounds Claude Code and Codex tool output, TypeSafe shadow judge optionalguardrailscoding-agentgithub★ 5442026-09-19
• jev-codex-routerper-turn Codex model routing, measured 60% savings on 237 turnsroutingcoding-agentgithub★ 8642026-09-19
• jev-rulesJev picks which CLAUDE.md rules apply to each promptcoding-agentclassificationgithub★ 4332026-09-19
• jev-prunertrims noisy Bash output before Claude sees it, keeps text verbatimcoding-agentopsgithub★ 12632026-09-19
• skillrankerRust CLI ranks Claude Code skills for the next stepclicoding-agentgithub★ 9632026-09-19
• save-token-jev-cleanJev picks which tool calls survive compaction across five agent hostscoding-agentopsgithub★ 5832026-09-19
• jev-semgrepgrep by meaning across languages, no translation step neededsearchcoding-agentgithub★ 11632026-09-19
• supercovscores code quality with Jev, pairs with test coverage gapscoding-agentevalgithub★ 7532026-09-19
• jev-siftclassifies files and URLs by relevance before the agent reads themclassificationcoding-agentgithub★ 4532026-09-19
• advocaattyped client for asking Jev questions about your datasdkdatagithub★ 8632026-09-19
• decideropen one-pass typed decision model, fine-tuned Qwen3.5-2B, HF benchmarks includedresearchevalgithub★ 15432026-09-19
• jeff (GliFormer)self-hosted drop-in Jev replacement, cheaper but less accuratesdkgithub★ 16732026-09-19
• fabricioctelles/skills26 agent skills, four use Jev for quality judgmentcoding-agentevalgithub★ 8832026-09-19
• commit-minerclassifies commit diffs by bug fix, CWE and change typeclassificationsecuritygithub★ 3032026-09-19
• vibecheckscores draft X posts on virality, cringe and regret risk before postingwritingclassificationgithub★ 4532026-09-19
• JevScoutJev scores job links and postings while Chrome drives the browsingbrowserclassificationgithub★ 2522026-09-19
• typesafe-ai-benchmarkcompares Jev against LLM-native structured output on latency, cost and qualityevalgithub★ 3432026-09-19
• mini-Jevpreregistered test reads option-letter logits instead of generating JSONresearchevalgithub★ 3332026-09-19
• building-with-jev-skillClaude Code skill for writing and debugging Jev-calling programscoding-agenteducationgithub★ 12632026-09-19
• fast-jev-compactionJev scores each tool call, drops stale ones, keeps rest verbatimcoding-agentopsgithub★ 5.1k32026-09-19
• compact-adviserJev judges when a Claude Code session is safe to compactcoding-agentopsgithub★ 14832026-09-19
• claude-code-traceJev scores Claude Code sessions on progress, focus and token efficiencycoding-agentevalgithub★ 37032026-09-19
• jev-review (dashboard)staged Jev code review workflow with a local dashboardcoding-agentgithub★ 40832026-09-19
• perchsemantic linter flags defects with confidence scores, CLI and agent skillcoding-agentcligithub★ 16432026-09-19
• foremanJev watches Codex workers, judges completion and drift, labeled an experimentcoding-agentevalopsgithub★ 43522026-09-19
• evotterminal coding agent, Jev prunes stale context instead of summarizing itcoding-agentcligithub★ 32632026-09-19
• openagents (Coder)terminal coding agent, Jev routes turns and judges shell outputcoding-agentcliroutinggithub★ 45022026-09-19
• vexjoy-agentJev routes plain-English requests across 43 specialist Claude Code agentsroutingcoding-agentgithub★ 42132026-09-19
• jev-router (per-turn)routes each Claude Code or Codex turn to the cheapest capable modelroutingcoding-agentgithub★ 25522026-09-19
• jev-mcp (10 tools)verify, screen, rerank, classify, review and gate as MCP toolsmcpguardrailsclassificationgithub★ 15742026-09-19
• tax-doc-classifierclassifies pages across 261 IRS forms, 100% accuracy, $0.001 per pageclassificationfinancegithub★ 29542026-09-19
• classifier.devzero-shot text classification over HTTP, no key, Jev backendclassificationsdkgithub★ 35232026-09-19
• llm-rankers (Jev arm)academic reranker benchmark adds Jev as a pointwise and listwise rerankerevalsearchresearchgithub★ 21132026-09-19
