The Agent Skills Directory

[SAFE]: The skill is composed entirely of instructional Markdown files that guide an AI agent in the safe and effective use of Git. It does not include any executable scripts, remote downloads, or obfuscated content.
[PROMPT_INJECTION]: The skill establishes an indirect prompt injection surface by instructing the agent to ingest data from Git repositories, which may include untrusted content like commit messages or branch names. * Ingestion points: Processes repository data through commands like git log, git show, and git diff. * Boundary markers: No specific delimiters are mandated to isolate repository data from instructions. * Capability inventory: Utilizes the git binary, which has capabilities for file system access and network operations (push/pull). * Sanitization: No explicit sanitization of repository data is provided in the documentation.

Git