skills/patinaproject/skills/ready-pr/Gen Agent Trust Hub

ready-pr

Pass

Audited by Gen Agent Trust Hub on Aug 12, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [DYNAMIC_EXECUTION]: The skill uses a Perl one-liner to implement a 10-minute timeout for watching GitHub checks. This script serves as a portable wrapper for the gh pr checks command when native timeout utilities are unavailable. The execution is limited to specific GitHub CLI commands.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub PR comments and reviews (Ingestion points: workflows/ready-for-merge.md). It mitigates risks using a triage state machine and an explicit human-authored thread handoff rule (Boundary markers: triage.md). Its capabilities include file writes, commits, and PR status changes (Capability inventory: git, gh). It relies on operator review for human feedback rather than automated parsing (Sanitization: triage state machine).
  • [DATA_EXPOSURE_&_EXFILTRATION]: While the skill interacts with the file system and GitHub APIs, it contains explicit guardrails to stop execution if secrets, ambiguous permissions, or out-of-scope files are detected. All network operations are directed to GitHub, a well-known service.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 12, 2026, 08:13 PM
Security Audit — agent-trust-hub — ready-pr