User repository on GitHub
Semantic PR gate: .jev.yml rules as TypeSafe Jev questions. Not a review bot.
- Forks
- 0
- Last pushed
- Updated Sep 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Integrations & Adapters
Pr Sieve
A GitHub Action that treats .jev.yml as a semantic linter for pull requests: each rule is a typed Jev question (Noul, Choice or Score) and the check fails, comments or passes from those numbers. It writes no reviews, suggests no patches and approves nothing.
- Stars
- 1
- Language
- TypeScript
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- thestral12
- Last verified
- Oct 2026
- Languages
- TypeScript
- Capabilities
- Open sourceRequires API key
Highlights
- Policy is read from the base ref, so a PR cannot rewrite the gate that judges it.
- Secret-shaped strings such as AWS AKIA keys and private-key armor fail in code with no Jev call; everything else is one systemOne batch capped at 12 questions.
- Reasons are templates like missing_tests 0.87 meets 0.80 and fail, not generated prose; a clean run still posts a comment reading clean.
- Sits closest to check-risk (risk score) and is explicitly not jev-review (review, approve, dashboard).
- TypeSafe 429, 5xx or timeout yields neutral with sieve degraded; fork PRs without a key skip as neutral; a same-repo missing key is a setup failure.
Quickstart
bash
pnpm install
pnpm sieve eval --live
pnpm sieve doctorWatch out
Copy reviewed against the repo README on 2026-10-15; it documents example rules (looks_like_secret, missing_tests, touches_auth, change_kind) but no full rules reference or benchmarks, and the local eval is described as a mocked gold bar unless run with --live. Pin a commit SHA, not a tag: tags move.
Questions about this entry
Short answers drawn from the linked project or page.
A GitHub Action that treats .jev.yml as a semantic linter for pull requests. Each rule is a typed TypeSafe Jev question — Noul, Choice or Score — and the check fails, comments or passes from the returned numbers. It does not write reviews, suggest patches or approve anything.
Rules like looks_like_secret, missing_tests, touches_auth and change_kind each map to a gate threshold. Reasons are templates such as missing_tests 0.87 meets 0.80 and fail, not generated prose, and the sticky table shows answer, gate and action per rule.
Copy examples/jev.yml to .jev.yml and examples/pr-sieve.yml to .github/workflows/pr-sieve.yml, add TYPESAFE_API_KEY as a repository Actions secret, then open a PR to get a Check Run named pr-sieve plus a sticky table comment.
Pin a commit SHA rather than a tag because tags move. It also says pull_request only, never pull_request_target, and warns against uses: ./ from an untrusted PR with secrets in scope.
TypeSafe 429, 5xx or timeout turns the check neutral with sieve degraded. Fork PRs without a key skip as neutral; a same-repo pull request missing its key is a job setup failure rather than a silent pass or fail.
Yes. Secret-shaped strings such as AWS AKIA keys and private-key armor fail in code with no Jev call. Everything else is one systemOne batch capped at 12 questions, and Jev only asks whether a change looks like it handles secrets carelessly.
No generated reviews, no suggested patches, no auto-approve, no reviewer assignment, no dashboard and no SARIF output. It is a gate that returns a verdict, and the README frames it as a semantic linter rather than a review bot.
More Jev resources like this
A Playwright reporter that uses Jev to score which existing E2E tests a code change could affect and runs only those above a confidence threshold before browser execution, failing open to the full suite.
Jev Integrations & Adapters#community#catalog-import#logicrw
A Pi extension that checks proposed code edits against your repository's Markdown rules with TypeSafe Jev and returns the exact instruction text and source line range for each violation. Informative mode reports; enforcement mode blocks reviewed code.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
reindeer-ai/pi-jev-guardTypeScript
reindeer-ai/pi-jev-guardTypeScriptAn OpenCode 2 plugin that sends every shell command — local or over FarHand's remote shell — to Jev first: six typed questions about global installs, host litter, harmful actions and privacy, and a permission prompt whenever something looks risky.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
cogflux/opencode-jev-guardTypeScript
cogflux/opencode-jev-guardTypeScriptA GitHub Action that reviews repository files with your own yes/no questions: Jev returns probabilities and the check passes only when every expected answer meets its configured threshold.
Jev Integrations & Adapters#community#catalog-import#logicrw
A provider-agnostic decision gateway that puts Jev in front of expensive generative LLMs: small bounded questions are batched through Jev, a deterministic policy layer builds the execution plan, and generation runs only when the policy says so.
Jev Integrations & Adapters#community#catalog-import#logicrw
A GitHub Action that asks Jev one yes/no question per thing worth checking on a pull request and approves only when every answer clears your confidence threshold; otherwise it skips and comments with the numbers.
Jev Integrations & Adapters#community#catalog-import#logicrw
From 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.