The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes shell commands to manage repository state and verify code quality. It executes git for version control, gh for GitHub interaction, and build tools like make and cargo to run tests and format checks.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests untrusted data from an external source and uses it to guide agent actions.
Ingestion points: The skill reads external content from GitHub issues and pull requests (titles, bodies, and comments) using gh issue view and gh pr view in SKILL.md.
Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the retrieved GitHub content as data only or to ignore embedded instructions.
Capability inventory: The agent has the authority to execute repository scripts (make, cargo), interact with the file system, and push changes directly to the main branch (git push origin main).
Sanitization: There is no evidence of sanitization or filtering of the content retrieved from GitHub before the agent processes it.

maintainer-sweep