keep-a-changelog

Pass

Audited by Gen Agent Trust Hub on Oct 23, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data by reading commit histories via git commands, existing CHANGELOG.md files, and project manifests such as package.json, Cargo.toml, and pyproject.toml. If these sources contain malicious instructions—for example, a maliciously crafted commit message—they could attempt to influence the agent's output or behavior during the changelog generation or release process.
  • Ingestion points: Processes content from CHANGELOG.md, package.json, Cargo.toml, pyproject.toml, and git log output as described in the 'Workflows' and 'Version source' sections of SKILL.md.
  • Boundary markers: The skill does not define specific boundary markers or instructions for the agent to ignore embedded commands when reading these external files.
  • Capability inventory: The skill is authorized to modify local files like CHANGELOG.md and package.json. It also mentions capabilities for git operations (commit, tag, push) and GitHub actions (gh release create) if explicitly requested by the user.
  • Sanitization: The instructions do not specify any sanitization, filtering, or validation for the content read from external files before it is processed or used in file updates.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 23, 2026, 01:56 PM