The Agent Skills Directory

[NO_CODE]: The skill is a pure markdown/YAML instruction set for the AI. It does not include any scripts (Python, JS, Shell), compiled binaries, or library dependencies.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns were detected. There are no curl | bash or similar commands.
[CREDENTIALS_UNSAFE]: No hardcoded credentials or secrets were found in the instructions or metadata.
[EXTERNAL_DOWNLOADS]: The skill does not perform any network operations or download external files.
[DATA_EXFILTRATION]: There is no functionality for accessing sensitive local data or transmitting data to external servers.
[PROMPT_INJECTION]: The content defines a specialized persona for code review but does not attempt to bypass core AI safety guidelines or extract system prompts.
[SAFE]: The logic for reviewing code is contained within the prompt instructions and does not introduce risk to the host system.

adversarial-reviewer