cull-release-publish

Pass

Audited by Gen Agent Trust Hub on Jul 15, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill follows security best practices for release management, including verification of provenance, signatures, and notarization before publication.
  • [COMMAND_EXECUTION]: The skill executes a local npm command to retrieve release state. This is a standard development operation and is constrained by strict verification of the resulting JSON envelope.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests data from a JSON envelope generated by the local environment.
  • Ingestion points: Step 1 in SKILL.md parses JSON output from npm run release:cull.
  • Boundary markers: Not explicitly defined in the input stream, but the skill enforces strict validation of the prepared state and exact metadata.
  • Capability inventory: Git tag creation and pushing, monitoring GitHub release workflows, and Homebrew promotion.
  • Sanitization: The skill requires exact inventory matches and valid signatures/provenance before proceeding, minimizing the risk of processing malicious data.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 15, 2026, 07:29 AM
Security Audit — agent-trust-hub — cull-release-publish