j5ik2o-gh-pr-review-follow-up
Pass
Audited by Gen Agent Trust Hub on Jul 15, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell commands via the GitHub CLI ('gh') to interact with the repository's pull requests. Specifically, it runs 'gh auth status', 'gh pr view', and 'gh api graphql'. These are standard operations for a developer tool and are executed using the user's existing local authentication. The use of 'subprocess.run' in 'scripts/fetch_comments.py' to invoke these commands is handled correctly, using list-based arguments to prevent shell injection.
- [DATA_EXFILTRATION]: While the skill interacts with the network via 'gh api', it only communicates with official GitHub endpoints to fetch pull request metadata and comments. It does not access sensitive local files (like '.ssh' or '.aws' credentials) nor does it transmit data to unknown external servers.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from GitHub PR comments. However, it incorporates a human-in-the-loop workflow where the agent summarizes the comments and explicitly asks the user for confirmation before taking any action. This significantly reduces the risk of the agent autonomously following malicious instructions embedded in review comments.
Audit Metadata