closeout
Pass
Audited by Gen Agent Trust Hub on Aug 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes multiple shell commands using
gitandgh(GitHub CLI) to manage the repository state. This includes destructive operations such as merging branches and removing git worktrees. - [PROMPT_INJECTION]: The skill performs a 'Review currency' check by fetching and parsing the body of a GitHub Pull Request, which is untrusted external data. This creates an indirect prompt injection surface where malicious content in a PR body could theoretically attempt to influence the agent's behavior.
- Ingestion points: The PR body is ingested in
SKILL.md(Step 2) usinggh pr view. - Boundary markers: No specific delimiters or safety instructions are used to isolate the untrusted PR body content from the agent's logic.
- Capability inventory: The skill possesses the capability to merge PRs, delete branches, and remove directories from the local filesystem.
- Sanitization: The implementation uses a strict regular expression to ensure only a 40-character hexadecimal string (a git SHA) is extracted from the PR body, which effectively prevents common command injection attacks.
Audit Metadata