specsfy-release-cli

Pass

Audited by Gen Agent Trust Hub on Sep 4, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes data from external files, such as release notes and the project changelog, which are then used as content for git commits, tags, and GitHub releases.\n
  • Ingestion points: The script release_changelog.py reads user-supplied release notes (e.g., from /caminho/notas.md) and the local cli/CHANGELOG.md file.\n
  • Boundary markers: The skill requires notes to follow a specific structure (avoiding nested version headers) and validates version strings against a strict numeric SemVer regular expression.\n
  • Capability inventory: The skill executes significant operations including git push, git tag, gh release create, and npm publish which could be influenced by malicious content in the ingested files.\n
  • Sanitization: Basic structural validation and whitespace stripping are performed by the accompanying Python script, though it does not perform advanced natural language filtering on the release notes content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 4, 2026, 06:09 PM
Security Audit — agent-trust-hub — specsfy-release-cli