The Agent Skills Directory

[SAFE]: The skill is primarily documentation-based, providing structured steps and checklists for an AI agent to conduct code reviews. The logic is purely instructional and aimed at providing feedback on provided source code.
[CREDENTIALS_UNSAFE]: A placeholder API key ('sk-1234567890abcdef') is included in an example section of SKILL.md. This is explicitly labeled as a 'Bad' practice example to teach users not to hardcode secrets and does not represent an actual credential leak.
[DATA_EXFILTRATION]: The skill configuration allows the use of 'Read', 'Grep', and 'Glob' tools for file analysis. There are no tools or instructions that enable network access or the transmission of analyzed data to external servers.

code-review