The Agent Skills Directory

[SAFE]: The skill implements standard development practices for reviewing code, running tests, and managing branches using well-known tools like git and gh.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and act upon data from external, potentially untrusted sources such as code diffs and chat history.
Ingestion points: Processes git diff output, commit messages, and recent chat history to determine intent and correctness.
Boundary markers: The instructions do not define specific delimiters or instructions to ignore embedded commands within the code it reviews.
Capability inventory: The skill uses git for branch management and gh (GitHub CLI) for pull request operations.
Sanitization: No explicit sanitization or filtering of the ingested diffs or chat content is specified in the workflow.

review-and-ship