The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or data exfiltration attempts were detected in the skill files.
[COMMAND_EXECUTION]: The skill uses git rev-parse to retrieve commit SHAs and git diff to display changes. These are standard, local-only development operations used to provide context for the code review.
[PROMPT_INJECTION]: The skill is designed to process external inputs such as implementation descriptions and code diffs, which represents an attack surface for indirect prompt injection. However, this is inherent to the functional purpose of a code review tool.
Ingestion points: Variables {WHAT_WAS_IMPLEMENTED}, {PLAN_OR_REQUIREMENTS}, and {EVIDENCE} in code-reviewer.md are populated from the primary agent's task context.
Boundary markers: The template utilizes Markdown headers (e.g., ## Requirements/Plan) to separate untrusted data from instructions.
Capability inventory: The reviewer agent is limited to reading repository state via git diff as defined in code-reviewer.md.
Sanitization: No explicit sanitization or instruction-ignoring delimiters are defined for the interpolated variables.

requesting-code-review