• jev-eval-agentcompares agent step counts when Jev picks the tool versus the LLMevalcoding-agentgithub★ 9732026-09-19
• jev-ultrafastJev picks click target and action, small LLM only types textbrowsergithub★ 11.7k32026-09-19
• jev-browserJev picks one browser action per step; Wikipedia nav in 4sbrowsergithub★ 18032026-09-19
• jev-browser-useJev clicks and navigates, Codex verifies; 5-10x faster in productionbrowsercoding-agentgithub★ 24932026-09-19
• SemIfopen 4B model reproducing Jev's typed-decision interface, runs via WebGPUresearchsdkgithub★ 2.4k22026-09-19
• kevtiny Jev-like model on Qwen2.5, trains and runs on a MacBookresearchsdkgithub★ 98222026-09-19
• layanon-autoregressive multilingual decision model, 33ms per call on a T4researchclassificationgithub★ 3.9k32026-09-19
• LangWatch Instant EvalsJev judges your whole production history on demand; 97% agreement with human labels on 300 support chats, $0.32 per 10k conversations (vendor's claim)evalopsdiscord2026-09-19
• JevSubRouterhook routes each Claude Code subagent dispatch to the cheapest model that can finish it; ~300ms warmroutingcoding-agentgithub★ 52026-09-19
• toolgateClaude Code hook and MCP proxy blocks, confirms or allows tool calls by riskguardrailsmcpsecuritygithub★ 22026-09-19
• jev-codeTypeScript coding CLI using Jev typed decisions and constrained AST generationcoding-agentcligithub★ 172026-09-19
• jevusecasescrowdsourced directory of what people replaced with Jevdemodiscord2026-09-19
• saypagedescribe a page in plain language, builds a standalone site in under 0.4sdemomediadiscord2026-09-19
• jev-cvssCVSS 3.1/4.0 vectors from a plain description, matches NVD on sample setsecurityclassificationgithub★ 22026-09-19
• jev-browser (MCP)LLM plans, Jev decides click target and action, Playwright executesbrowsermcpgithub★ 5032026-09-19
• vonopen-source non-autoregressive decision model, sub-15ms local alternative to Jevresearchsdkgithub★ 18832026-09-19
• jevbetterstronger one-pass option scorer, benchmarked head-to-head against jevlikeevalresearchgithub★ 1332026-09-19
• OpenDecisionopen-source Choice/Noul/Score engine over zero-shot modelssdkclassificationgithub★ 4032026-09-19
• duckdb-jevDuckDB extension, Jev answers typed as real SQL columns per rowdatasdkgithub★ 1532026-09-19
• openjevlocal bilingual probability decisions on a frozen Qwen3-4B backendresearchsdkgithub★ 2132026-09-19
• tidepoolHaskell notebook agent harness, Jev supplies System 1 judgmentscoding-agentresearchgithub★ 1232026-09-19
• oxlint-plugin-jevplain-English yes/no lint rules, Oxlint finds the code, Jev judges itcoding-agentguardrailsgithub★ 4132026-09-19
• jev-benchmarkscalibration, selective risk and latency eval, Jev vs GLiNER2.5evalclassificationgithub★ 1242026-09-19
• jevilJev + agent-device QA agent, reads a mobile app and picks the next actionbrowserevalgithub★ 1132026-09-19
• jev-macos-loopnative macOS GUI automation, OmniParser plus Jev for action choicebrowseropsgithub★ 1532026-09-19
• jgrepgrep by natural-language description, pipeable, ~200ms per lineclisearchgithub★ 1242026-09-19
• macbrowvoice-controlled Mac, Jev picks the AppleScript or browser toolbrowsercligithub★ 8132026-09-19
• choosekittyped choices and probabilities from local llama.cpp modelssdkgithub★ 1332026-09-19
• beam-cli (AgentBeam)hooks and local policy layer across Claude Code, Codex, Jevcoding-agentguardrailscligithub★ 1132026-09-19
• Turncompiled language, Choice/Score/Noul decisions as durable VM effectssdkresearchgithub★ 1122026-09-19
• system-onebatched single-token choice inference for open models, Jev-compatiblesdkresearchgithub★ 2032026-09-19
• open-alternative-jevopen System One layer for open-weights LLMs, benchmarked on RACE-Hresearchevalgithub★ 3442026-09-19
• jotgeneral-purpose agent loop where Jev picks the tool and argumentscoding-agentroutinggithub★ 1632026-09-19

