Submits binary, categorical and ordered-score questions and returns typed answers with probabilities.
Directory.337
Submit a projectConverts strategy questions to System One requests and normalizes native answers to the project’s result format.
ai-hedge-fund
JEV ROLEConverts strategy questions to System One requests and normalizes native answers to the project’s result format.
Maps requests to configured complexity classes that drive backend routing.
Sends agent state and typed questions to Jev and parses structured answers.
Evaluates task tier, reasoning needs and production risk; local policy maps results to invocation settings.
jev-model-router
JEV ROLEEvaluates task tier, reasoning needs and production risk; local policy maps results to invocation settings.
Turns tool or action conditions into structured questions and passes Jev answers to local invocation logic.
Maps choice, score, and yes/no questions to TypeSafe System One requests and parses typed results.
Reads DOM or supported visual-region descriptions and returns a supplied candidate action ID.
Only an explicit systemOne call sends state and questions, returning validated typed answers.
Uses invoke to call TypeSafe and parse choice, noul, score and probability fields.
langchainjs
JEV ROLEUses invoke to call TypeSafe and parse choice, noul, score and probability fields.
Selects components and layout from a predefined catalog for a renderer to display.
json-render
JEV ROLESelects components and layout from a predefined catalog for a renderer to display.
Jev selects a task category; local category mappings determine the model configuration.
openchamber
JEV ROLEJev selects a task category; local category mappings determine the model configuration.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
rig-typesafeai
JEV ROLESends application state and questions to Jev and parses Choice, Score or Noul answers.
Sends the task brief and candidate rules to Jev, then resolves execution profiles with confidence and local conditions.
Selects the next browser operation and target element; a separate model writes free text.
Jev Ultrafast
JEV ROLESelects the next browser operation and target element; a separate model writes free text.
Jev scores jailbreaks, harmful content and secret disclosure; thresholds or evaluation errors reject requests.
agentgateway
JEV ROLEJev scores jailbreaks, harmful content and secret disclosure; thresholds or evaluation errors reject requests.
Judges which checks apply and can add checks when thresholds and rate limits permit.
latitude-llm
JEV ROLEJudges which checks apply and can add checks when thresholds and rate limits permit.
Selects stale tool calls and results for removal while keeping useful original text.
fast-jev-compaction
JEV ROLESelects stale tool calls and results for removal while keeping useful original text.
Maps supported signatures to Jev questions or sends native System One requests.
Submits state and question bundles to System One and returns structured answers to the Assistant.
vellum-assistant
JEV ROLESubmits state and question bundles to System One and returns structured answers to the Assistant.
Jev selects a target and action and estimates presence and risk; local policy decides whether to execute.
jev-desktop
JEV ROLEJev selects a target and action and estimates presence and risk; local policy decides whether to execute.
Judges whether a finding was introduced by the change, is supported and merits a fix.
Chooses a buy or sell direction from an order book; code manages orders and controls.
Jev Trader
JEV ROLEChooses a buy or sell direction from an order book; code manages orders and controls.
Sends predefined questions to Jev and decodes answers for callers; regular models still generate code.
Assesses utterance completeness and whether the user is likely to continue speaking.
aiavatarkit
JEV ROLEAssesses utterance completeness and whether the user is likely to continue speaking.
Evaluates multiple questions and dynamic candidate spaces concurrently in a single forward pass, logging navigation choices.
Sends a Score question per candidate, reorders and drops low scores; the model id is typesafe/jev.
Adds a destructive-risk judgment after local shell checks and refuses commands above the configured threshold.
Sends state and questions, normalizes answers and retains the raw provider response.
Chooses controls and checks completion or blockage before the test runner operates the UI.
After deterministic routing guards, Jev judges the remaining candidates and required workflow components.
vexjoy-agent
JEV ROLEAfter deterministic routing guards, Jev judges the remaining candidates and required workflow components.
Judges coding-task completion, test results and the need for human review.
Jev evaluates the task against eligible specialists; local dispatch rules use the result.
WrongStack
JEV ROLEJev evaluates the task against eligible specialists; local dispatch rules use the result.
Attaches a parallel decision head to an open 0.5B model to answer discrete questions directly from token activations.
Selects the next desktop action using OCR text; code operates the mouse.
typesafe-computer-use
JEV ROLESelects the next desktop action using OCR text; code operates the mouse.
Makes focused judgments about diffs and codebases to guide staged review.
Classifies the first task prompt and branch state by work type; local policy accepts the result or falls back.
orchestkit
JEV ROLEClassifies the first task prompt and branch state by work type; local policy accepts the result or falls back.
Chooses a legal controller action from structured emulator state.
Judges risks such as exposed credentials or destructive changes; local rules warn or block a commit.
jev-experiments
JEV ROLEJudges risks such as exposed credentials or destructive changes; local rules warn or block a commit.
Sends shared state and typed questions to Jev and parses structured answers.
openai-scala-client
JEV ROLESends shared state and typed questions to Jev and parses structured answers.
Programmed questions feed Jev judgments into local action logic under runtime budgets.
Makes bounded transcript judgments for checkpoints, note-taking actions or screening before remote text cleanup.
OpenWhisper
JEV ROLEMakes bounded transcript judgments for checkpoints, note-taking actions or screening before remote text cleanup.
Sends row content to Jev for a match decision, category, or score used by SQL conditions and ordering.
Assesses inputs such as test names, documentation claims and supporting evidence.
interlinked-cli
JEV ROLEAssesses inputs such as test names, documentation claims and supporting evidence.
Judges each coding turn and routes it to an appropriate model tier.
Extracts log-probabilities of candidate tokens from model vocabulary logits, formatting them into standard Jev responses.
simple-jev
JEV ROLEExtracts log-probabilities of candidate tokens from model vocabulary logits, formatting them into standard Jev responses.
