pr-comment-triage
Pass
Audited by Gen Agent Trust Hub on Aug 22, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted data from pull request comments, which presents a surface for indirect prompt injection. A malicious comment could attempt to influence the agent's code modifications or pull request management actions. \n
- Ingestion points: Pull request comments and review summaries are fetched using
gh apiandgh pr viewcommands (Step 2: Build a complete comment inventory). \n - Boundary markers: The skill does not define explicit delimiters or instructions to ignore instructions embedded within the ingested comment content. \n
- Capability inventory: The agent has the authority to modify the filesystem to address feedback and perform GitHub API mutations to post replies and resolve threads. \n
- Sanitization: No automated sanitization of the comment text is performed, though the skill provides instructions for the agent to manually verify claims against the code (Step 4: Verify against the current code). \n- [COMMAND_EXECUTION]: The skill relies on the
gh(GitHub) CLI to interact with external PR data and perform state changes. This includes executing GraphQL mutations for resolving threads and posting replies, which are standard for the tool's purpose.
Audit Metadata