meta-github-pr-watch-digest

Pass

Audited by Gen Agent Trust Hub on Jun 30, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides legitimate productivity functionality by triaging GitHub PRs, comments, and CI statuses using standard CLI tools.
  • [PROMPT_INJECTION]: The skill processes untrusted external data (PR bodies, comments, and CI logs) which presents a standard surface for indirect prompt injection.
  • Ingestion points: Data is pulled from GitHub repos in the pull_prs step of SKILL.md.
  • Boundary markers: Not explicitly used for the external PR content before it is passed to the summarizer.
  • Capability inventory: The skill uses memory_save to persist content to the local filesystem at memory/pr-watch.md.
  • Sanitization: User-supplied filtering instructions are escaped using xml_escape, though the PR content itself is not sanitized before being stored in the memory file.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 30, 2026, 01:08 AM
Security Audit — agent-trust-hub — meta-github-pr-watch-digest