Jev scores task relevance among skills the client is allowed to access.
Chooses the next action from Android screen state.
evaluate, choice, score and noul actions submit questions and retain answers and usage.
Extracts PDF page text and sends it to Jev for form type, page type, and confidence.
tax-doc-classifier
JEV ROLEExtracts PDF page text and sends it to Jev for form type, page type, and confidence.
Selects sources, time ranges and queries, then ranks retrieved results.
systemOne submits state and named questions and infers answer types from the questions.
typesafe-sdk-js
JEV ROLEsystemOne submits state and named questions and infers answer types from the questions.
Sends page state and permitted candidates to Jev; the existing browser connection executes the actions.
jev-browser-use
JEV ROLESends page state and permitted candidates to Jev; the existing browser connection executes the actions.
Forwards authorized structured requests to TypeSafe and returns answers with usage accounting.
Scores code quality across correctness, complexity and changeability.
Jev answers probability, choice and score questions for application transitions or model selection.
effect-agent
JEV ROLEJev answers probability, choice and score questions for application transitions or model selection.
Chooses DOM actions and judges goal completion or stalled progress; a separate model can supply typed text.
jev-browser
JEV ROLEChooses DOM actions and judges goal completion or stalled progress; a separate model can supply typed text.
Identifies unnecessary page elements so the extension can reuse hiding rules.
system_one sends context and choice, score or yes/no questions to the API, then parses answers by question name.
typesafe-sdk-python
JEV ROLEsystem_one sends context and choice, score or yes/no questions to the API, then parses answers by question name.
Selects a tool before each model step while a language model fills arguments.
jev-eval-agent
JEV ROLESelects a tool before each model step while a language model fills arguments.
Chooses intent, elements, URLs or verbatim spans and judges command completeness and sensitive actions.
jev-voice-browser
JEV ROLEChooses intent, elements, URLs or verbatim spans and judges command completeness and sensitive actions.
When experimental AI is enabled, Jev evaluates replies missed by rules and local thresholds decide whether to hide them.
Sends state and typed questions to Jev and returns probabilities for branching.
typesafe-mcp
JEV ROLESends state and typed questions to Jev and returns probabilities for branching.
Jev reads road and traffic state and selects motion; geometry, collision prediction and braking remain local.
Converts SQL input into System One requests and returns answers through database functions.
pg_typesafe
JEV ROLEConverts SQL input into System One requests and returns answers through database functions.
Combines yes/no, Choice and Score questions and maps responses to typed values.
Provides claim verification, injection screening and semantic ranking to agents.
jev-mcp (jkudish)
JEV ROLEProvides claim verification, injection screening and semantic ranking to agents.
Sends scan rules, matched snippets, and file context to Jev to assess likely true positives and suggested handling.
captaincore
JEV ROLESends scan rules, matched snippets, and file context to Jev to assess likely true positives and suggested handling.
Judges tool calls and outputs and offers a typed question tool for Pi.
Checks writes against project rules and reports violations to the agent.
Makes tactical drone judgments from structured state; code retains flight control.
Identifies sponsored segments in transcripts so code can skip them during playback.
Sponsor Skip
JEV ROLEIdentifies sponsored segments in transcripts so code can skip them during playback.
Scores transcribed sentences for an overlay on a video.
Ranks shell history against current input to suggest completions.
Selects a model, reasoning depth and speed setting for each Codex turn.
Batches about thirty lines and asks jev-latest a Score or Noul per line for the stated meaning.
jev-semgrep
JEV ROLEBatches about thirty lines and asks jev-latest a Score or Noul per line for the stated meaning.
Judges whether DOM elements are ads before the extension hides them.
Chooses targets and actions and assesses completion and risk; local policy controls execution or confirmation.
Asks Jev about file-quality properties; code composes scores and ordering.
Asks whether predefined traits are present or how strongly they appear, recording scores, latency and Token usage.
goodwatch-monorepo
JEV ROLEAsks whether predefined traits are present or how strongly they appear, recording scores, latency and Token usage.
A shared Client sends state and typed questions and parses choices, probabilities, scores and usage.
ruby_decision_model
JEV ROLEA shared Client sends state and typed questions and parses choices, probabilities, scores and usage.
Sends workflow state and named questions to Jev and returns structured answers to agents.
synkora-ai
JEV ROLESends workflow state and named questions to Jev and returns structured answers to agents.
Evaluates entry distribution, toxic flow, holding and stress signals for calibration logs.
Maps the same tasks to Jev Choice/Noul questions and normalizes answers to a shared result format.
typesafe-ai-benchmark
JEV ROLEMaps the same tasks to Jev Choice/Noul questions and normalizes answers to a shared result format.
Sends the draft and reply or quote context to Jev for multiple scores and a posting suggestion.
One System One request per row: Noul, Choice, or Score answers land as ordinary typed columns with confidences.
polar_llama
JEV ROLEOne System One request per row: Noul, Choice, or Score answers land as ordinary typed columns with confidences.
Matches incoming requests to routes described in natural language.
Scores Discord messages for spam, scams and social-engineering signals.
Jev Moderation Bot
JEV ROLEScores Discord messages for spam, scams and social-engineering signals.
Jev makes a Choice; the router separately checks availability, permissions, risk and confirmation policy.
Converts the current message into decision questions and receives answers from TypeSafe or a compatible Decisions endpoint.
Assesses the first task’s capability needs; local policy chooses a same-gateway model and keeps the route sticky.
Classifies commit diffs by change type, bug fixes and security issues.
Jev uses Choice to classify pages and Noul to assess careers links, role relevance, filter controls and candidate fit; local thresholds select navigation, job details and saved results.
Selects graph relationships for a search algorithm to follow in Neo4j.
Scores file and directory relevance to guide codebase search.