Do now

• jev-watchbaseline today's answers, flag drift when TypeSafe updates the modelevalopsgithub★ 02026-09-19
• Jev on OpenRouter$0.042/M input, $0 output; plugs into the broker with no new adapterroutingsdkdiscord2026-09-19
• jev-sec-bench96.5% on deepset prompt-injection set, 325ms p50; second opinion beside the intake regexsecurityevalgithub★ 22026-09-19
• jev-belayStop hook that catches "tests pass" when nothing ranguardrailscoding-agentgithub★ 142026-09-19
• jev-commitcommit-msg hook: message vs diff, leaked creds, debug leftoversguardrailssecuritycoding-agentgithub★ 82026-09-19

Coding-agent guardrails (Claude Code hooks)

• pi-wardenjudges every tool call, output and reply in ~250msguardrailssecuritygithub★ 1062026-09-19
• second-thoughtshell commands judged pre-execution, blocks past 85% confidenceguardrailssecuritycligithub★ 42026-09-19
• jev-guardauto-approval layer for Claude Code and Codex tool calls, ships a bypass CTFguardrailssecuritycoding-agentgithub★ 22026-09-19
• omp-auto-modesafe/unsafe/ask tool-call classifier emulating auto modeguardrailsclassificationgithub★ 22026-09-19
• pi-heedenforces "never touch prod" constraints on tool calls, 46 testsguardrailsopsgithub★ 52026-09-19
• tenetjudges every agent commit against plain-English rules in AGENTS.mdguardrailscoding-agentgithub★ 42026-09-19
• abidemakes the coding agent obey project rulesguardrailscoding-agentgithub★ 1962026-09-19
• Reapersemantic linter for silent failures, weakened tests, scope creepcoding-agentguardrailsgithub★ 12026-09-19
• jevlintplain-English convention rules checked in the agent loopcoding-agentguardrailsdiscord2026-09-19
• commentlintchecks that agent-written comments still match the codecoding-agentguardrailsgithub★ 12026-09-19
• wincescores diffs by blast radius and auth/data-write for review routingsecuritycoding-agentroutinggithub★ 22026-09-19
• Migration Guardianhalts destructive SQL migration plansguardrailsdataopsgithub★ 22026-09-19
• is-maliciousscreens a repo or PR for obviously malicious code before running itsecurityguardrailsgithub★ 172026-09-19
• Gorgonazero-dependency Python guardrail engine, halts agents on leaked keys or loopsguardrailssecuritygithub★ 22026-09-19

RAG, grading, classification (GWEDU)

• Jev vs rerankersties Cohere rerank-4-pro on NDCG@10 across 8 datasetssearchevaldiscord2026-09-19
• GBrain Jev reranker(PR by Daniel Andrade, see channel)Score-based reranking, 1.9s vs 14.4s LLM at similar cost2026-09-19
• ReadyBase rerank eval28-scenario gold-labeled harness for code-context selectionevalsearchcoding-agentdiscord2026-09-19
• extraction-marker-recoveryrestores OCR-flattened footnote and citation markers in passagesdataresearchgithub★ 02026-09-19
• jev-document-classificationdocument classes plus injection detectionclassificationsecuritygithub★ 22026-09-19
• book-aurora601 passages, 9 emotion scores, 25s, $0.034classificationmediagithub★ 52026-09-19
• AI Elo rankertournament scoring for text, reusable for rubric gradingevalgithub★ 52026-09-19
• sqlite-jevbatched judgments as SQL functions in SQLitedatasdkgithub★ 22026-09-19
• pg-redacttags PII spans in Postgres text, redacts only the spansecuritydatagithub★ 22026-09-19
• FairCheckRAG-grounded guardrail citing regulation paragraphs; pattern for cited refusalsguardrailsragdiscord2026-09-19
• sentence-relationship classifiernon-LLM segmentation plus 3 Jev calls, ~0.5sclassificationdiscord2026-09-19

Eval and adoption method

• Sniff Testprose linter as CLI, pre-commit, Action and Claude Code skill; 182ms, $0.013 per 100 paragraphswritingcligithub★ 252026-09-19
• shadow-mode adoption write-uprun Jev beside the LLM for 3 days, promote only where it winsopsevalx2026-09-19
• Working with Jevpractitioner notes from a real PR reviewercoding-agentwritingdiscord2026-09-19
• CultivarPinecone's agent-skill eval harness with Jev gradingevalcoding-agentgithub★ 402026-09-19
• dinostompverifies the scorer, not just the score, with a blind control per runevalgithub★ 52026-09-19
• Vercel eve-agent failure classifier320 sessions, 2M tokens, classified in 14s for $0.08classificationevalcoding-agentx2026-09-19
• Empryo harness write-upfive decision points in a coding agent benchmarked against frontier modelsevalcoding-agentdiscord2026-09-19
• Swamp alert triage91% cheaper triage before routing to Clauderoutingopsclassificationdiscord2026-09-19
• ZeroSweepJev vs LLM inbox triage, live comparisonclassificationevalgithub★ 22026-09-19
• optimaizrranks LLM spend waste by dollar savings on real trafficfinanceopsdiscord2026-09-19

