The Agent Skills Directory

[Data Processing Surface]: The skill is designed to ingest and analyze external content from GitHub issue reports, comments, and pull request diffs. This represents a potential surface for indirect prompt injection where malicious instructions could be embedded in the data being reviewed.
[Mitigation Strategies]: The skill includes robust procedural safeguards to address untrusted input. It instructs the agent to treat proposed mechanisms as hypotheses rather than requirements, requires explicit proof of user impact before accepting claims, and enforces a desk-review-only workflow.
[Operational Constraints]: The instructions explicitly limit the agent to read-only access and forbid the execution of the GitHub CLI (gh) or any remote write operations (e.g., merging, commenting, labeling) without separate user authorization. These constraints effectively minimize the risk of the agent being manipulated into performing unauthorized actions based on the content of a pull request.

maintainer-review