New: Decision Runtime — closed-set judgments for Agent OS. Read the guide →
HazelJSStarter
QA Agent
Generate tests, execute evaluation scenarios, validate behavior, and report regressions.
What to build
A QA agent that generates and runs eval scenarios against agents or APIs.
Suggested architecture
- 1Target system↓
- 2QA Agent↓
- 3Generate scenarios↓
- 4Run eval / tests↓
- 5Detect regressions↓
- 6Report
Minimum acceptance criteria
- Public GitHub repository
- Uses @hazeljs/agent
- At least one eval or test generation path
- README with setup instructions
- Uses @hazeljs/testing or custom eval harness
Optional advanced goals
TestingEvalObservabilityCI integration
Submit QA Agent
This form saves your project to our database. We review it for the community showcase — you do not open a pull request.
Submit your repository
Public GitHub repos only. This form saves your project to our database. We review it before it can appear on the community showcase — you do not open a pull request.