create-pull-request-with-reviewers
Pass
Audited by Gen Agent Trust Hub on Jun 30, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses common command-line utilities
gitandgh(GitHub CLI) to perform repository analysis and interact with GitHub APIs. These operations are restricted to the local development environment and the user's authorized GitHub repositories. - [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because it reads untrusted content from git commit messages and diffs to generate pull request descriptions.
- Ingestion points: Git logs and diff output are read from the repository in Step 6.
- Boundary markers: The instructions do not define specific delimiters to isolate git data from generation prompts.
- Capability inventory: The skill can perform actions via the
gh pr createcommand. - Sanitization: No explicit sanitization of the commit data is performed.
- Mitigation: A human review step in Step 8 requires the user to verify all generated PR content before submission, providing a critical safety check against malicious output.
Audit Metadata