Connects typed Jev judgments to ACP and MCP workflows.
After length and content checks, Jev judges which chunks to retain while original output is archived.
jev-pruner
JEV ROLEAfter length and content checks, Jev judges which chunks to retain while original output is archived.
Jev answers scoring, category and clarity questions; local weights and gates calculate the final label.
killmyidea
JEV ROLEJev answers scoring, category and clarity questions; local weights and gates calculate the final label.
Supplies board state, fruit location, and legal direction choices to Jev, selecting the optimal move direction each tick.
typesafe-snake
JEV ROLESupplies board state, fruit location, and legal direction choices to Jev, selecting the optimal move direction each tick.
Jev scores whether Read, Bash and Grep output blocks are needed; local thresholds keep relevant or uncertain blocks.
Evaluates smart-home entity state to produce sensor values and automation inputs.
Experiments send action candidates or typed questions to Jev, then execute or record the answers.
jev-playground
JEV ROLEExperiments send action candidates or typed questions to Jev, then execute or record the answers.
Jev selects search phrases and highlights references using titles and descriptions, not image pixels.
Intercepts request payloads and uses Jev to score historical messages, discarding outdated tool noise before forwarding to upstream LLMs.
Chooses tactics and unit actions from structured StarCraft state.
Judges how candidate tools, skills, and history relate to the task so local rules can load or retain them.
Adds batched typed evaluation tools to Pi through a shared TypeSafe client.
Assesses paragraphs against style rules such as redundant closings, clichés and excessive hedging.
Sends state and typed questions to TypeSafe, then delivers answers as messages.
Converts input state and options into structured questions, returning results or risk scores for local policy.
Implements the /v1/systemone endpoint (Choice, Score, Noul) by calculating candidate token log-probabilities without text generation.
Chooses concrete browser actions from Playwright page state.
Judges whether shell, write and edit actions may proceed, failing closed on uncertainty.
pi-jev-auto-mode
JEV ROLEJudges whether shell, write and edit actions may proceed, failing closed on uncertainty.
Builds Noul, Choice and Score questions with Schema and returns answers to the RubyLLM application.
ruby_llm-typesafe
JEV ROLEBuilds Noul, Choice and Score questions with Schema and returns answers to the RubyLLM application.
Each asset account sends market state to Jev; execution code places or cancels orders based on the decisions.
Jev reviews completed minute candles and indicators; Python applies confidence and position limits to a simulated portfolio.
Jev-Trades
JEV ROLEJev reviews completed minute candles and indicators; Python applies confidence and position limits to a simulated portfolio.
Feeds file contents and security criteria into Jev to assess risks such as data exfiltration, obfuscated loaders, and credential leakage.
is-malicious
JEV ROLEFeeds file contents and security criteria into Jev to assess risks such as data exfiltration, obfuscated loaders, and credential leakage.
Runs the same labeled text tasks through both backends and records probabilities, latency and failures.
jev-benchmarks
JEV ROLERuns the same labeled text tasks through both backends and records probabilities, latency and failures.
Selects a fixed workflow and makes structured judgments over bounded diffs, code, or log excerpts.
Exposes classification, scoring, matching and screening as agent tools.
jev-mcp (blakestone-x)
JEV ROLEExposes classification, scoring, matching and screening as agent tools.
Sends state and typed questions to System One and parses answers for terminal workflows.
Submits custom states and questions to Jev via Next.js API routes, displaying discrete probabilities, latency, and confidence scores.
typesafe-ai-playground
JEV ROLESubmits custom states and questions to Jev via Next.js API routes, displaying discrete probabilities, latency, and confidence scores.
Judges support messages or chooses lanes and target speed from structured simulated sensors.
typesafe-playground
JEV ROLEJudges support messages or chooses lanes and target speed from structured simulated sensors.
Asks Choice, Score, and Noul questions about row text and maps answers to SQL types defined by the questions.
duckdb-jev
JEV ROLEAsks Choice, Score, and Noul questions about row text and maps answers to SQL types defined by the questions.
Judges injection and dangerous-action signals for configurable guard policies.
Sends multiple Jev questions over shared state and parses typed answers, probabilities and usage.
typesafe-ai
JEV ROLESends multiple Jev questions over shared state and parses typed answers, probabilities and usage.
Scores diagnostic value and priority to route logs to later models.
Jev answers lane, speed, hazard and pedestrian-yield questions; local rules apply the actions.
Validates requests, calls System One, and decodes typed answers.
Compares the request with the skill roster; inserts a suggestion only when suitable and abstains on failure.
typesafe-skill-router
JEV ROLECompares the request with the skill roster; inserts a suggestion only when suitable and abstains on failure.
Combines semantic conditions from one parse into Jev Noul questions and maps probabilities to validation results.
Screens and verifies code relevance before whole files enter the agent context.
Uses inferred types and labeled handlers to connect Choice answers to predefined program branches.
Chooses goal actions, tick interventions and the next polling interval from game state.
Sends Jev questions through systemOne and exposes answers as Swift types or dynamic maps.
swift-typesafe
JEV ROLESends Jev questions through systemOne and exposes answers as Swift types or dynamic maps.
Chooses guard actions and reactions from structured game state.
Choice selects the next word; Noul evaluates candidate chunks and stopping conditions.
Jev selects controls or groups from locally observed candidates; coordinates and input execution stay on the Mac.
jev-macos-loop
JEV ROLEJev selects controls or groups from locally observed candidates; coordinates and input execution stay on the Mac.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Runs fixed questions and measures accuracy, calibration, coverage and escalation rates.
Uses Jev for ranking, verification, action checks and next-step selection, with code applying policy.
Each decision tick asks about movement, view, aiming, firing, and jumping, with a heuristic fallback on API failure.
OneVOneJev
JEV ROLEEach decision tick asks about movement, view, aiming, firing, and jumping, with a heuristic fallback on API failure.