SDKs, adapters, glue

• JevRouterJev-powered router for models, tools and subagentsroutingcoding-agentgithub★ 1122026-09-19
• daf-jevPython toolkit: confidence gates, tiered routing, calibration module, MCP serverroutingmcpguardrailsgithub★ 42026-09-19
• dspy-typesafeifyone decorator for Jev-backed DSPy signaturessdkcoding-agentgithub★ 612026-09-19
• jev-mcpMCP server exposing classify, score, check, match, screenmcpclassificationgithub★ 142026-09-19
• typesafe-mcpdrop-in MCP connector for Claude Codemcpcoding-agentgithub★ 1332026-09-19
• jev-cliverify, screen, classify, extract, route, rerank from the shellcliclassificationgithub★ 122026-09-19
• jev-axipick, rate, check, rank, triage, guard shell commandscliguardrailsgithub★ 172026-09-19
• simple-jevturn any open model into a classifier endpoint; fallback if TypeSafe goes awayclassificationsdkgithub★ 3892026-09-19
• System One SearchJev-only codebase retrieval and reference graphsearchcoding-agentdiscord2026-09-19
• everyask a yes/no question of every function in a codebasecoding-agentresearchgithub★ 32026-09-19
• jev-browsercheap Playwright browser agent driven by Jevbrowsergithub2026-09-19
• typesafe-computer-useOCR, classify, click loop at ~$0.0002 per stepbrowserclassificationgithub★ 6182026-09-19
• vgi-typesafeDuckDB extension, Jev as LATERAL joinsdatagithub★ 22026-09-19
• pg-jevPostgres extension, plain-English per-row WHERE judgmentsdatadiscord2026-09-19
• HA-JevHome Assistant sensors and automationsopsgithub★ 362026-09-19
• awesome-jev-typesafe250-entry curated indexresearchgithub★ 1082026-09-19
• awesome-jev-projects287-entry index with English viewresearchgithub★ 2122026-09-19
• TypeSafe primitives quiz(dball9, quiz_handover.md in channel)self-test on Choice, Noul, Score and the cascade patterns2026-09-19

