The Agent Skills Directory

Indirect Prompt Injection (LOW): The skill is designed to analyze and act upon repository data (diffs, logs, and files) which may contain untrusted content from external contributors.
Ingestion points: Data enters the agent's context through tools like git diff, git log, and git show (via the Bash tool), as well as the Read and Grep tools.
Boundary markers: There are no explicit instructions or delimiters used when processing Git output to prevent the agent from following instructions embedded within commit messages or code changes.
Capability inventory: The skill allows the use of the Bash tool, which provides a high-privilege surface for executing commands based on interpreted data.
Sanitization: The skill does not implement sanitization or filtering for the Git data it processes, relying on the underlying agent's own safety filters.

git-workflow