The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates complex Git operations such as worktree listing, detaching branches, and switching branches using the scripts/worktree-context.sh helper script. These actions involve direct subprocess calls to the git binary and coordination with a sibling Node.js utility.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-controlled branch names and directory paths, which are then used in shell scripts and tool arguments, creating a potential vector for instruction injection.
Ingestion points: User-provided <branch> names and the --from <directory> path are ingested via the skill's command interface as described in SKILL.md.
Boundary markers: The skill's instructions do not specify delimiters or boundary markers to isolate user-supplied data from the instruction context.
Capability inventory: The skill possesses filesystem modification capabilities via Git and command execution capabilities via git and node subprocesses in both SKILL.md and scripts/worktree-context.sh.
Sanitization: While variable quoting is employed in the bash script to mitigate simple shell injection, the awk command in the holder_record function passes the branch name using a -v flag, which can be sensitive to crafted strings containing double quotes.

move-branch-here