Cool

New 2026-09-20

• 1v1 JevFPS bot where Jev decides movement, aim, and firing at 9Hz.gamedemox★ 612026-09-20
• Jev's FlyJev steers a Three.js flying-game character in real time.gamedemogithub★ 912026-09-20
• RoboJEVTwo-stage Jev control of a simulated Franka Panda arm, evaluated.roboticsevalgithub★ 932026-09-20
• ST-jevedJev reads roleplay replies and triggers narrator rules or rerolls.gamemediagithub★ 912026-09-20
• J++Experimental language where questions and methods compose as values.researchgithub★ 812026-09-20
• JEVfireBatches typed-variable decisions in parallel on CUDA LLMs, benchmarked.classificationgameresearchgithub★ 832026-09-20
• RuneBench with Jev (jevscape)Lets Jev play RuneScape through a bounded 50-action harness.gamedemogithub★ 712026-09-20
• xtagsChrome extension tags X posts with Jev-derived intent labels.browserclassificationgithub★ 822026-09-20
• syft-listeningReal-time speech analysis using Jev, tested live in browser.mediaresearchgithub★ 122026-09-20
• slop-filterChrome extension hides AI-generated posts on X and LinkedIn.browserclassificationgithub★ 322026-09-20
• Jev plays F-ZeroJev plays F-Zero from raw screen pixels, no game code.gamex12026-09-20
• crush-monitorAnalyzes WeChat chat sentiment and rates reply quality with Jev.classificationmediagithub★ 3512026-09-20
• Nemotron_JevServes a diffusion model behind a Jev-shaped decision API.mediasdkgithub★ 1212026-09-20
• jev-reflex-autonomy-labMulti-drone sim where fast Jev reflexes escalate to a slower planner.roboticsgithub★ 1212026-09-20
• feelingsAdds a typed .feels() method to any value via Jev.sdkgithub★ 1222026-09-20
• JevthovenGenerates editable multitrack MIDI music one Jev decision at a time.mediagithub★ 1022026-09-20
• PlayJev0.8B model plays ten browser games directly from raw pixels.gamegithub★ 1322026-09-20
• JevinikStock terminal predicts 30-day price direction using Jev over evidence.financegithub★ 1422026-09-20
• jev-robot-control (openroboto-ai)Jev vs GPT-6 Astra vs GPT-4.1 mini placing an appleroboticsgithub★ 2732026-09-20
• transcript-lensJev finds chapters, claims and key passages in YouTube transcripts, Turkish UImediaresearchgithub★ 1522026-09-20
• jev-seofree SEO/GEO CLI, Jev classifies intent instead of paid Semrush-style scoringclisearchgithub★ 1922026-09-20
• jev-game-tools (Brotato)Jev picks every-frame movement, Claude handles long-term shop strategygamegithub★ 1422026-09-20
• Jev OthelloJev plays Othello against random moves, heatmap vs minimax comparisongamex12026-09-20
• Jev + Stagehand browser useaccessibility tree state, Jev decides next click, task cost $0.001browserx22026-09-20
• minecraft-agentAstra plans, Jev picks moves in a verified 8m43s dragon killgamecoding-agentgithub★ 26822026-09-20
• embodied-jevMuJoCo robot arm workbench compares Jev against local and cloud modelsroboticsgithub★ 8922026-09-20
• SmartMoney-Cubread-only trading journal uses Jev for typed judgments, no live ordersfinancegithub★ 2522026-09-20
• MKUltraScale logic circuitsbuilt logic circuits out of Jev gatesresearchx12026-09-20
• Shader from text (0xFotex)turns any text into a visual shader with Jevmediax12026-09-20
• Jev color palette generator (fran)Jev scores hue, warmth and energy, then picks a fontmediademox12026-09-20
• jevmojitype anything, get related emojis scored 0-3 by Jevmediademogithub★ 112026-09-20
• Adventures of JevJev plays an RPG adventurer across 43 places and 31 peoplegamediscord12026-09-20
• Jev's Sprint PlanningJev negotiates sprint scope, coordinates four developers in a 3D officeopsdemodiscord12026-09-20
• Fieldnotesemantic physics and math solver, Jev solved 3 MIT Integration Bee questionseducationresearchdiscord22026-09-20
• jevchatturns Jev into a chatbotdemogithub★ 1712026-09-20
• jev-liberoJev controls a simulated robot arm to close a microwave and drawerroboticsgithub★ 2322026-09-20
• Quarrag and the Sun-Heart of MordanneSonnet writes the story, Jev makes every choicewritinggamediscord12026-09-20
• Jev-X-Sentiment-Analysisscores crypto tweets bullish or bearish, combines with funding rate and RSIfinanceclassificationgithub★ 13222026-09-20
• jev-ncr-demosuggests defect codes from a plain-English non-conformance report, Rust/Leptos/Axumclassificationdemogithub★ 122026-09-20
• jev-311-heatmapNYC 311 complaint heatmaps classified and mapped by Jevclassificationdatagithub★ 212026-09-20
• HN reranker (danprice.ai)reranks Hacker News with a plain-English query like "jealous author"searchroutingdiscord22026-09-20
• ORIGIN-CIVILIZATIONinspectable life-and-civilization sim, every voluntary NPC action needs a Jev decisiongameresearchgithub★ 212026-09-20
• JevLM Studioplayground compares three experimental models against ten promptsevaldiscord12026-09-20
• canyoubeatjev.fyiunfinished dashboard for comparing Jev against other models head to headevaldemodiscord12026-09-20
• HazAlerts near-melive AU fire and emergency map triaged by Jev, concept democlassificationdemodiscord12026-09-20
• Odyssey (Jev flight director)moon and Mars launch sim narrated by Jev's decisions with confidence scoresdemomediadiscord12026-09-20
• jamseshMIDI jambox agent follows your lead while you play musicmediagithub★ 412026-09-20
• jev-mermaidJev-driven Mermaid diagram generatormediacoding-agentgithub★ 312026-09-20
• flappyaireversed Flappy Bird, Jev controls the missiles, you move the guardrailgamediscord12026-09-20
• memefyfinds a matching meme for your text using Jevmediademodiscord12026-09-20
• Vesper living-town demoUltima Online fan demo, Jev picks each NPC's next actiongamedemodiscord12026-09-20
• Jev vs Jev Ultima Online duelstwo Jev-controlled characters duel each othergamediscord12026-09-20
• jevfishchess engine, every move a Jev judgment call, no search, 4-0 vs humans so fargamediscord22026-09-20
• jev-chess (SyedZawwarAhmed)Jev picks a move from a code-generated legal-move listgamegithub★ 122026-09-20
• City Dispatchdriving game, bot cars use Jev and sensor input to drivegamediscord22026-09-20
• Jevilishword game, guess the phrase Jev mangled into synonymsgamediscord12026-09-20
• From a movietype a line or scene, Jev finds the moviesearchmediadiscord12026-09-20
• Jev MIDI phrase pickerJev selects the next musical phrase from generated candidatesmediadiscord12026-09-20
• The Trolley Problemput anything on the tracks and find out whether Jev pulls the leverdemoguardrailsdiscord12026-09-20

