The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes data from the local git environment (branch names, worktree lists, and status output) which can be influenced by external contributors in shared repositories.
Ingestion points: The agent reads output from commands like git branch -vv, git worktree list, and git status in SKILL.md and reference files.
Boundary markers: The skill explicitly instructs the agent to present a comprehensive plan to the user and wait for confirmation before executing destructive commands, providing a human-in-the-loop safety mechanism.
Capability inventory: The skill utilizes the Bash tool to perform branch deletions (git branch -d), worktree removals (git worktree remove --force), and remote ref pruning (git push origin --delete).
Sanitization: The skill implements strict classification rules in references/branch-classification.md and references/worktree-cleanup-rules.md to ensure only merged or identical content is targeted for removal.

git-post-merge-cleanup