finishing-a-development-branch
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes local shell commands for repository management, including Git operations (
git checkout,git pull,git merge,git push) and GitHub CLI interactions (gh pr create). It also invokes project-specific test runners such asnpm test,cargo test,pytest, andgo testto verify code quality before proceeding. - [INDIRECT_PROMPT_INJECTION]: The skill operates on repository metadata that could be influenced by external contributors, presenting a theoretical attack surface for indirect prompt injection.
- Ingestion points: Processes branch names, commit lists, and test suite outputs retrieved from the local environment.
- Boundary markers: Does not implement specific delimiters or 'ignore' instructions for repository-sourced strings, though it uses a highly structured workflow to guide agent behavior.
- Capability inventory: Includes network operations (
git push,gh pr create), file system modifications (git branch -D), and worktree removal (git worktree remove). - Sanitization: No explicit sanitization or escaping is defined for dynamic values like branch names or PR titles before they are interpolated into shell commands.
Audit Metadata