Jev in production › Evaluation and testing
jev-browse (cooper667)
Claude Code plugin uses Cloudflare-hosted Jev to judge browser QA steps. A run costs about a cent (author).
a centmeasured, as published by the source
- Use
- Evaluation and testing
- Industry
- Developer tools
- Form
- Open-source tool
- Stage
- Announced
- Plugs into
- Claude Code
- Listed
- 2026-09-25
- Found via
- github
- Repository
- cooper667/jev-browse
- Stars
- 7
- Forks
- 1
- Last push
- 2026-10-09
- Language
- TypeScript
- License
- MIT
The README opens with
Runs a checklist written in plain English, one action or check per line, in a real browser. Playwright drives the page and a classifier decides each step: which element an action means, and whether a check holds. An agent uses it to check its own work after a user-visible change, to run a pull request's QA steps as written, or to compare two builds. It starts from any URL, or logged in through an app's own e2e fixture. A 24-step flow takes about 35 seconds.
Badge
For the project's own README, linking back here:
Also used for evaluation and testing
- jevals (Openlayer) p50 244ms
- grounded-ai (CascadeEvaluator) 51% faster
- Pinecone Cultivar ~30x cheaper
- jev-test-filter $0.00098
- Jev Arena 7,671
- touchpress 29 seconds
- SmartMoney-Cub 78.43%
- MetaCog 86.7 %