One AI skill + MCP server for Claude Code, Codex and Hermes: Jev (TypeSafe) gates large-model turns (event triage, review verdicts, owner questions, tool choice), runs build-and-repair loops with inde
View on GitHub ↗One workflow for your AI coding tools: build, skip the large-model turns you don't need, choose the right help, review and publish.
Jev-AI-Skill is one skill and one MCP server for Claude Code, Codex and Hermes. Jev, the fast decision model from TypeSafe System One, makes the small judgments; your AI coding tool does the work. Describe what you want to achieve, and your assistant gets a structured way to work toward it, check the result and prepare the next Git step, while Jev keeps large models out of turns a single cheap decision can settle.
- Loop builds, finds and fixes code with configured checks and an independent review. - Gate asks Jev before a large model runs, with four MCP tools: - gatetriage decides whether a cron run, poll, notification or chat event needs an agent at all. - gateverdict checks which review findings the evidence already closes before another review round. - gatedecide answers a question you would otherwise be asked, and says when you really must decide. - gatepick chooses a tool, skill or session from one-line descriptions. - --watch runs a recurring check with no agent: silent on success, and a new failure is reported once. - --http and JEVMCPURL keep the TypeSafe key on one always-on server for your private network, while Loop still runs on each machine next to its own projects. - Route picks a model tier, a skill and whether to delegate. - Git recommends commit, push or pull-request steps within repository rules. - findcode.py ranks code-search results so a coding model reads less.
From the project README.
Add the radar badge to your README — it shows your project was picked up by MCP Radar and links to this page:
[](https://mcp.liqiwa.com/s/King4s--Jev-AI-Skill.html)
Browser use for coding agents, 3-5x faster than browser-use. MCP server + CLI; TypeSafe Jev decides every step in ~300 ms.
NiazMorshed2007/jev-reviewLocal-first MCP plugin for continuous software-quality review by AI coding agents, powered by Jev.
mcpdelta/mcpdeltaDelta MCP is a free app that sits between your AI apps and their MCP servers: one program per task instead of one tool call per step, up to 24.1× fewer tokens in our benchmarks. Every change is record
agent-chaperone/agent-chaperoneScreens an AI agent's tool calls before they run and tool results before the agent reads them. An MCP proxy plus a hooks adapter for a client's built-in tools.
rashedInt32/jev-mcpMCP server exposing TypeSafe Jev as typed, calibrated judgment tools: classify, score, check, batched ask. Ships as a Claude Code plugin.
abhishekswe/agent-fastpathJev MCP server: a decision layer for coding agents, built on TypeSafe Jev (System One model). Ship gates, risk checks, file triage that keeps files out of context, and a safe headless browser, with ca
The top new MCP servers of the week, every Monday. No spam, unsubscribe anytime.