changelog-orchestrator
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local Python scripts (validate_config.py, render_template.py, quality_score.py) and standard development tools (git, gh) to process repository data and manage pull requests.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data that could contain malicious instructions aimed at overriding agent behavior during changelog synthesis.
- Ingestion points: Merged PR titles/descriptions, issue summaries, git commit logs, and Slack channel messages.
- Boundary markers: The skill utilizes markdown templates with double-brace placeholders (e.g., {{features}}) to structure data.
- Capability inventory: The skill is permitted to read/write files and execute git/github commands for repository management.
- Sanitization: Content is injected into templates via direct string replacement in render_template.py without specific sanitization of the external input.
Audit Metadata