The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a variety of standard git commands (e.g., git worktree add, git worktree list, git worktree remove) to perform its primary function of repository management. It includes safeguards such as checking for uncommitted changes and unpushed commits before allowing the removal of a worktree, and it explicitly requires user confirmation for destructive actions.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from the local environment, specifically output from git commands like branch names and status information, to construct tables and subsequent commands.
Ingestion points: Git repository state (branch names, status short-format, and commit logs) and user-provided worktree names via interactive prompts.
Boundary markers: The instructions do not define explicit boundary markers for the data interpolated into shell commands.
Capability inventory: The skill has the capability to execute shell commands for git operations and directory switching.
Sanitization: There is no explicit sanitization step for the branch names or worktree paths before they are used in command strings, representing a theoretical surface for command injection if the underlying repository metadata is maliciously crafted. However, given the primary purpose and the inclusion of manual confirmation steps, this is assessed as a low-risk surface.

worktree-ops