The Agent Skills Directory

[SAFE]: The skill provides a set of instructions for managing Git worktrees with appropriate safety guards.
[COMMAND_EXECUTION]: The skill executes Git shell commands to manage worktrees. Security is maintained through the use of native Git tools and explicit state verification steps.
[INDIRECT_PROMPT_INJECTION]: The skill processes external data from Git outputs and user inputs. It employs a sanitization strategy for folder names to prevent command injection and uses repository state checks to ensure operations are performed on valid targets. 1. Ingestion points: Git command output (git worktree list, git status) and user-supplied branch/task names. 2. Boundary markers: The skill uses Git's porcelain output format which is structured. 3. Capability inventory: Shell command execution (git, mkdir, printf), directory traversal, and context switching (cd). 4. Sanitization: Employs a specific slugging algorithm to sanitize folder names, rejecting reserved labels and shell-sensitive characters.

git-worktree