useqai/QAI Agent
Stop guessing why CI failed. QAI posts root cause, fix suggestions, and a merge verdict directly on your PR — automatically.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| junit-path | Path or glob to JUnit XML file(s), e.g. "test-results/results.xml" or "**/junit-*.xml" | yes | — |
| github-token | GitHub token for posting PR comments. Defaults to the built-in GITHUB_TOKEN. | no | ${{ github.token }} |
| post-comment | Whether to post a PR comment with the analysis results (true/false) | no | true |
| qai-url | QAI cloud platform ingest URL for historical intelligence (optional). Example: https://ingest.your-domain.com | no | — |
| qai-api-key | QAI API key for cloud platform (required when qai-url is set) | no | — |
| trace-path | Glob to Playwright trace zip files for RCA analysis, e.g. "test-results/**/*.zip" | no | — |
| playwright-report | Path to Playwright JSON report file for richer AI fix suggestions, e.g. "test-results/report.json". Requires qai-url and qai-api-key. | no | — |
| fail-on-high-risk | Fail the action step when risk level is high (true/false). Default false — the action reports but never blocks. | no | false |
| suite-name | Display name for this test suite (e.g. "Selenium Python", "WebdriverIO"). When set, each suite gets its own PR comment instead of overwriting others. | no | "" |
| slack-webhook-url | Slack incoming webhook URL. When provided, posts a plain-text High Risk alert when risk level is high. For richer alerts with cluster trends and AI analysis, use the QAI platform. | no | — |
Outputs
| name | description |
|---|---|
| risk-level | Computed risk level: low | medium | high |
| risk-score | Computed risk score: 0.0–1.0 |
| failed-tests | Number of failed tests |
| total-tests | Total number of tests |
| cluster-count | Number of unique failure clusters detected |