rbkunnela/FriendlyAI Review
Public GitHub Action wrapper that sends PR diffs to a private validation API and reports the returned verdict.
View on GitHubTrust Signals
- Scorecard Score
- not yet scored
- Maintenance Recency
- Stale
- License
- None
Inputs
| name | description | required | default |
|---|---|---|---|
| review-api-key | API key for the private validation service. Provide via secrets.FRIENDLYAI_REVIEW_API_KEY. | yes | — |
| review-api-url | Absolute URL of the private review endpoint. | yes | — |
| github-token | GitHub token for reading PR files and posting comments/check-runs. | no | ${{ github.token }} |
| max-diff-tokens | Approximate max token budget for inline PR patches sent to the API. | no | 15000 |
| request-timeout-seconds | Timeout for the review API request. | no | 120 |
| bypass-label | Deliberate soft-override label that bypasses validation. | no | friendlyai-bypass-ack-by-maintainer |
| methodology | Review methodology: "standard" (normal code review — default) or "story-driven" (adds linked-story and acceptance-criteria checks). Set "story-driven" only on repos that use a story-driven process. | no | standard |
Outputs
| name | description |
|---|---|
| conclusion | Final check-run conclusion: success | failure | neutral. |
| finding-count | Number of findings returned by the review service. |
| blocking-finding-count | Number of medium/high/critical findings returned by the review service. |
| max-risk | Maximum risk level returned by the review service, when provided. |
| report-url | Optional report URL returned by the review service. |
| unavailable-reason | Set when the review could not run and the check failed open to neutral: credits | timeout | upstream | malformed. |