User repository on GitHub
Test your AI agents the way your users talk to them (alpha)
- Forks
- 0
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Hafen
A testing app and CLI for chat endpoints: send messages to any chat API, watch replies stream in as a conversation, and check every reply with rule checks (JSONPath, latency, tokens, cost, tool calls) and judged checks including TypeSafe's System One. Export the suites as files and run them in CI.
- Stars
- 0
- Language
- JavaScript
- License
- MIT
- Reading time
- 3 min
- Published by
- Community
- Author
- abhimanyupandian
- Last verified
- Oct 2026
- Languages
- JavaScript
- Platforms
- CLI
- Capabilities
- Open sourceRequires API key
Highlights
- Plain HTTP endpoints (one JSON reply, SSE or NDJSON streams) plus presets for OpenAI Chat Completions and Responses, Anthropic Messages, Gemini and Ollama.
- Rule checks cover JSONPath, status, latency, tokens, cost, tool calls and event order; judged checks cover System One and LLM judges with rubrics.
- Every check result is pass, fail, unsure, skipped or error with a reason, and judged results carry a confidence.
- The app stores suites, providers and plugins as YAML and JavaScript; Export writes them to a repository folder, where the CLI runs them unchanged.
- Secrets are referenced as {{secret.NAME}} and redacted from every run file and report, including where they appear escaped, URL-encoded or in base64.
Watch out
Grounded in the truncated README and GitHub metadata reviewed on 2026-10-15; the README is cut off in the screenshots section, install instructions were not in the captured excerpt, and the project labels itself a proof of concept with feature testing still in progress.
Questions about this entry
Short answers drawn from the linked project or page.
A proof-of-concept app and CLI for testing chat endpoints the way users talk to them: send messages to any chat API, watch the replies stream in as a conversation, check every reply, and then run the same test files in CI.
Plain HTTP endpoints that return one JSON reply or stream SSE/NDJSON, plus presets for OpenAI Chat Completions and Responses, Anthropic Messages, Gemini and Ollama. Anything else takes a small plugin, and an endpoint behind a login takes a fixture that logs in first.
Rule checks evaluate JSONPath, status, latency, length, tokens, cost, tool calls, thinking and the order of events; judged checks include TypeSafe's System One, LLM judges with rubrics and reference answers. Each result is pass, fail, unsure, skipped or error, always with a reason.
The app keeps suites, providers and plugins as plain YAML and JavaScript; Export writes them with everything they use to a folder in your repository, where the command-line runner executes them unchanged, exits with a clear code and writes JUnit, Markdown, JSON and run files.
Endpoints that require authentication take secrets, and judged checks through TypeSafe's System One need a TypeSafe key. Secrets are named as {{secret.NAME}} in files rather than stored as values there, and the desktop app keeps them encrypted by the system per scope.
Jev is the judge for System One checks: the README's screenshot shows a check's value of 0.91 with confidence 0.82 and the judge's full answer shown in the panel, alongside rule checks that carry their own badges. It runs as one of the judged check types.
Not according to its own README: Hafen is an early-stage proof of concept with testing of its features in progress, and it asks you to expect rough edges and report what you find. Install instructions were not in the captured excerpt.
More Jev resources like this
Local-first test selector that asks a judge — classifier.dev by default, or Jev with a TypeSafe key — whether each test could be affected by the current code change, then hands only the selected files to Playwright or Vitest. Uncertainty always runs the full suite.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
baronunread/leanestTypeScript
baronunread/leanestTypeScriptA linter for Jev queries that takes the request you would send to System One — a state and the questions asked about it — and reports where the query is written in a way the model is documented to handle badly. Static checks run free without a key; the rest ask Jev.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
fox-islam/jevlintPython
fox-islam/jevlintPythonLatitude is an open-source observability platform for AI agents: one line of telemetry captures traces, failing traces group into tracked signals, and fixes are dispatched to coding agents and replayed against real traces.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
latitude-dev/latitude-llmTypeScript
latitude-dev/latitude-llmTypeScriptAn enterprise admin-console test tool that generates replayable Excel test cases from live browser pages, or replays existing Excel cases, driving Playwright with local Laya or a compatible decision API to match fields, options, and targets.
Jev Tools & CLIs#community#catalog-import#madewithjev
Community
anthonyli/laya-pilotJavaScript
anthonyli/laya-pilotJavaScriptWrite end-to-end tests for web apps in plain English; jev-e2e turns each case into a plan, has Jev pick the page controls, and runs it with Playwright, returning PASS, FAIL or BLOCKED with HTML, JSON and masked screenshot evidence.
Jev Tools & CLIs#community#catalog-import#madewithjev
Community
perixtar/jev-e2eTypeScript
perixtar/jev-e2eTypeScriptBrowser QA where Playwright drives and films, and Jev judges: natural-language assertions, self-healing locators and goal-driven click loops, with a JSON-flow CLI for agents, a local dashboard and an agent skill.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
jonymusky/jev-browser-qaTypeScript
jonymusky/jev-browser-qaTypeScriptFrom 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.