Older

• [Jev live demo (val.run)](typesafe-demo.val.run), [killmyidea](github.com/monteduro/killmyidea), [typesafe-snake](github.com/sorrycc/typesafe-snake), [jev-voice](github.com/kevinbadi/jev-voice), [tsai-sc](github.com/phyous/tsai-sc), [refgarden](github.com/AlbionaHoti/refgarden), [jev-playground](github.com/mizchi/jev-playground), [jevpilot](github.com/standardagents/jevpilot), [jev-recruiter](github.com/skeptrunedev/jev-recruiter), [jev-trade (Hyperliquid)](github.com/aowang-ai/jev-trade)
• [ghosthands](github.com/affirmitv/ghosthands), [jev-chat](github.com/w3cj/jev-chat), [jev-paint](github.com/achimala/jev-paint), [hono-jev-router](github.com/yusukebe/hono-jev-router), [neo4jev](github.com/jexp/neo4jev), [jev-shell-history](github.com/mrnugget/jev-shell-history), [jevmeter](github.com/ChetasLua/jevmeter), [typesafe-adblock](github.com/realZachi/typesafe-adblock), [youtube-sponsor-detection](github.com/trungdq88/youtube-sponsor-detection), [jev-trader](github.com/jarrodwatts/jev-trader)
• [mobile-jev](github.com/droidrun/mobile-jev), [jev-voice-browser](github.com/moritzkremb/jev-voice-browser), [unclutter](github.com/kitze/unclutter), [jev-visual](github.com/hr98w/jev-visual), [typesafe-mario](github.com/fhshaik/typesafe-mario), [Jev Search](github.com/superagents-lab/jev-search), [Jev-Trades](github.com/zadescoxp/Jev-Trades), [live-jev (driving sim)](github.com/vinilana/live-jev), [typesafe-playground](github.com/kavehmz/typesafe-playground), [mario-jev](github.com/shantanugoel/mario-jev)
• [live-jev (Ableton)](github.com/okinaaudio/live-jev), [x-scanner](github.com/oso95/x-scanner), [Piano in real time](x.com/chandamamz/status/2101066294816920063), [Jev vs Opus competitive Pokémon](x.com/sid19arya0/status/2100458351440048258), [JevPlaysPokemon](github.com/anxkhn/JevPlaysPokemon), [Mario from NES memory](x.com/rherton/status/2101005942913729001), [TerraBlind](github.com/Reisenbug/TerraBlind), [Wordle x32](x.com/RichardSuwandi), [Jev vampire survivors](github.com/oldmoldycake/jev_vampire_survivors), [Jev arena](jev-arena.vercel.app)
• [Coup](coup.clairbridge.tech), [jev-chess](jev-chess.tgr.rs), [Doom](x.com/CompleteSkeptic/status/2099925687465570372), [Minecraft zombie escape](youtu.be/3G14OU00crI), [Ultima Online combat AI](youtube.com/watch?v=cP-GWKteHKo), [Ember Hollow living village](lifelike.build), [NPC judgment village](x.com/haiderTheDev/status/2100419750459519156), [Coffee-shop sim](x.com/soya_da_yoot/status/2100451279222755616), [DJ/VJ director](x.com/shannon_fano/status/2100639390468833379), [Real-time video effects](x.com/ponyo877/status/2101139914419290345)
• [JevPixelArt](github.com/rivianpratama/JevPixelArt), [jev-drone](github.com/RomanSlack/jev-drone), [SO-101 arm via quackd](github.com/rokbenko/quackd/blob/main/docs/jev.md), [Robot manipulation gating](x.com/JonyShaik/status/2100827590781214981), [jev-mobile](github.com/Friedjof/jev-mobile), [WikiRace](x.com/princecaarlo/status/2100313645116952701), [Football, Jev vs Jev](x.com/attractmodeio/status/2100723340021276739), [date-with-jev](date-with-jev.vercel.app), [Piano, failed](x.com/BaselAshraf81/status/2100400028711805242)

