demoauthor demo
Pull-request review gate
Jev evaluates a pull request against bounded review questions before results reach the user.
PR GATES AND ROUTERS
Coding workflows
Jev inside coding agents: reviewing diffs against fixed checks, choosing models per turn, deciding what stays in context.
From this task to a first run
Choose a project, read its requirements, then follow its setup reference or a related guide.
- Oko: local code search and prompt-time retrieval: How to run →
Supported macOS or Linux binary; a local source checkout
Try a walkthrough
- Connect Jev to your app →
Set up the Python SDK, make a typed routing decision and keep the result under your application's control.
- Connect Jev with JavaScript and TypeScript →
Make a server-side structured decision with the official SDK, then check the response before your app uses it.
PR gates, routers and review tools that turn a code question into a typed decision.
88 entries
reposource-backed
Jev Codex Router
Per-turn routing chooses a model, reasoning depth and speed mode for Codex.
demoauthor demo
Instant context compaction
Jev scores which tool calls and results remain relevant before context is compacted.
demoauthor demo
Agent routing in FirstMate
A model or agent route is chosen from a fixed catalog based on the current task.
reposource-backed
Building with Jev skill
An agent skill for finding and designing bounded Jev decisions in software.
postcommentary
Find every decision in a workflow
A prompt asks coding agents to inventory every bounded semantic decision where Jev might replace a heavier model call.
postcommentary
Early Jev GitHub projects
An early community roundup of repositories and Jev experiments.
demoauthor demo
Vercel's fx safety classifier on Jev
Vercel benchmarked the auto-mode safety reviewer of its fx coding agent with Jev in place of GPT Luna and says it is likely to become the default.
reposource-backed
toolgate: a tool-call firewall for agents
Screens agent tool calls for destructive, exfiltration, privilege-escalation and off-task actions using Jev probabilities, ships as an npm package and a Claude Code PreToolUse hook, and logs an audit trail.
reposource-backed
jev-rules for Claude Code
A Claude Code plugin that asks Jev which standing rules and codebase maps are relevant to the current prompt or file edit, injects only those, and falls back to all rules if the API is down.
reposource-backed
blink: codebase search with Jev-scored paths
Jev scores file and folder names so an ensemble of walkers is allocated across the most probable paths; a Bun CLI with recursive and multi-walker modes and JSON traces.
demoauthor demo
Claude Code hook routing tasks with Jev
A Claude Code hook where Jev decides whether a task goes to Haiku, an Opus sub-agent, or an external harness such as Codex or OpenCode.
articleauthor demo
Empryo: Jev inside a coding-agent harness
Wires Jev into five decision points of a coding-agent harness; the main test is failure triage on 102 real provider-API failures from session logs.
demoauthor demo
Request router that picks the serving LLM
A router where Jev decides which LLM should serve each incoming request; the same author hosts a Pac-Man page where Jev chooses a direction at every junction.
reposource-backed
jev-review: staged code review with a local panel
A step-by-step code-review workflow built on Jev with a local dashboard, listed in the Studio Yebisu roundup.
reposource-backed
jev-router for Claude Code
Routes Claude Code tasks to a cheaper or more suitable model based on a Jev decision; listed in the Studio Yebisu roundup.
postsource-backed
Foreman: a supervisor for coding agents
Foreman watches coding agents and uses Jev to continuously assess progress, completeness, tests, drift and verification, intervening when needed.
demoauthor demo
OpenCode browser-use CLI with Jev
A preview of browser automation for OpenCode powered by Jev plus its browser-use CLI, framed around speed; the demo video sits in Neriousy's post and dax's root post.
integrationauthor demo
skillbox: skill routing with Jev
skillbox reads the query and returns the relevant agent skills immediately with Jev, instead of agents spending many turns searching for them.
reposource-backed
jev-pruner: trim tool output before the LLM sees it
A Claude Code and Codex CLI plugin that intercepts Bash output above roughly 10K estimated tokens, chunks it and asks Jev which lines are still needed, dropping progress logs and boilerplate.
integrationauthor demo
mcp_jev: local MCP server with Jev packs
A local MCP server that runs Jev "packs" (code audit, PR review, GUI automation) for Cursor, Claude Code and Codex.
integrationauthor demo
Agent Handoff Gate
An experimental protocol where a worker agent sends requirements, code and test results to Jev, which returns an evidence verdict for the team-lead agent before results are accepted.
demoauthor demo
Secret detection test with Jev
A quick credential and secret-detection experiment with Jev; the author reports consistent results across runs.
