User repository on GitHub
MCP server that puts TypeSafe Jev on the coding loop in Cursor, Codex, and any MCP client
- Forks
- 10
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Jev MCP — burnigtm
An MCP server that puts Jev in the coding loop for Cursor, Codex, and other MCP clients: coding-loop checks, ranking, verification, and review tools.
- Stars
- 64
- Language
- TypeScript
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Last verified
- Sep 2026
This is one of several Jev listings for the same project. The canonical entry is Jev MCP — jkudish.
About this project
A local stdio MCP server that gives Cursor, Codex, and other MCP clients typed TypeSafe Jev judgments. Jev returns Choice, Score, and Noul answers; the host agent still edits files and runs commands. Results include typed answers, token usage, and an action: auto, review, or escalate. Confidence measures model certainty, not factual truth. Incomplete context never permits auto; reduce the input and submit it again for a complete judgment.
Quoted from the project's README. Read the full README.
Highlights
- Ships nine tools including jev_step, jev_review, jev_verify, jev_gate, jev_screen, and jev_rank.
- Every result carries an action of auto, review, or escalate alongside token usage.
- The prepared-call router accepts at most 32 candidates; empty or ineligible lists return locally with zero Jev usage.
- Rank handles up to 5,000 candidates with at most 250 options per upstream call and 2,000 characters per candidate.
- State plus questions must fit an estimated 64,000-token budget; calls default to a 30-second deadline set by JEV_MCP_TIMEOUT_MS.
Quickstart
bash
npm ci && npm run build
export TYPESAFE_API_KEY=ts_...
node dist/index.js doctor --json
codex mcp add jev --env TYPESAFE_API_KEY=ts_... -- node /abs/path/jev-mcp/dist/index.jsWatch out
Live quality and cost savings have not been measured, and JEV_MCP_MOCK=1 is for tests and demos rather than production decisions.
Questions about this entry
Short answers drawn from the linked project or page.
No, and the README is explicit: the host agent still edits files and runs commands, and Jev never invents arguments or executes tools. It returns an action of auto, review or escalate, and incomplete context never permits auto, so reduce the input and submit it again rather than trusting a judgment made on omitted context.
No. The README says live quality and cost savings have not been measured. What is concrete is structural: jev_step fuses routing and call selection, so a host that would call jev_coding_loop and then jev_tool_route spends one MCP round-trip instead of two, and a partner model is invoked only when partner_model.required is true.
It refuses rather than guessing. State plus questions must fit an estimated 64,000-token budget with a 32,000-token state-plus-longest-question budget, results expose incomplete coverage, and questions alone over the budget are rejected. A 30-second call deadline is tunable with JEV_MCP_TIMEOUT_MS.
More than that. It ships nine tools with a specific division of labour: jev_review assesses a proposed patch, jev_verify checks claims against supplied evidence, jev_gate combines both in one upstream call, and jev_rank takes up to 5,000 candidates with at most 250 options per call. Arithmetic and date calculation are deliberately left to your code.
Install Node 20+, checkout the repo, run `npm ci` and `npm run build`, set TYPESAFE_API_KEY and check with `node dist/index.js doctor`. Register it in Codex with `codex mcp add jev --env TYPESAFE_API_KEY=ts_... -- node /absolute/path/to/dist/index.js`, or copy the Cursor MCP example.
Nine: jev_step, jev_coding_loop, jev_tool_route, jev_review, jev_verify, jev_gate, jev_screen, jev_rank and jev_evaluate. Question packs ship as MCP resources under jev://packs/, and results include typed answers, token usage and an action of auto, review or escalate.
Yes, for a deterministic local demo set JEV_MCP_MOCK=1; the README says mock mode is for tests and demos, not production decisions. API failures and invalid responses return typed errors rather than fabricated judgments, and diagnostics do not log request content or API keys.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: jev-review: a local-first score loop for coding agents
Reddit: One command to wire Jev into Claude Code, Claude Desktop, and Codex.
X: Decision outputs in Grep
X: Jev in front of the HeyGen MCP
More Jev resources like this
An MCP server exposing ten Jev judgment tools, including verify, screen, find, rerank, classify, review, and gate, each returning typed probabilities in roughly 150 to 500 ms.
Jev Tools & CLIs#community#typescript#mcp
Community
jkudish/jev-mcpJavaScript
jkudish/jev-mcpJavaScriptAn MCP server and agent skill that verifies coding-agent work by driving the real running app and returning pass, fail, or couldn't-tell verdicts with the file:line to fix; its listing frames the approach as Jev-style machine-native runtime perception.
Jev Tools & CLIs#community#typescript#mcp
Community
reticlehq/reticleTypeScript
reticlehq/reticleTypeScriptAn MCP server for Antigravity IDE, Cursor, and Claude Code that uses Jev (jev-latest) to locate code semantically and audit git diffs, replacing broad grep sweeps of entire files.
Jev Tools & CLIs#community#python#mcp
A local SQLite memory layer for CLI agents with an opt-in Jev reranker that batches Noul judgments over the top 40 recalled memories and falls back to a local cross-encoder on errors.
Jev Tools & CLIs#community#typescript#agents
Community
kitfunso/hippo-memoryTypeScript
kitfunso/hippo-memoryTypeScriptA self-hosted, versioned skills library for AI agents with HTTP MCP, scoped client keys, and opt-in Jev recommendations that rank skills for a task using your own OpenRouter, TypeSafe, or Vercel key.
Jev Tools & CLIs#community#typescript#mcp
Community
kitze/skillboxTypeScript
kitze/skillboxTypeScriptA router that turns models, subagents, skills, MCP tools, CLIs, and plugins into one candidate set Jev answers with a single typed Choice, while the router code owns availability, permissions, and confirmation.
Jev Tools & CLIs#community#typescript#routing
Community
BillionsBobby/JevRouterTypeScript
BillionsBobby/JevRouterTypeScriptFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.