Sends input to Jev and applies local probability thresholds to outputs and exit codes.
Submits typed questions, validates answers, and lets domain rules choose registered tools with permission and argument checks.
Wraps System One API endpoints in TypeSafeClient to issue parallel question evaluations and parse responses into immutable C# records.
typesafeai-dotnet-sdk
JEV ROLEWraps System One API endpoints in TypeSafeClient to issue parallel question evaluations and parse responses into immutable C# records.
Jev evaluates bounded questions in the background while Hermes retains reasoning and execution responsibilities.
hermes-jev
JEV ROLEJev evaluates bounded questions in the background while Hermes retains reasoning and execution responsibilities.
Jev returns APPROVE, DENY or ESCALATE, with local thresholds and policy determining handling.
hermes-jev-approvals
JEV ROLEJev returns APPROVE, DENY or ESCALATE, with local thresholds and policy determining handling.
Classifies email as invoice-related or general and routes it to a handler.
Jev email intent workflow
JEV ROLEClassifies email as invoice-related or general and routes it to a handler.
Submits noul, choice and score questions and emits JSON or a primary value.
Sends input and bounded answer options to Jev, then reports judgments and probabilities using local thresholds.
Chooses from bounded syntax and action options; local code constructs programs or invokes tools.
Selects click, type or select actions from an indexed ego lite element table.
Invokes Jev judges to grade multi-level relevance on bilingual queries, evaluating ranking metrics and judge-circularity effects.
jev-search-rerank-eval
JEV ROLEInvokes Jev judges to grade multi-level relevance on bilingual queries, evaluating ranking metrics and judge-circularity effects.
For each new pipe, Jev selects a route and the game executes its scheduled flaps.
Selects form, instruments, chords, and rhythm candidates; local code renders the choices as notes.
Judges whether calls and outputs should survive in the context sent to the model.
pi-fast-jev-compaction
JEV ROLEJudges whether calls and outputs should survive in the context sent to the model.
Posts state plus Choice, Score, or Noul questions to `/v1/systemone`, defaulting to jev-latest.
Analyzes live speech transcripts with Jev to judge whether the current slide's topics are covered, triggering slide transitions.
slidepilot
JEV ROLEAnalyzes live speech transcripts with Jev to judge whether the current slide's topics are covered, triggering slide transitions.
Jev evaluates candidate capabilities, outputs and task state; per-feature switches control applying or recording advice.
Builds Choice, Score and Noul questions and exposes parsed answers by type.
Selects actions from current page controls and assesses risk and irreversibility.
Builds Noul, Choice and Score questions, then combines returned probabilities through local functions.
Sends controlled variants of fixed tasks and records choices, probabilities and raw request-response evidence.
jev-behavior-study
JEV ROLESends controlled variants of fixed tasks and records choices, probabilities and raw request-response evidence.
Compares stepwise Choice decoding with selection from complete candidate replies.
Judges dataset rows so the pipeline can retain or reject samples.
Calls TypeSafe with Choice, Score and Noul questions and returns structured answers.
Classifies search intent, direct answers and content gaps, with a custom citation-likelihood rubric.
Jev selects the next action and approves the concrete call; tool results feed the next state.
Designs Noul, Choice, and Score question packs for an application, with a script for API-backed case runs.
The CLI or service layer evaluates jev_* calls over row text before completing the query.
Sends row data and typed questions to Jev and maps answers into queryable SQL results.
Jev judges relevance; local policy runs tests on uncertainty, API failure or changes to the test file itself.
One Noul relevance question per document; long lists are chunked and scored, then filtered by threshold or sorted.
Jev judges prompt injection or unsafe content; local rules forward, flag or block the request.
safer-with-jev
JEV ROLEJev judges prompt injection or unsafe content; local rules forward, flag or block the request.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
typesafe-ai-rs
JEV ROLESends application state and questions to Jev and parses Choice, Score or Noul answers.
Sends Java state and Noul, Choice, and Score questions to System One and parses typed results.
typesafe-sdk-java
JEV ROLESends Java state and Noul, Choice, and Score questions to System One and parses typed results.
Breaks requests into choice and score questions and maps answers to a bounded component catalog.
ui-generator-instinct-jev
JEV ROLEBreaks requests into choice and score questions and maps answers to a bounded component catalog.
Encodes state and typed questions as a System One request and decodes Probability, Choice distributions, and Score values.
zio-typesafe-ai
JEV ROLEEncodes state and typed questions as a System One request and decodes Probability, Choice distributions, and Score values.
Scores candidate defects, explanations or options against preset questions and local policies.
claude-jev
JEV ROLEScores candidate defects, explanations or options against preset questions and local policies.
Uses window state to select an action, target, text-input need, and completion status.
computer-use-jev
JEV ROLEUses window state to select an action, target, text-input need, and completion status.
Scores X reply text with Jev Noul questions for sexual content, solicitation, and spam, then combines probabilities with local thresholds and structural signals.
Assesses past conversation turns with Jev when context limits approach, selecting critical tool outputs and user requirements for retention.
fast-dev-compaction
JEV ROLEAssesses past conversation turns with Jev when context limits approach, selecting critical tool outputs and user requirements for retention.
Benchmarks Jev response accuracy and latency via a TypeSafe provider integration, deriving empirical routing policies.
Selects legal chess moves or judges whether an utterance addresses each NPC.
Classifies text that passes local gates; local mapping decides whether to allow or suppress it.
jev-block-android-ad
JEV ROLEClassifies text that passes local gates; local mapping decides whether to allow or suppress it.
Jev chooses actions, matches form fields or extracts structured content from page observations; Playwright executes.
jev-browser
JEV ROLEJev chooses actions, matches form fields or extracts structured content from page observations; Playwright executes.