reposource-backed
Stanley: a Jev-first, self-evolving coding agent
Bounded TypeSafe Jev workflows for coding agents: small-model judgments combined with deterministic workflows so coding work stays bounded, from the author of jev-review.
integrationsource-backed
jev-preflight: a bounded risk check for Claude Code
Born from Claude Code changing an auth check to return true with tests still passing: eight risk axes scored by Jev in one request, with one optional re-inspection before the agent declares done.
integrationsource-backed
pi-jev-guard: check Pi code edits against repo rules
For the Pi coding agent: every code edit is checked by Jev against the repository's Markdown rules (agents.md and friends) so the agent stops ignoring them.
integrationsource-backed
abide: make your coding agent follow project rules
Uses Jev to enforce the project rules linters cannot catch — the judgment calls in a style guide — on a coding agent's output.
integrationsource-backed
pi-jev: Jev-assisted file retrieval and request caching for Pi
Speeds up Pi coding-agent workflows by letting a decision model pick which files matter and cache requests; the author also tried DiffusionGemma in place of Jev and documents the results in VALIDATION.md.
integrationsource-backed
A Jev-based model and effort router for Codex
An autorouter that lets Jev pick both the model and the reasoning effort for each Codex turn — 'I cannot use codex the same anymore'.
integrationsource-backed
hunk-triage: read the important changes first
An extension for bentlegen's hunk that classifies the files in a diff (core, tests, docs…) with Jev and reorders the review so the important changes come first.
demosource-backed
jev-lexer: syntax highlighting with no language knowledge
Split like gpu-lexer, classify every token with Jev, render Shiki-compatible tokens to HTML and ANSI — a language-agnostic highlighter, experimental only.
demoauthor demo
Greptile realtime: a Jev bug checker as you type
A simple real-time bug checker built with Jev for people who still code by hand; it also underlines inaccurate English sentences.
integrationsource-backed
winnow: a calibrated context sieve for Claude Code
Every tool result is judged by a System One model before it enters Claude Code's context, so noise never gets a chance to bloat the window.
integrationsource-backed
Canny: stop coding agents from claiming work is done without evidence
Deterministic hooks decide, Jev advises: an append-only ledger with zero runtime dependencies that blocks an agent's 'done' until there is evidence.
integrationsource-backed
pi-jev (y0usaf): a measured tool-call gate for Pi
Jev as a decision layer for the Pi coding agent: a measured tool-call gate plus a jev_ask tool for typed, calibrated answers; distinct from madeye/pi-jev.
articlecommentary
The Jev + Claude Code stack
An X Article on wiring Jev into a Claude Code workflow for cheaper, faster agents.
articleauthor demo
Jev in front of Grok Build: 22–40% cheaper on the same tasks
An X Article on putting Jev as a gate in front of Grok Build and measuring the cost difference on identical tasks.
integrationsource-backed
pi-typesafe-jev: Jev as five native Pi tools
A Pi extension exposing TypeSafe judgments as five tools (choice / bool / score and friends) so the model makes narrow semantic calls while your code keeps thresholds and weights.
integrationsource-backed
opencode-jev-orchestrator: cheap sticky model, Jev escalates hard turns
Keeps OpenCode on one cheap model for a warm cache and lets Jev decide which turns are hard enough to spawn stronger subagents.
integrationsource-backed
switchboard: a cache-aware decision router for Claude Code and Codex
Model-agnostic router that takes prompt cache into account when Jev picks the model per turn; other System One models planned.
integrationsource-backed
pi-jev-router: automatic model routing for Pi via Vercel AI Gateway
Open-source Pi extension that lets Jev route each turn to a model through Vercel AI Gateway; the author pairs it with Astra adaptive thinking.
integrationsource-backed
jev-guard: auto mode for every coding agent
Risk-scores every tool call (rm, curl, bash…) with session context into deny / ask / allow, flags prompt injection in tool results and checks skills and plugins; works with Claude Code, Codex and others.
integrationsource-backed
is-malicious: scan a repo before you run it
A codebase scanner that checks third-party repositories and CI scripts for malicious behaviour before execution, with Jev as the judge.
integrationsource-backed
qc-use: Jev-assisted UI tests from plain-language steps
Turns plain-language critical-path tests into bounded UI actions chosen by Jev. The runner checks a fresh screenshot after each action and records a step-by-step result.
integrationsource-backed
cmd-mod-jev-nudge: decide when an agent should continue
A Command Code mod asks Jev whether a gentle nudge would help an agent continue useful work within the user's existing request.
integrationsource-backed
CodeNib: Jev reranking for code search
CodeNib uses Jev to rerank code-search results and publishes integration docs plus an evaluation on 100 CodeNib Base issues.
