The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content from various sources, which could potentially contain malicious instructions intended to influence the agent's behavior during a review.
Ingestion points: The instructions in SKILL.md direct the agent to process staged/unstaged diffs, untracked files, pull request metadata, and requirements fetched from user-supplied URLs or issue trackers.
Boundary markers: The skill does not define specific delimiters or "ignore instructions" warnings to separate untrusted code or external documentation from the agent's operational instructions.
Capability inventory: The skill utilizes tools for fetching codebase context, Library/API context, and performing Web Searches to inform its findings.
Sanitization: There are no instructions for sanitizing or validating external content (e.g., stripping HTML/Markdown comments) before it is analyzed by the agent.

code-reviewing