finishing-a-development-branch
Pass
Audited by Gen Agent Trust Hub on Sep 28, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands including
git checkout,git merge,git branch, andgh pr createas documented inreferences/decision-rules.mdandreferences/examples.md. These are standard development operations. It also usesmr_worktree_remove, which is a vendor-specific tool for managing git worktrees. - [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could contain malicious instructions. 1. Ingestion points: The agent reads branch names and target base branches from the environment, and captures output from
npm testas shown inreferences/examples.md. 2. Boundary markers: The skill includes a mandatory announcement "I'm using the finishing-a-development-branch skill..." and requires the user to type "discard" to confirm deletion (Option 4). 3. Capability inventory: The skill can perform local file system changes (git) and network requests (gh CLI). 4. Sanitization: No sanitization is performed on the ingested branch names or test outputs before they are processed.
Audit Metadata