The Agent Skills Directory

[PROMPT_INJECTION]: The skill documents commands to ingest data from external sources (GitHub issues, pull requests, code, and logs), creating a surface for indirect prompt injection where malicious instructions could be embedded in the retrieved content. * Ingestion points: gh issue view, gh pr view, gh search code, and gh run view --log referenced in SKILL.md. * Boundary markers: The skill includes a clear directive: "IMPORTANT: This skill is for READ-ONLY operations only. DO NOT execute any commands that create, modify, or delete resources." * Capability inventory: The skill provides documentation for read-only CLI commands to be executed by the agent's host environment. * Sanitization: No sanitization or output filtering is specified within the reference.

gh-reference