The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands such as git diff, npm run test, and make check to determine the scope of changes and verify code quality.
[PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface. Ingestion points: untrusted data enters the context via PR descriptions, linked issues, task specifications, and file diffs. Boundary markers: the skill lacks explicit delimiters or instructions to ignore embedded commands in the data being reviewed. Capability inventory: the agent is authorized to run subprocesses such as npm run test and make check. Sanitization: there is no mention of validating or escaping scripts within configuration files like package.json or Makefile before they are executed. A malicious codebase could leverage these scripts to perform unauthorized actions when the agent performs a review.

code-review