resolve-review

Pass

Audited by Gen Agent Trust Hub on May 6, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it ingests and processes GitHub PR comments to guide code modifications and verification steps.
  • Ingestion points: PR comments and review summaries are fetched via GitHub CLI tools in SKILL.md and REFERENCE.md.
  • Boundary markers: The skill uses reasoning-based safeguards like 'State your understanding' but lacks programmatic delimiters for untrusted data.
  • Capability inventory: The skill has extensive capabilities including file system access, Git push operations, GitHub API write access, and execution of local test suites.
  • Sanitization: No explicit validation or structural escaping is performed on the content of the review comments before processing.
Audit Metadata
Risk Level
SAFE
Analyzed
May 6, 2026, 09:16 AM