reposource-backed
mizchi/jev-playground: MoonBit client and experiments
A MoonBit Jev client and CLI with game, shell-risk and browser experiments and the author's measurement reports.
integrationsource-backed
Caliber ai-setup: optional Jev transcript compaction
Caliber's agent-configuration CLI includes an optional Jev plugin that scores tool-call history for retention during transcript compaction.
integrationsource-backed
system-one: Promise and Effect clients for typed decisions
A TypeScript client with shared typed question definitions and Promise and Effect APIs for Jev and other System One-compatible models.
integrationsource-backed
Spring AI TypeSafe: Java SDK and Jev integrations
A Java System One client plus Spring AI modules for Jev judging, guardrails, retrieval filtering and reranking, and tool selection.
integrationsource-backed
kataras/jev: a Go client for System One
A community Go client for Jev's Noul, Choice and Score questions with typed response helpers, configurable retries and rate limiting.
productsource-backed
Jev Workbench: versioned judgment functions
A local browser workbench for defining, previewing and publishing versioned Jev functions that callers can invoke over HTTP or MCP.
integrationsource-backed
Jevify: an agent skill for typed-decision workflows
An independent agent skill for spotting Jev opportunities, designing typed questions, preserving source evidence and testing a small decision workflow.
reposource-backed
Community jev-harness: proposal review with four questions
A research-stage coding harness where an LLM proposes an action, Jev answers four bounded questions and code records evidence for the host to review.
integrationsource-backed
jevgrep: semantic code search for coding agents
A CLI that uses Jev to judge relevance across repository folders, files and declarations, returning file leads and verbatim source excerpts to a coding agent.
articlecommentary
Jev and Codex: a Japanese implementation guide
A Japanese guide to the TypeSafe skill, Jev's three question types, separation of generation from judgment, and tests for a Codex-assisted workflow.
repocommentary
Beacon: cross-harness memory for coding agents
Beacon captures sessions across coding-agent harnesses and turns reviewed workflows, corrections and debugging patterns into reusable knowledge.
integrationsource-backed
System One for TypeScript: provider-neutral decisions
A TypeScript System One runtime provides shared Choice, Noul and Score builders for Jev, Reflex and compatible endpoints, with a companion Pi extension.
reposource-backed
JevGrep: semantic source retrieval through CLI and MCP
A TypeScript CLI and local MCP server ask Jev to rank source fragments and return exact excerpts with paths and line numbers to coding agents.
reposource-backed
JevFind: two-stage semantic code finder
A Rust code-search CLI first scores source paths with Jev, then ranks overlapping windows from the selected files and prints matching snippets.
integrationsource-backed
pi-jev-tools: ranking and retrieval tools for Pi
A Pi extension exposes Jev ranking, classification, local snippet retrieval and general typed evaluation while retaining the normal coding model.
reposource-backed
jevgrep jg: Rust natural-language code search
A standalone Rust binary searches a codebase using natural-language questions and returns relevant files and line ranges with scoring details.
integrationsource-backed
hyspacex Jev Router: session model admission
A self-hosted router asks Jev to assess fresh coding tasks and applies YAML policies to choose an upstream model and reasoning effort. Strict sessions retain their admitted model, provider and limits through later tool calls.
integrationsource-backed
pi-jev-auto-mode: typed tool permission gate
A Pi extension applies deterministic policies before asking Jev to judge uncertain bash, write and edit calls. It records semantic decisions and blocks calls when the default policy cannot reach a decision.
integrationsource-backed
genigrep: source evidence packs for coding agents
A CLI, MCP server and library combine ripgrep recall with Jev judgments over candidate files and passages. Results contain original source excerpts with paths, line numbers and explicit omitted ranges.
integrationsource-backed
Jev Navigator: code relationships and semantic search
A Python code-navigation library builds a structural index and composes small Jev judgments into best-first source searches. Its commands can trace known workflows or find matching functions while saving source provenance and partial results.
integrationsource-backed
Jev Playwright: change-aware browser test selection
A Playwright reporter asks Jev which discovered browser tests are relevant to a code change before browser execution. It combines changed paths and patches with test metadata and optionally test source.
reposource-backed
jevcal: fit confidence thresholds and detect decision drift
A CLI evaluates Jev against labelled application data, chooses per-question thresholds on one split and checks them on held-out examples. It writes a threshold lock and report, supports low-confidence fallback, and fails CI when accepted accuracy or coverage drifts.