Maps attack conditions and impacts to discrete metrics; local CVSS formulas calculate the numeric score.
Jev classifies engineering text and candidates; calculations and final actions stay in Python rules.
jev-for-engineers
JEV ROLEJev classifies engineering text and candidates; calculations and final actions stay in Python rules.
Jev selects observed controls and operations; test code owns expected values and pass criteria.
jev-frontend-qa
JEV ROLEJev selects observed controls and operations; test code owns expected values and pass criteria.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
Delegates bounded coding judgments to Jev while the main model writes code.
Jev classifies change evidence against configured rules; code maps answers to review outcomes.
Jev selects candidate line IDs; code copies the original text with file and location references.
jev-reviewer
JEV ROLEJev selects candidate line IDs; code copies the original text with file and location references.
Batches candidate DOM elements for TypeSafe and uses Noul probabilities with thresholds to decide which to collapse.
jev-shield
JEV ROLEBatches candidate DOM elements for TypeSafe and uses Noul probabilities with thresholds to decide which to collapse.
Assesses skills against the project stack, directory, and README before adjusting description visibility.
jev-skill-gate
JEV ROLEAssesses skills against the project stack, directory, and README before adjusting description visibility.
Judges response mode, depth, uncertainty and draft quality inside a LangGraph workflow.
jev-system-one
JEV ROLEJudges response mode, depth, uncertainty and draft quality inside a LangGraph workflow.
Traverses a taxonomy through successive choices to handle large candidate sets.
Encodes classification or evidence checks as Noul, Choice, or Score questions sent to TypeSafe.
jev-workbench
JEV ROLEEncodes classification or evidence checks as Noul, Choice, or Score questions sent to TypeSafe.
Selects a direction-button target from allowed controls using structured board state.
Employs Jev to classify and select the next tool action based on execution state, delegating parameter generation to DeepSeek.
Wraps a community PHP SDK for typed judgments and preserves promises and exceptions.
laravel-typesafe-jev
JEV ROLEWraps a community PHP SDK for typed judgments and preserves promises and exceptions.
Judges flight diversions, emergencies and landing priority.
Uses Score for passage relevance and Choice for selecting query engines or tools.
llama-index-jev
JEV ROLEUses Score for passage relevance and Choice for selecting query engines or tools.
Sends proposed actions and permission rules to Jev, then uses local logic to allow, deny, or ask the user.
oc-plugins
JEV ROLESends proposed actions and permission rules to Jev, then uses local logic to allow, deny, or ask the user.
Jev judges tool-call and output utility; the extension truncates selected content with recovery notes.
omp-jev-compaction
JEV ROLEJev judges tool-call and output utility; the extension truncates selected content with recovery notes.
Scores historical fragments and hides low-scoring content from future requests while preserving session history.
pi-jev-context
JEV ROLEScores historical fragments and hides low-scoring content from future requests while preserving session history.
Runs Choice, Score and Noul over samples and uses measured results to design local thresholds.
Uses declarative YAML policies evaluated by Jev discrete classification without requiring pre-collected training data.
Jev classifies requests and matches skills, with heuristic fallback when unavailable.
Builds state and Noul, Choice, and Score requests and parses System One responses.
typesafe_sdk
JEV ROLEBuilds state and Noul, Choice, and Score requests and parses System One responses.
Builds requests from state and Noul, Choice or Score questions, then validates and parses responses.
typesafe-go
JEV ROLEBuilds requests from state and Noul, Choice or Score questions, then validates and parses responses.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
typesafe-go
JEV ROLESends application state and questions to Jev and parses Choice, Score or Noul answers.
Builds Jev requests and handles authentication, transport and structured response parsing.
typesafe-sdk-rust
JEV ROLEBuilds Jev requests and handles authentication, transport and structured response parsing.
The Agent supplies action candidates; Jev picks an ID, then the Agent executes and verifies it through Aside.
Sends selected records to Jev for judgments about execution claims and evidence sufficiency.
Jev can identify completion claims and semantic-rule issues; stop gates depend on ledger facts and local rules.
Judges relevance of eligible read-only tool records; local rules protect required content and assemble a handoff packet.
codex-jev-compaction
JEV ROLEJudges relevance of eligible read-only tool records; local rules protect required content and assemble a handoff packet.
Jev selects tactical goals and actions while local geometry handles fine aiming.
Local checks handle machine-verifiable requirements; Jev assesses the remaining semantic conditions.
hermes-jev-north-star
JEV ROLELocal checks handle machine-verifiable requirements; Jev assesses the remaining semantic conditions.
Uses Jev Choice to compare bounded options and return probabilities for further review.
jev_ampcode
JEV ROLEUses Jev Choice to compare bounded options and return probabilities for further review.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
Builds candidate sets from traces and submits three choice questions.
jev-agent-failure-benchmark
JEV ROLEBuilds candidate sets from traces and submits three choice questions.
Maps natural-language goals to predefined robot-arm primitives.
Four questions: whether the closing message claims done, claims checks passed, whether verification applies, and a complete / partial / blocked / other choice.
Chooses from locally generated legal chess moves; Stockfish scores are shown to the operator rather than Jev.
jev-broadcast-lab
JEV ROLEChooses from locally generated legal chess moves; Stockfish scores are shown to the operator rather than Jev.
Jev selects among available tools; adapters may record suggestions or influence actual tool selection.
jev-classifier
JEV ROLEJev selects among available tools; adapters may record suggestions or influence actual tool selection.
POSTs https://api.typesafe.ai/v1/systemone as jev-latest with a closed action Choice each step.
Evaluates typed questions over one input and returns choices and probabilities as JSON.
Judges passage relevance and withholds No and Unknown results.
Selects tactical Doom actions that a local controller translates into controls.
jev-doom-agent
JEV ROLESelects tactical Doom actions that a local controller translates into controls.
