git-worktrees
Pass
Audited by Gen Agent Trust Hub on Feb 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION] (SAFE): Utilizes routine Git commands (worktree, list, prune, cherry-pick) and filesystem operations (mkdir) to create isolated development environments. These actions are consistent with the skill's stated purpose and are limited to the local codebase.
- [PROMPT_INJECTION] (LOW): Indirect Prompt Injection surface (Category 8) detected in the Task tool invocation workflow. 1. Ingestion points: The [plan] placeholder in the Task prompt template in SKILL.md. 2. Boundary markers: Absent; the template does not use delimiters to wrap the untrusted plan input. 3. Capability inventory: Spawns sub-agents with background execution capabilities via the Task tool. 4. Sanitization: None; the workflow does not suggest any validation or sanitization of the plan text before interpolation into the agent prompt.
Audit Metadata