reposource-backed
Oko: local code search and prompt-time retrieval
A local Rust search tool combines keyword discovery with Jev ranking and delivers source excerpts through a CLI or MCP. Its setup command connects Codex, Claude Code and OpenCode and can supply retrieval context when a code question is submitted.
reposource-backed
jevr: BM25 recall and Jev reranking for code and documents
A Rust CLI searches source trees and text documents with local BM25 recall, then uses Jev to filter and order candidate files. Results carry paths, line ranges and relevance scores in a grep-style interface.
integrationsource-backed
oxlint-plugin-jev: plain-English rules over source code
An experimental Oxlint plugin evaluates natural-language yes/no rules over functions, calls, JSX elements or files using Jev. Violations are reported when the configured probability cutoff is met.
reposource-backed
patdown: Markdown rule linting and agent write checks
A CLI checks files against fuzzy rules written in Markdown, using a swappable judge with TypeSafe as the default backend. Companion packages apply the same rules to Pi writes and Claude Code Write/Edit hooks.
integrationsource-backed
pi-heed: conversational constraints for Pi tool calls
An experimental Pi extension maintains quoted rule and task ledgers and checks side-effecting tool calls against them. The main agent records user receipts; Jev handles bounded semantic checks while code enforces resource, scope and precedence rules.
integrationsource-backed
xpressabhi Jev Browser: a harness-independent decision core
A TypeScript browser decision core chooses operations and observed targets from snapshots without importing a particular coding harness. CLI and harness adapters share the same Jev questions and step validation.
reposource-backed
every: ask a yes/no question of each source function
A Python CLI parses source files into functions and asks Jev the same yes/no question about each unit. It ranks the returned probabilities and uses chunks when a language cannot be parsed at function level.
integrationsource-backed
xafold Jev Router: cache-aware model routing for Claude requests
A Rust Anthropic-compatible proxy uses Jev judgments to select Claude model and effort settings. Deterministic routing rules also account for continuations, side requests and prompt-cache considerations.
integrationsource-backed
Jev Model Router: turn-level model and effort hooks
Claude Code function hooks ask Jev for a model tier and effort at turn start and apply the selection as the turn runs. A sticky cache guard and usage footer expose the selected route and actual model.
integrationsource-backed
Jev Agent Tools: bounded conversation context for Pi and Codex
A shared TypeScript decision tool integrates natively with Pi and through local MCP with Codex. It collects bounded conversation context from a known session and returns compact structured results through OpenRouter Decisions.
reposource-backed
Jev Agent Desk: local control of Claude and Codex sessions
A Linux and WSL2 tmux controller provides a local web interface for Claude and Codex sessions. Jev maps typed or spoken requests to bounded operations such as sending, switching, reading or interrupting.
reposource-backed
wellposed: lint the structure and meaning of Jev questions
An offline linter checks Jev request design before inference, including missing escape options, broken state references and malformed criteria. An optional semantic layer asks Jev whether a structurally valid question has missing or ambiguous answer boundaries.
integrationsource-backed
Oh My ClaudeCode: opt-in Jev judgments
Claude Code orchestration includes optional Jev judgment points alongside existing heuristics. The registry defines question schemas and distinguishes blocking decisions from background observations.
integrationsource-backed
OpenCodex: Jev model routing
OpenCodex adds Jev decision services to its provider and combo-routing system. Optional automatic routing chooses a configured model and reasoning effort before the main coding request.
integrationsource-backed
LLPhant: PHP Jev classifier
LLPhant exposes Jev through a PHP classifier and typed request objects. Applications can send a state with classification questions through the native TypeSafe endpoint.
integrationsource-backed
ReqLLM: Elixir TypeSafe evaluation
ReqLLM adds a TypeSafe provider for structured evaluation requests in Elixir. Its evaluate operation handles state and questions separately from the library’s text-generation operations.
integrationsource-backed
Pi extensions: TypeSafe decisions and recall
A Pi extension collection includes TypeSafe decision tools, Jev-assisted context compaction and searchable local conversation history. The compaction evaluator scores bounded batches of history units.
integrationsource-backed
Pro Workflow: advisory System One checks
Pro Workflow can add Jev judgments to coding-agent correction hints and risk suggestions. Its optional classifier also supports a compatible local Laya server.
integrationsource-backed
Pi MCP Adapter: Jev semantic tool search
Pi MCP Adapter can rank MCP tools by meaning using a System One evaluator. Project policy specifies which servers may share data for semantic search, alongside the existing word and regex search modes.
integrationsource-backed
GoModel: native Jev gateway provider
GoModel includes a Jev provider that forwards native System One decision requests. Its passthrough path preserves request and answer structures through the gateway.