github-comment-cleanup

Pass

Audited by Gen Agent Trust Hub on Nov 7, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSPERSISTENCE
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The automation processes untrusted file content from external GitHub repositories to perform comment cleanup.
  • Ingestion Points: Candidate files from configured GitHub repositories (SKILL.md).
  • Boundary Markers: None explicitly defined in the setup instructions.
  • Capability Inventory: File editing, git commit/push, and Pull Request creation (SKILL.md).
  • Sanitization: Includes a verification step via verify_comment_only.py to ensure only comments are modified (SKILL.md).
  • [EXTERNAL_DOWNLOADS]: The skill instructs the creation of an automation using code sourced from a remote GitHub repository.
  • Evidence: source: "github:OpenHands/extensions" (SKILL.md).
  • Context: The source points to the vendor's official extensions repository.
  • [PERSISTENCE]: The skill facilitates the creation of a recurring automation triggered by a cron schedule.
  • Evidence: The request template includes a cron trigger for weekly execution (SKILL.md).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 7, 2026, 11:27 AM