The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes several bash snippets using standard Unix utilities (grep, find, head, ls) and git. These are used appropriately to gather architectural context, detect coding patterns, and analyze file churn within the local repository to provide informed code reviews.
[DATA_EXFILTRATION]: No network-enabled tools or commands (such as curl or wget) are used. The skill's analysis remains entirely local to the execution environment.
[CREDENTIALS_UNSAFE]: No hardcoded credentials or secrets were detected. While the skill specifically instructs the agent to look for security-sensitive paths (e.g., searching for 'auth', 'token', 'password'), this is performed as part of a legitimate security review process and the data is not sent externally.
[PROMPT_INJECTION]: The instructions are focused on providing high-quality technical feedback and do not attempt to bypass safety filters or override the agent's core instructions.

code-review