Calls Jev on defined questions and labeled cases, then analyzes errors, calibration, and difficulty effects.
jev-exploration
JEV ROLECalls Jev on defined questions and labeled cases, then analyzes errors, calibration, and difficulty effects.
Jev checks the diff, selects evidence from actual hunks and adjudicates candidate findings.
jev-flash-review
JEV ROLEJev checks the diff, selects evidence from actual hunks and adjudicates candidate findings.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
Jev chooses from locally generated move candidates while the experiment varies board and tactical information.
jev-gomoku
JEV ROLEJev chooses from locally generated move candidates while the experiment varies board and tactical information.
Classifies location, narrative beat, mood, danger and whether text remains in fiction.
Runs typed questions and uses labeled samples to measure accuracy and calibration.
Submits multiple Noul, Choice, and Score questions in one request.
A Choice over legal actions at each branch; battle turns also ask Noul faints_this_turn and should_flee; failures take the code default.
jev-plays-pokemon-red
JEV ROLEA Choice over legal actions at each branch; battle turns also ask Noul faints_this_turn and should_flee; failures take the code default.
Jev first assesses whether a closed decision is possible, then selects among the target skill’s outcomes.
jev-predict-skill
JEV ROLEJev first assesses whether a closed decision is possible, then selects among the target skill’s outcomes.
Ranks candidate passages with Choice, Noul and rubric scores, then computes retrieval metrics.
jev-rerank-bench
JEV ROLERanks candidate passages with Choice, Noul and rubric scores, then computes retrieval metrics.
Uses task and repository state to choose an executor and assess risk and dispatch readiness.
jev-research
JEV ROLEUses task and repository state to choose an executor and assess risk and dispatch readiness.
Checks whether a nominally successful response actually contains a failure.
Answers policy questions over pinned repository evidence or PR patches, then applies classification rules.
jev-review-action
JEV ROLEAnswers policy questions over pinned repository evidence or PR patches, then applies classification rules.
Compares candidate information with the request, scores fit and maintenance signals, and chooses a candidate.
One Choice per caption segment among content, sponsor, intro, outro, self_promo, recap, and other. Painted categories sponsor, self_promo, intro, outro, and recap skip when probability is at or above the threshold; content and other never skip.
Jev returns judgments; application policies choose automatic handling, fallback or human review.
jev-starter
JEV ROLEJev returns judgments; application policies choose automatic handling, fallback or human review.
Invokes Jev at workflow checkpoints to evaluate change validity, test sufficiency, and package safety before advancing agent state.
jev-superpowers
JEV ROLEInvokes Jev at workflow checkpoints to evaluate change validity, test sufficiency, and package safety before advancing agent state.
Asks Choice and Noul questions about eligibility, then combines them into include or exclude decisions.
jev-synergy-screening
JEV ROLEAsks Choice and Noul questions about eligibility, then combines them into include or exclude decisions.
Deterministic rules run first; Jev assesses remaining call risks and tool relevance for threshold-based handling.
jev-tool-permissions
JEV ROLEDeterministic rules run first; Jev assesses remaining call risks and tool relevance for threshold-based handling.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
jev-web-analyzer
JEV ROLEJev returns a structured decision for the local program; consult the source for the exact decision policy.
Maps answer styles to Choice options and selects chat titles from fragments of the question.
Calls TypeSafe through aiohttp and parses choices, scores and probabilities into objects.
Jev classifies posts with Choice, rates signal, actionability and originality with Score, and assigns Noul labels; local thresholds and noise rules decide whether to fold each post.
JevFilterForX
JEV ROLEJev classifies posts with Choice, rates signal, actionability and originality with Score, and assigns Noul labels; local thresholds and noise rules decide whether to fold each post.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
POSTs to /v1/systemone as jev-latest: a goal Noul and a Choice over registered Flutter actions.
Uses a deterministic local engine by default; a TypeSafe key enables real Jev judgments over features.
When configured with a key, sends project state and candidate categories to Jev; otherwise uses local classification rules.
jevsome-projects
JEV ROLEWhen configured with a key, sends project state and candidate categories to Jev; otherwise uses local classification rules.
Jev evaluates extracted resume text against typed questions; a Worker combines the results into review output.
Jev judges complete read-only tool pairs; local code removes those rated unnecessary.
pi-jev-compaction
JEV ROLEJev judges complete read-only tool pairs; local code removes those rated unnecessary.
Before each tool call, Jev receives the user's request (or pinned task), the proposed command, the agent's explanation, and recent conversation. It answers a Choice for intent (on task / off task / injected / needs more context) and a Score for risk (safe / needs approval / harmful). Local code turns these into run, ask the user, or warn. If Jev says more context would help, the extension adds it and asks again.
pi-jev-sentinel
JEV ROLEBefore each tool call, Jev receives the user's request (or pinned task), the proposed command, the agent's explanation, and recent conversation. It answers a Choice for intent (on task / off task / injected / needs more context) and a Score for risk (safe / needs approval / harmful). Local code turns these into run, ask the user, or warn. If Jev says more context would help, the extension adds it and asks again.
Jev suggests skills, reranks actual source candidates and triages failures without automatic retries or permission approval.
Rules become at most twelve questions (`MAX_JEV_QUESTIONS` in src/types.ts). AKIA and private-key armor from src/redact.ts fail locally in src/pipeline.ts with no Jev call.
Preserves probabilities and option types and routes uncertainty to a caller-supplied fallback.
Sends text to Jev for separate Noul probabilities, scores or category choices.
TypeSafe AI Playground
JEV ROLESends text to Jev for separate Noul probabilities, scores or category choices.
