User repository on GitHub
Jev-powered Playwright test selection
- Forks
- 0
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Integrations & Adapters
Jev Playwright
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.
- Stars
- 7
- Language
- TypeScript
- License
- MIT
- Reading time
- 2 min
- Published by
- Community
- Author
- arthurfiorette
- Last verified
- Oct 2026
- Languages
- TypeScript
- Platforms
- PlaywrightOpenRouter
- Capabilities
- Open sourceRequires API key
Highlights
- Requires Node.js 24.16+ and Playwright 1.62+ for reporter preprocess(); install with `pnpm add -D jev-playwright @playwright/test`.
- Reads staged, unstaged and untracked changes locally or a CI baseline; directly changed specs are always included.
- Tests at or above threshold (default 0.5) run; if selection cannot complete, the entire suite runs instead.
- A green run can execute zero tests, so the README says to check the 'Selected 0/N tests' message before trusting it.
- Changed paths, git patches, test titles and callback bodies may leave your machine; secrets in diffs are not redacted.
Quickstart
bash
pnpm add -D jev-playwright @playwright/test
export JEV_PLAYWRIGHT_PROVIDER_KEY="your-typesafe-key"Watch out
Copy reviewed against the repo README on 2026-10-15; the README was truncated mid-sentence in the 'How selection works' section, so calibration details are incomplete. The README itself warns it is a relevance filter, not a guarantee, and can still miss tests.
Questions about this entry
Short answers drawn from the linked project or page.
It adds a Playwright reporter that compares a code change with the tests Playwright discovered and uses Jev to score which ones are relevant, so a CI run executes a smaller set of existing E2E journeys before browser execution.
The reporter reads staged, unstaged and untracked changes locally or uses a configured CI baseline, always includes directly changed specs, and lets Jev score the remaining tests using their titles and file paths; tests at or above the threshold, 0.5 by default, run.
It fails open: if selection cannot be completed, the reporter runs the full suite. The README also warns that a green run can execute zero E2E tests, so check the 'Selected 0/N tests' message before relying on the run as a required CI check.
By default Jev receives changed paths, git patches, test titles and test callback bodies and comments; PR or commit descriptions may also be sent. Secrets in diffs or test source are not redacted, so filter sensitive paths with include/exclude or set includeTestSource to false.
Node.js 24.16+ as declared by the package, and Playwright 1.62+ because the reporter uses preprocess(). Install with `pnpm add -D jev-playwright @playwright/test` and set JEV_PLAYWRIGHT_PROVIDER_KEY, or configure the OpenRouter provider instead.
No. The README calls it a relevance filter, not a guarantee that every affected test will run: its goal is to skip clearly unrelated tests while keeping plausibly affected ones, and it can still miss tests, so keep full-suite coverage where completeness matters.
defineConfigWithJev(jevOptions, playwrightConfig) wraps your existing configuration and keeps your settings and other reporters intact. For CI-only selection, enable it with a process.env.CI check instead of enabled: true.
More Jev resources like this
A Claude Code mod that adds $.jev to the engine interface, letting plugin hooks ask Jev for Choice, Score and Noul answers over one state via TypeSafe, OpenRouter, Vercel, Cloudflare, LiteLLM or a custom endpoint.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
chrishan17/claude-jev-modTypeScript
chrishan17/claude-jev-modTypeScriptJev-scored context recovery for the Cursor CLI agent: hooks capture tool call and result I/O, two noul questions per call decide what to keep, and kept facts are re-injected verbatim after native compaction finishes.
Jev Integrations & Adapters#community#catalog-import#logicrw
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.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
thestral12/pr-sieveTypeScript
thestral12/pr-sieveTypeScriptA Cline plugin that gives the agent an isolated Playwright Chromium through a jev_run tool: Jev chooses a concrete operation and target per step via AI SDK 7's experimental_evaluate through the Vercel AI Gateway, and when it picks TYPE_TEXT, gemini-2.5-flash-lite writes the field value.
Jev Integrations & Adapters#community#catalog-import#logicrw
A host-agnostic routing layer that sits between a coding harness and its models, choosing capacity per round from trajectory evidence like tool calls, failures and context pressure, running as a user-scoped daemon and local OpenAI-compatible proxy.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
vizuh/sabiTypeScript
vizuh/sabiTypeScriptAn opencode plugin that turns an agent's small decisions — is this command safe, which tool fits, is this PR ready — into typed Jev questions with calibrated confidences, labelling each answer auto or escalate so the agent knows when to ask for help.
Jev Integrations & Adapters#community#catalog-import#logicrw
Community
darwintechlab/openjevJavaScript
darwintechlab/openjevJavaScriptFrom 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.