using-git-worktrees
Pass
Audited by Gen Agent Trust Hub on Sep 23, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes system commands to manage Git worktrees and project environments. This includes directory creation, Git operations (git worktree add, git add, git commit), and running project-specific setup and test tools (npm, cargo, pip, poetry, go, pytest). These commands are standard for development automation and are used to fulfill the skill's primary purpose.
- [INDIRECT_PROMPT_INJECTION]: The skill retrieves configuration and preferences from project-local files, which serves as a potential vector for indirect prompt injection.
- Ingestion points: Reads CLAUDE.md to identify preferred worktree directories and .agent/config.yml to check for auto_commit preferences.
- Boundary markers: Not explicitly defined; the agent is instructed to use the information found in these files to guide its subsequent actions.
- Capability inventory: The skill possesses capabilities for file system access, command execution (package managers and test runners), and Git repository modification.
- Sanitization: No specific sanitization or validation logic is mentioned for the content parsed from external configuration files before it influences command paths or logical branching.
Audit Metadata