Dispatches state and criteria from ActiveRecord models or services to Jev System One endpoints, logging structured outcomes.
typesafe-ai-rails
JEV ROLEDispatches state and criteria from ActiveRecord models or services to Jev System One endpoints, logging structured outcomes.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
typesafe-go
JEV ROLESends application state and questions to Jev and parses Choice, Score or Noul answers.
Sends application state and questions to Jev and parses Choice, Score or Noul answers.
typesafe-sdk-php
JEV ROLESends application state and questions to Jev and parses Choice, Score or Noul answers.
Scores relevance, substance, usefulness, and promotion; local weights and thresholds control highlighting, collapsing, or hiding.
your-signal
JEV ROLEScores relevance, substance, usefulness, and promotion; local weights and thresholds control highlighting, collapsing, or hiding.
Choice clean / suspicious_pua / malicious, Score 0–4, and Noul questions for packing, C2, persistence, and injection; local policy maps the answers.
Jev selects operations from a DOM target table; a separate text model supplies typed text when needed.
cline-plugin-jev-browser
JEV ROLEJev selects operations from a DOM target table; a separate text model supplies typed text when needed.
Selects only current snapshot candidates and supported actions; another model can supply typed text.
Jev returns a structured decision for the local program; consult the source for the exact decision policy.
fast-jev-compaction-pi
JEV ROLEJev returns a structured decision for the local program; consult the source for the exact decision policy.
Jev assesses progress and completion signals; deterministic acceptance checks still decide whether the work passes.
foreman-jev
JEV ROLEJev assesses progress and completion signals; deterministic acceptance checks still decide whether the work passes.
Submits typed questions through a separate client for tools, turn routing and a mail-triage example.
Jev classifies the utterance and exposed devices; code chooses a lighting service or conversation path.
ha-conversation-jev
JEV ROLEJev classifies the utterance and exposed devices; code chooses a lighting service or conversation path.
Collects Noul and Choice outputs and compares them with labels for error, accuracy and stability.
jev-calibration-audit
JEV ROLECollects Noul and Choice outputs and compares them with labels for error, accuracy and stability.
One systemOne call with a should_play Noul plus card, lane, and depth Choices.
jev-clash-royale-test
JEV ROLEOne systemOne call with a should_play Noul plus card, lane, and depth Choices.
Five Noul questions in one request: checkable message, hunk match, debug leftovers, unmentioned work, and credential-shaped added lines. Default blocking is the regex belt; secret_shaped participates only under --strict.
jev-commit
JEV ROLEFive Noul questions in one request: checkable message, hunk match, debug leftovers, unmentioned work, and credential-shaped added lines. Default blocking is the regex belt; secret_shaped participates only under --strict.
TypeSafeClient.systemOne answers policy questions; the adapter then decides whether a generative model runs.
jev-decision-gateway
JEV ROLETypeSafeClient.systemOne answers policy questions; the adapter then decides whether a generative model runs.
Assesses classification levels, human-handoff intent and frustration, with low-confidence escalation.
Chooses moves from direction candidates, then checks collisions, paths and completion.
Collects judgments on matched tasks and computes confidence intervals and repeat-input stability.
Chooses jumping, ducking, movement, aim, or firing from game state, with local execution rules.
jev-experiments
JEV ROLEChooses jumping, ducking, movement, aim, or firing from game state, with local execution rules.
POSTs /v1/systemone as jev-latest with one type:choice question per tree level.
Chooses voices, pitches, durations, or chords; code grades exercises or synthesizes playback.
jev-music-theory-1
JEV ROLEChooses voices, pitches, durations, or chords; code grades exercises or synthesizes playback.
For each text chunk, asks twelve PII-category Nouls and a three-level sensitivity Score.
jev-pii-checker
JEV ROLEFor each text chunk, asks twelve PII-category Nouls and a three-level sensitivity Score.
Reads structured telemetry and selects contact and pace through Choice before code executes inputs.
jev-play-ping-pong
JEV ROLEReads structured telemetry and selects contact and pace through Choice before code executes inputs.
Chooses among candidate targets, grasp or release, and a done Noul.
Selects a model from the task and candidate profiles and records probabilities and runs.
jev-router-playground
JEV ROLESelects a model from the task and candidate profiles and records probabilities and runs.
Jev answers bounded direction and risk questions; local policy decides simulated positions.
One System One request asks Nouls for worth-engaging plus optional tags such as spam, buzz, and AI-ish.
jev-tweet-radar
JEV ROLEOne System One request asks Nouls for worth-engaging plus optional tags such as spam, buzz, and AI-ish.
Scores each Treesitter function against the user question, splitting across requests if the buffer does not fit. Every hit is listed in quickfix; only hits at or above the threshold get buffer marks.
Chooses CLICK, TYPE_TEXT, SELECT, SCROLL_DOWN, SCROLL_UP, PRESS_ENTER, WAIT, DONE, or BLOCKED and the matching DOM target in one request; PRESS_ENTER is a separate key control. Separate yes/no checks cover goal completion and a stuck loop.
JevBrowserExt
JEV ROLEChooses CLICK, TYPE_TEXT, SELECT, SCROLL_DOWN, SCROLL_UP, PRESS_ENTER, WAIT, DONE, or BLOCKED and the matching DOM target in one request; PRESS_ENTER is a separate key control. Separate yes/no checks cover goal completion and a stuck loop.
Assesses tasks, filters research evidence, rates capability needs and selects UI components; local rules select models and execute work.
The script compares separate and batched answers, usage and latency; the full language engine remains a design goal.
Jev reads game state and chooses among the same seven moves offered to other controllers.
mk-jev-fly-brain
JEV ROLEJev reads game state and chooses among the same seven moves offered to other controllers.
Builds state and questions from input items, calls Jev and appends or separately outputs answers.
n8n-nodes-typesafe-jev
JEV ROLEBuilds state and questions from input items, calls Jev and appends or separately outputs answers.
