v1.0.3
ModelBound Skill Check Action v1.0.3
Lint, trust-score, and optimize-dry-run agent skill files in GitHub Actions — with public README badges on modelbound.co.
Highlights
- Report drill-down — PR comments and workflow outputs link to per-skill findings on modelbound.co
scan-all-on-main— optional input to rescan all tracked skill files on main when a push does not touch skills (keeps badges fresh)- Node 24 — matches current GitHub Actions runner defaults
- Richer PR comments — trust findings, MCP lint output, badge and report URLs
Quick start
- uses: ModelBound/skill-check-action@v1.0.3
with:
mode: full
scan-all-on-main: true
env:
MODELBOUND_API_KEY: ${{ secrets.MODELBOUND_API_KEY }}Marketplace
This release is published for the GitHub Marketplace. After installing, add a MODELBOUND_API_KEY repository secret (mb_live_… from API Keys).