Findings and gotchas

New 2026-09-20

• intikhab49: An open 150M reproduction scored 0.697 vs Jev's 0.727 on the same benchmark, while being 2.5x better calibrated and 4x faster.
• QuicqDev: Benchmarked Jev 1.13 against 11 classical ML pipelines on 8 datasets: Jev hit 96.3% balanced accuracy on IMDb sentiment vs 88.4% for the best classical pipeline, but classical pipelines won on all 4 tabular datasets.
• zeeshan8281: Routing through Jev preserved the same routes and accuracy as local deterministic features but raised p95 end-to-end latency from 77.93ms to 490.38ms.
• DECRUX9812: Jev-based skill routing for Hermes Agent costs about $0.001 per routed turn.
• openroboto-ai: on a single seed-0 xArm7 pick-and-place trial (one apple, one plate), Jev 1.13 placed it in 226 API calls for $0.019 and 182s wall time; GPT-6 Astra also placed it but took 707s at $5.93, 312x the cost; GPT-4.1 mini hit the 160-cycle limit without finishing.
• Kyle Jeong: driving Stagehand from an accessibility-tree state instead of screenshots, letting Jev pick the next click completed a browser task for $0.001 at near-instant speed.
• Bartosz Mikulski: converting 400 hand-drawn doodles into SVG coordinate text, Jev named the drawing correctly about 35% of the time versus 10% chance, but lost to Claude Sonnet 5 and guessed "airplane" for more than half.
• yuwakisa benchmark (via hypnoticfuzzwave): on a two-step benchmark of stating a principle from examples then applying it cold in a new context window, Jev scored ahead of every other tested model.
• bartlomein: oko's 108-session pilot on ranking local grep matches before an agent reads them saw up to 46% fewer agent tokens and 18% less wall time with a warm cache.
• Fatalfencer: swapping a Monte Carlo bot's option-picking step for Jev beat a "hard" difficulty tactics-game bot but not the hardest hand-tuned one, at about $0.10 per game.
• loop (dev rel @ openrouter): OpenRouter's Ori Eval judging benchmark found Jev over 5x faster than the next fastest model, and Jev's slowest requests still beat every other model's median latency.
• _FailSafe: a WIP Jev-scored retrieval triage sidecar over Noema's hybrid search kept 2 of 8 candidates on a clear query (relevant scores ~0.95/0.91 vs ~0.03-0.06, ~3700 in/666 out tokens) and 4 of 8 on a messier preference query (preference-flagged items scored 0.86-0.98), with a soft decision band identified around 0.55.
• Zawwar: playing chess by having Jev pick from a code-generated list of legal moves, it found the correct move on 16% of 600 rated Lichess puzzles versus 5% for random guessing, and accuracy did not fall as puzzles got harder (16% at 800-rated vs 23% at 2200-rated, the opposite of a human), because Jev isn't calculating; it found mate-in-one only about 1 time in 10.
• patebry: jevfish's move-by-move confidence score cannot separate a mate-in-2 from a waiting move; up a queen against a human it took 65 moves to deliver mate.
• askmuyukani: four concurrent Jev calls filtering real FAISS retrieval results in JarvisCore completed in 1.37s wall time for $0.000074.
• shitianfang: routing agent steps that need no text output straight to Jev instead of an LLM measured p50 ~230ms and ~$0.02 per 1,000 judgments.
• Zaious: on the same historical multiple-choice question, Jev gave a wrong answer at 0.90 confidence with no supporting passage, then a correct answer at 0.97 confidence once the background text was included in state; it is only as good as the facts you give it.
• goodrahstar: labelling 1,000 Android app reviews across 4 typed questions each, Jev finished in 4.6s for $0.023 versus Gemini 3.8 Flash's 18.8s and $0.158 (4.1x faster, 7x cheaper), with near-identical sentiment agreement against star ratings (rho 0.80 vs 0.82).
• JoshuaSP: an open DiffusionGemma reproduction of Jev's typed-decision interface matched 48/48 relevance labels and 6/6 top-1 retrieval on a code-retrieval eval, and hit 138/144 (95.8%) agreement with saved Jev decisions on a customer-voice eval, using a single denoising step.
• Argos1111: a local zero-shot LFM 1.2B backend scored JNLI 17% / JComQA 69% on JGLUE, versus a fine-tuned ModernBERT cross-encoder backend at JNLI 93% / JComQA 92%; zero-shot small models lag furthest on entailment judgments, not commonsense QA.
• phuryn: hardened Jev's own invoice-classification showcase to 50 documents; Jev scored 50/50 at $0.025 per 1,000 decisions, tied by Claude Haiku 4.5 and nearly matched by open 48/50 self-hosted models, while Opus 5 landed one answer behind at 115x the cost; strip the written definitions out and every Jev mistake falls below 0.80 confidence.
