The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from local git history and file diffs. If a commit message or the content of the files being analyzed contains malicious instructions, the agent might interpret them while planning the reorganization.
Ingestion points: The skill reads external data using git log, git diff, and git diff --name-only (SKILL.md).
Boundary markers: The skill includes a human-in-the-loop check by instructing the agent to show the plan to the user and ask for confirmation before proceeding (Step 3).
Capability inventory: The skill uses the Bash tool restricted to git commands, along with Read, Grep, and Glob tools (SKILL.md frontmatter).
Sanitization: No specific sanitization or delimiters are mentioned for the content retrieved from git history before it is processed by the agent.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute standard git commands locally (git reset, git add, git commit). These are legitimate for the skill's primary purpose of commit reorganization and include warnings about the destructive nature of history rewriting.

cleanup-commits