The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to interact with the local repository and the GitHub API.
Employs git commands (e.g., git merge-base, git log, git diff) to identify code changes.
Uses gh (GitHub CLI) for PR operations, including gh pr diff, gh pr checkout, and gh api to post review comments.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted data from external code changes.
Ingestion points: Reads git diff outputs, gh pr diff content, and raw source code files from the local branch.
Boundary markers: The skill does not define explicit delimiters or instructions to the agents to ignore embedded commands within the code being reviewed.
Capability inventory: The skill can execute shell commands, read local files, and communicate with the GitHub API to post comments.
Sanitization: Content is aggregated into a JSON payload for GitHub API reviews, but there is no evidence of filtering or escaping malicious instructions embedded in the input code.

review-crdb