• 0xNatoshi: per-turn Jev routing for Codex cut cost roughly 60% against an all-frontier baseline across a 7-day replay of 237 real turns, at about $0.00003 and 0.6s per routing decision.
• yuzushi-dev: deterministic tool-output bounding (Sando) cut Codex shell tokens 90.9% across 40,000+ recorded results and Claude PostToolUse tool results 14.5-18.9% across two machines, fitting roughly 2.4x more files in a 200,000-token window.
• coldteadotai: replaying 93 real Claude Code sessions found agents break an unwritten project rule on 1 turn in 13; Jev flagged 39 edits and 15 turns, an independent reviewer confirmed 10 and 11.
• kyotofin: a Jev tax-form classifier hit 100% strict accuracy across 261 IRS forms at $0.001 per page, 34x cheaper and 6x faster than their prior LLM pipeline.
• jaz: toolgate's held-out adversarial set (injection, forged approval, encoded payloads) scored 23/24 with zero permissive errors, both frozen sets 20/20; about a second and a fraction of a cent per call.
• Suraj Phanindra: a Choice with no matching option still answers at confidence 1.00. Always include an "other/none" option.
• ubhillutd, Eddy Vinck: phrasing shifts probabilities on identical state ("save money" vs "save money long term" flips the answer). Per-call is stochastic, the average is stable.
• Zurban: calibration under domain shift is unproven. A miscalibrated gate is worse than no gate.
• parallax: 1,565 emails, 10 classes. All 737 predictions at 99%+ confidence were correct; mistakes cluster at low confidence. Trailed Gemini slightly on accuracy, 10 to 22x cheaper.
• Cleista: ~90% accuracy at ~12k classifications per dollar; the last 3% of hard cases still need an LLM at 10 to 100x the cost.
• identityTorn: zero-shot Jev within ~5 points recall of a fine-tuned Qwen 9B; the fine-tune still wins on latency and cost.
• Atanas: 94% tool-pick accuracy on typos vs 28 to 34% for keyword rules. Cannot fill open values; hand those to a small LLM (0.2s added).
• venkat (LessWrong): one yes/no call reaches ~0.97 AUROC catching backdoors, but accuracy drops once an attacker can generate and score variants.
• Sub: EU to Jev infra adds up to 320ms network latency. Killed a real-time NPC use case.
• stbenjam: no temperature setting, so repeated generative-style tasks come out repetitive.
• Tawfek Sraj: knowledge cutoff reverse-engineered to between 2024-07 and 2025-01.
• sh0rtythegreat: token counts ran unexpectedly high on a line-by-line task; check the bill, not just the per-token price.
• Sid Arya: Jev is greedy in multi-turn setups. Frame whole actions, not letter-by-letter steps.
• Yonatan Gross: hand-rolled client sent Score criteria as a dict and got silent 422s. Use the SDK.
• azakhary: 60% of a 10x-speed game playtest was Jev thinking; effective 3x. Server capacity is the bottleneck at scale.
• timpratim: an early version of a voice-controlled Mac agent, told to "clean up my desktop," moved every file on the Desktop into one folder; the project now ships an explicit safety policy after that run.
• cramforce: swapped Jev into a classifier eval that previously ran on Gemini 2.5 Flash Lite; Jev matched or beat it on quality, saturating the eval, and ran roughly 6x faster.
• Manjunath Janardhan: a 200-decision benchmark against Claude Fable 5.1, GPT-6 Astra, Kimi K3, MiniMax M3 and DeepSeek V4.1 Flash on BANKING77, BoolQ, Yelp and ChaosNLI found only the two flagships clearly ahead of Jev (11.5 and 6.5 points), while the mid-tier models were within noise but cost 6-150x more and took 2-15x longer; Jev tied every model on BoolQ at 94% (AUROC 0.970); on Yelp its errors are nearly all one star off but arrive at 0.98-0.99 confidence, with 18 of 55 errors above 0.9 confidence; on ChaosNLI, where 100 people split their labels, Jev's probabilities sat further from the human distribution (JSD 0.149) than a blind guess (0.127), while Claude Fable 5.1 hit 0.043.
• ASHu2: an independent 8-dataset benchmark against classical ML found Jev strong on sentiment classification but mixed across other tasks, with fine-tuned ML staying cheaper wherever fine-tuning is an option; few-shot prompting did not reliably improve Jev's results.
• yzfly (awesome-jev-zh): an independent evaluation cited in the list found that asking Jev a single direct question about a phishing email scored 62.6% accuracy, while two lines of regex reached 91.8%.