changelog-automation

Pass

Audited by Gen Agent Trust Hub on Aug 10, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill recommends installing standard development packages from official NPM and PyPI registries, such as semantic-release, standard-version, and commitizen.
  • [COMMAND_EXECUTION]: Includes shell commands for Git configuration, environment setup, and automated release workflows, which are standard for the skill's stated purpose of changelog automation.
  • [PROMPT_INJECTION]: The skill instructions involve processing untrusted data from commit history and Pull Requests (Indirect Prompt Injection surface).
  • Ingestion points: Commit messages and Pull Request descriptions processed by automation tools (referenced in implementation-playbook.md).
  • Boundary markers: The skill relies on standard parsing delimiters and logic built into third-party tools like git-cliff and semantic-release.
  • Capability inventory: Capability to write to the filesystem (CHANGELOG.md), perform Git operations (tagging, pushing), and interact with the GitHub API for release creation.
  • Sanitization: The instructions explicitly mandate a manual review step for accuracy and include safety warnings to avoid exposing secrets or internal details in the generated notes.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 10, 2026, 01:43 PM
Security Audit — agent-trust-hub — changelog-automation