The Agent Skills Directory

[SAFE]: The skill's functionality is purely analytical and instructional. It utilizes read-only tools (Read, Grep, Glob) to examine local source code files and does not possess capabilities for file modification, network communication, or arbitrary command execution.
[SAFE]: No obfuscated code, hardcoded credentials, or suspicious remote dependencies were found in the skill files or metadata.
[SAFE]: Indirect prompt injection risk is inherently low as the skill processes data as text for review purposes and lacks the high-privilege tools (like shell access or network requests) required to execute malicious payloads embedded in reviewed code.

code-reviewer