The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses standard command-line tools for repository management and GitHub interaction.
Evidence: The instructions in SKILL.md (Steps 1, 2, 7) utilize git diff, git log, gh pr view, gh pr diff, gh pr list, gh issue list, and gh api for reading PR data and posting review comments.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from external pull requests that could contain malicious instructions designed to influence the agent's behavior.
Ingestion points: In SKILL.md (Step 1), the agent is instructed to read the PR title, body, author, diff, and earlier review comments using gh pr view, gh pr diff, and gh api.
Boundary markers: The skill does not define specific delimiters or instructions to ignore potentially malicious embedded content within the ingested PR data.
Capability inventory: The skill possesses read capabilities for local and remote repository data and write capabilities to the GitHub API (posting reviews/comments).
Sanitization: There are no explicit sanitization or validation steps for the retrieved external data before it is processed by the agent.

pr-review