branch-pr
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill interacts with the local file system and network via standard developer tools (
gitandgh). - It utilizes a temporary file (
/tmp/pr_body.md) to pass data to the GitHub CLI, which is a recognized safety pattern to prevent shell-quoting vulnerabilities and command injection from content containing special characters. - The instruction to
unset GITHUB_TOKENis a localized environment modification designed to resolve authentication conflicts within the GitHub CLI's internal credential management. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface (Category 8) by processing untrusted data from the repository.
- Ingestion points: Reads content from
git log,git diff, and repository-hosted templates (.github/pull_request_template.md). - Boundary markers: None identified; the skill directly interpolates analyzed code changes into the generated PR title and body.
- Capability inventory: Possesses the ability to write to remote repositories via
gh pr createandgh pr editcommands. - Sanitization: No explicit sanitization or filtering of commit messages or code diffs is implemented before they are processed by the LLM for summarization.
Audit Metadata