The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified during the analysis of the skill's instructions and reference documents. The skill provides standard developer guidance for improving code quality and security.
[PROMPT_INJECTION]: The skill's instructions define a strict persona and set of constraints for conducting reviews. No attempts to override safety guardrails or bypass system instructions were detected. While the skill is designed to process untrusted code (Category 8 surface), it lacks the high-privilege capabilities such as network access or write permissions required to facilitate an exploit.
[DATA_EXFILTRATION]: There are no network requests, hardcoded credentials, or sensitive file access patterns. The skill restricts its operations to reading local files for analysis purposes using standard tools like Read, Grep, and Glob.
[REMOTE_CODE_EXECUTION]: The skill does not attempt to download external scripts or execute arbitrary commands. It lacks any package dependencies or dynamic code execution logic, adhering to its read-only analytical purpose.

code-reviewer