worktree
Warn
Audited by Gen Agent Trust Hub on Apr 21, 2026
Risk Level: MEDIUMCOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill relies on shell command execution to interact with the Git repository and the local filesystem.
- Evidence: The
SKILL.mdfile contains several bash script blocks designed to executegit worktree list,git worktree add, andgit worktree removecommands. - [COMMAND_EXECUTION]: User-provided input is interpolated directly into shell scripts, posing a risk of command injection.
- Evidence: In the 'Create New Agent Worktree' section, the skill instructions state
NAME="" # Set from ARGUMENTS if present. This user-controlled variable is then used to construct theWT_PATHandBRANCHvariables, which are passed to thegit worktree add "$WT_PATH" -b "$BRANCH"command. If the user input includes shell metacharacters such as backticks, semicolons, or subshell expansions (e.g.,$(...)), it could lead to the execution of arbitrary commands within the agent's environment.
Audit Metadata