Sends caller-provided state and questions to TypeSafe and returns typed Jev decisions.
openclaw-typesafe-ai
JEV ROLESends caller-provided state and questions to TypeSafe and returns typed Jev decisions.
Scores task, reasoning, and tool complexity, then chooses fast / balanced / strong / long; local policy stays, escalates, or fans out.
opencode-jev-orchestrator
JEV ROLEScores task, reasoning, and tool complexity, then chooses fast / balanced / strong / long; local policy stays, escalates, or fans out.
Jev judges which candidates remain useful; surviving text is returned verbatim.
pi-jev-compact
JEV ROLEJev judges which candidates remain useful; surviving text is returned verbatim.
**Deleting a feature made it more accurate.** The first build made a second Jev call to pin each tell to a sentence. Removing it took precision from 0.80 to 0.95 and latency from 1,365 ms to 575 ms. Jev returns typed answers rather than text, so a quote has to come from a second round of typed questions about lines, and that round was where every defect lived. The README documents the two intermediate designs that were worse.
slopcheck-jev
JEV ROLE**Deleting a feature made it more accurate.** The first build made a second Jev call to pin each tell to a sentence. Removing it took precision from 0.80 to 0.95 and latency from 1,365 ms to 575 ms. Jev returns typed answers rather than text, so a quote has to come from a second round of typed questions about lines, and that round was where every defect lived. The README documents the two intermediate designs that were worse.
For each level, Jev evaluates release probability plus plea, logic, and paradox signals, then selects a persuasion tactic and scores persuasiveness.
turing-jail
JEV ROLEFor each level, Jev evaluates release probability plus plea, logic, and paradox signals, then selects a persuasion tactic and scores persuasiveness.
system_one POSTs to /v1/systemone, defaulting to https://api.typesafe.ai and jev-latest.
typesafe-ai-ruby
JEV ROLEsystem_one POSTs to /v1/systemone, defaulting to https://api.typesafe.ai and jev-latest.
Sends state and named questions to TypeSafe and parses choice, score and noul answers.
typesafe-rs
JEV ROLESends state and named questions to TypeSafe and parses choice, score and noul answers.
Sends state and typed questions to System One and receives structured answers through async/await.
typesafe-sdk-swift
JEV ROLESends state and typed questions to System One and receives structured answers through async/await.
TypeSafeClient.SystemOneAsync POSTs {Endpoint}/v1/systemone, defaulting to https://api.typesafe.ai and jev-latest.
TypeSafeSDK
JEV ROLETypeSafeClient.SystemOneAsync POSTs {Endpoint}/v1/systemone, defaulting to https://api.typesafe.ai and jev-latest.
Adds bounded decisions and uncertainty escalation to Hermes.
Judges tool risk, edit quality and loops while deterministic code applies policy.
Scores whether cited paper passages support a sentence, with final human review.
citation-verifier
JEV ROLEScores whether cited paper passages support a sentence, with final human review.
Checks files for code smells and complexity before another model writes the review.
Clean Code Judge
JEV ROLEChecks files for code smells and complexity before another model writes the review.
Answers a semantic question about each function and ranks results by Noul probability.
Reads generated scenes and judges location, danger and soundtrack state.
Judges intent, clarity, hostility and information density for code-generated diagnostics.
Human Compiler
JEV ROLEJudges intent, clarity, hostility and information density for code-generated diagnostics.
Selects a Claude model and reasoning effort for each message.
Selects city, research, diplomacy and unit actions in Civilization II.
Jev × Civilization II
JEV ROLESelects city, research, diplomacy and unit actions in Civilization II.
Selects tools and arguments; code executes them and renders result cards.
Selects suitable agent skills and declines uncertain matches or requests review.
jev-agent-skill-router
JEV ROLESelects suitable agent skills and declines uncertain matches or requests review.
Identifies Spanish profanity in timestamped text for audio redaction.
Scores outgoing Wikipedia links to find a path between two articles.
Selects Gomoku moves through a MoonBit client.
Combines Jev judgments with confidence gates, policies and shadow mode.
Optionally selects which language model should serve an incoming request.
jev-router (prismhq)
JEV ROLEOptionally selects which language model should serve an incoming request.
Judges whether file snippets contain real secrets in repeatable evaluations.
jev-secret-detection
JEV ROLEJudges whether file snippets contain real secrets in repeatable evaluations.
Chooses the next action in the Chrome dinosaur game.
Checks plain-language completion rules when an agent is about to stop.
Selects useful context to retain and routes each turn to a suitable model.
Routes Pi requests to models through typed evaluations on Vercel AI Gateway.
pi-jev-router
JEV ROLERoutes Pi requests to models through typed evaluations on Vercel AI Gateway.
Exposes Jev judgments through five Pi tools; the application owns thresholds and actions.
pi-typesafe-jev
JEV ROLEExposes Jev judgments through five Pi tools; the application owns thresholds and actions.
Judges whether an agent trajectory is stalled so code can continue or replan.
ProgressGate
JEV ROLEJudges whether an agent trajectory is stalled so code can continue or replan.
Selects bounded robot-arm joint steps within a decision-run budget.
Ranks agent skills against the current execution context.
Evaluates the Snake board and directions; Python combines the answers into a move.
Exposes typed judgments, choices and scores as SQLite functions.
Runs multiple checks on model responses; configured policy controls the outcome.
Judges issue type, severity, urgency, duplication and next steps.
Evaluates CVs against editable criteria and returns structured assessments.
typesafe-jev CV screener
JEV ROLEEvaluates CVs against editable criteria and returns structured assessments.
EDITORIAL NOTES & SOURCES
Project leads and selected descriptions from logicrw/awesome-jev-projects (MIT) and Awesome Jev, cross-checked against project documentation and implementations. Stars and forks are dated snapshots. Inclusion is not a runtime test or a performance endorsement.