User repository on GitHub
- Forks
- 8
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Agents & MCP
JevScout — hqman
A demo-MVP coding-agent skill and CLI that hunts AI and software jobs on real company sites: Chrome sees and acts, Jev scores every link and job, and the host LLM never picks what to click.
- Stars
- 38
- Language
- Python
- Reading time
- 2 min
- Published by
- Community
- Author
- hqman
- Last verified
- Oct 2026
Highlights
- jjh drives a visible Chrome over CDP; Jev (TypeSafe System One) scores every link and job, and the host LLM never picks what to click.
- One command: uv run jjh run --url https://openai.com --query 'Applied AI Engineer, Codex'.
- Runs with --mock and no API key, plus --max-pages N, --tab reuse, and a --companies shortcut.
- Packaged as a skill at skills/jev-job-hunter/SKILL.md for a coding-agent host.
- A fixtures-based slow-motion mode replays recorded pages with jjh step for narrated demos.
Quickstart
bash
cp .env.example .env # then set TYPESAFE_API_KEY (optional for --mock)
uv sync
uv run jjh run --url https://openai.com --query "Applied AI Engineer, Codex"Watch out
Copy reviewed against the repo README on 2026-10-15; the repo calls itself a demo MVP and the GitHub snapshot shows no license, pricing, or benchmark numbers. Real runs need a visible Chrome and a TYPESAFE_API_KEY; --mock and the fixture replay work without one.
Questions about this entry
Short answers drawn from the linked project or page.
A coding-agent skill plus a CLI named jjh that hunts AI and software jobs on real company sites. Chrome sees and acts, Jev (TypeSafe System One) scores every link and job, and the host LLM never picks what to click.
The host CLI runs one command; jjh drives a visible Chrome over CDP and calls Jev to score every link and job it finds, so the clicking policy comes from Jev rather than from the host model. The README's examples target sites like openai.com and typesafe.ai with a role query.
Copy .env.example to .env and set TYPESAFE_API_KEY, run uv sync, then uv run jjh run --url https://openai.com --query 'Applied AI Engineer, Codex'. Use uv run jjh log to review the run; this needs a visible Chrome window, not headless.
Yes, add --mock and run with no API key. The fixtures-based slow-motion mode replays recorded pages with jjh step for narrated demos, and the offline suite runs with uv run pytest -q.
Neither. jjh talks to Chrome over CDP directly and the README says Chrome DevTools MCP is optional, so the tool works with the visible Chrome you already have open.
A skill file at skills/jev-job-hunter/SKILL.md that documents the workflow for the host coding agent, so the agent runs the CLI while Jev makes the scoring calls rather than the language model.
The README calls it a demo MVP, and the GitHub snapshot shows no license. Real runs need a TypeSafe key, while --mock and the fixture replay work without one; no benchmark numbers are published.
Other listings of this project
Forks and copies of this project in the directory. This entry is the canonical one.
An open-source toolkit that reduces large MCP responses in Claude Code through a PostToolUse hook and maps Codex session requests to reviewed operations: Jev ranks what stays verbatim, and without an API key results pass through unchanged.
Jev Agents & MCP#community#catalog-import#logicrw
Community
yctimlin/jevscoutTypeScript
yctimlin/jevscoutTypeScriptMore Jev resources like this
An independent agent skill that walks a coding agent through designing a bounded Jev decision layer: contract, criteria, error behavior, fallback, and evaluation plan first, then scaffolding through TypeSafe or OpenRouter.
Jev Agents & MCP#community#catalog-import#logicrw
Community
grandamenium/jev-anythingPython
grandamenium/jev-anythingPythonA dependency-free MCP server that exposes TypeSafe's Jev to Claude Code, Codex, Cursor, Hermes, and other agents through the ask_jev and list_jev_models tools, so your code gets typed decisions with probabilities instead of prose to parse.
Jev Agents & MCP#community#catalog-import#logicrw
Community
codeia-academy/jev-mcpPython
codeia-academy/jev-mcpPythonA local Python service that keeps your skill catalog out of the main model's context: it reads your SKILL.md library, asks Jev which skills fit, and returns only the selected instructions through a single read-only MCP tool, with Claude Code, Codex, and Cursor adapters.
Jev Agents & MCP#community#catalog-import#logicrw
Community
himomohi/jev-skill-routerPython
himomohi/jev-skill-routerPythonpi extension that routes every prompt to a task-appropriate model tier using four TypeSafe Jev typed judgments, with budget-aware downgrades, candidate fallback chains, and the decision recorded in the transcript.
Jev Agents & MCP#community#catalog-import#logicrw
Lets AI agents use Windows applications through MCP or the command line: hand off a bounded task and Jev observes windows, chooses controls, types supplied text and navigates locally, returning a result or a request for help.
Jev Agents & MCP#community#catalog-import#logicrw
Community
jacks3tr/jev-desktopPython
jacks3tr/jev-desktopPythonOne skill and one MCP server for Claude Code, Codex and Hermes: Jev gates large-model turns with gate_triage, gate_verdict, gate_decide and gate_pick, runs build-and-repair loops with independent review, routes model tiers and skills, and guards Git steps.
Jev Agents & MCP#community#catalog-import#logicrw
Community
king4s/jev-ai-skillPython
king4s/jev-ai-skillPythonFrom 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.