clean-gone-branches
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes output from Git commands which can contain untrusted data in the form of branch names created by other users. Ingestion points: Data enters the agent context through the output of
git branch -vvandgit worktree list(SKILL.md). Boundary markers: The instructions do not define clear boundaries or provide instructions to ignore embedded commands within the tool output. Capability inventory: The agent is authorized to execute branch and worktree deletion commands based on the processed output. Sanitization: There is no explicit requirement for the agent to sanitize or validate branch names before execution. - [COMMAND_EXECUTION]: The skill involves the execution of local Git commands to modify the repository state. Evidence: Instructions explicitly call for running
git fetch --prune,git branch -vv, andgit worktree list, followed by deletion commands for identified branches and worktrees.
Audit Metadata