natmart/AgentQA
Autonomous QA Engineer - AI-powered test generation, execution, and self-healing
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Maintainedlast commit Jan 28, 2026
- License
- MIT
Pinned Snippet
uses: natmart/agentqa@fecec0edd3db5babba0f6e4dcc895cfe74fd365e # no releases — HEAD as of 2026-07-11tags can be moved; commit SHAs can't. why a SHA?
Inputs
| name | description | required | default |
|---|---|---|---|
| path | Path to scan for untested code | no | ./src |
| framework | Test framework (jest, vitest, pytest) | no | vitest |
| openai-api-key | OpenAI API key for AI test generation | yes | — |
| github-token | GitHub token for posting PR comments | no | ${{ github.token }} |
| max-retries | Maximum healing retries per test (1-5) | no | 3 |
| dry-run | Run without writing files (preview mode) | no | false |
| auto-commit | Automatically commit generated tests | no | true |
| coverage-threshold | Fail if coverage drops below this percentage | no | "" |
| limit | Maximum number of files to process | no | "" |
| model | OpenAI model to use (gpt-4, gpt-4o, gpt-3.5-turbo) | no | gpt-4o |
| verbose | Enable verbose logging | no | false |
| post-comment | Post results as PR comment | no | true |
Outputs
| name | description |
|---|---|
| coverage-before | Coverage percentage before test generation |
| coverage-after | Coverage percentage after test generation |
| coverage-diff | Coverage difference (positive = improvement) |
| tests-generated | Number of tests generated |
| tests-healed | Number of tests that required healing |
| healing-success-rate | Percentage of failed tests successfully healed |
| summary | Summary of AgentQA run in markdown format |