charliechenye/SkillGate

Pre-merge and pre-install trust checks for AI-agent skills and MCP configurations. Scan capabilities, detect risky changes, and block unapproved agent behavior in CI.

View on GitHub

Trust Signals

Scorecard Score
not yet scored
Maintenance Recency
Activelast commit Jul 9, 2026
License
MIT

Pinned Snippet

workflow.ymlSHA-pinned
uses: charliechenye/skillgate@14c48ef558e87a9eadd7366c0100b4389bd22767 # v0.1.2

tags can be moved; commit SHAs can't. why a SHA?

namedescriptionrequireddefault
pathRepository path to scan.no.
policyPolicy file path in the repository being scanned. Required for blocking policy checks.no""
baselineOptional baseline lockfile path for diff checks.no""
fail-on-driftExit nonzero when baseline drift is detected without requiring a policy file.nofalse
provenanceOptional provenance manifest path for policy and baseline checksum verification.no""
sarif-outputOptional SARIF output path.no""
summary-outputOptional Markdown review summary output path.no""
json-outputOptional machine-readable review JSON output path.no""
step-summaryAppend the Markdown review summary to the GitHub Step Summary.nofalse
mcpb-pathOptional committed MCPB artifact path to scan.no""
mcpb-fail-onOptional MCPB severity threshold that exits nonzero when matched.no""
mcpb-sarif-outputOptional MCPB SARIF output path for separate code scanning upload.no""

no outputs