review-wall-of-apps-prs

Pass

Audited by Gen Agent Trust Hub on Jul 16, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes the make check-wall-of-apps command on the branch head of a pull request to validate the schema. To mitigate risk, the instructions require the agent to inspect the full file list and diff before checkout, specifically rejecting unexpected code, scripts, or binary changes.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external pull requests, including JSON content and PR metadata, creating a potential indirect prompt injection surface. The instructions provide a defense-in-depth approach by:
    • Ingestion points: Accessing file content from docs/wall-of-apps.json and PR metadata.
    • Boundary markers: Using explicit instructions to treat submissions as "untrusted external contributions."
    • Capability inventory: Utilizing tools to list, checkout, and merge PRs, alongside the make execution.
    • Sanitization: Mandating a manual diff review to ensure no code or executable content is introduced alongside the intended data changes.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 16, 2026, 09:31 PM
Security Audit — agent-trust-hub — review-wall-of-apps-prs