The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on executing local shell commands and a bundled audit script to analyze skill structure and file organization. This includes using grep to find skill callers and git for version control operations during skill updates.
[PROMPT_INJECTION]: The skill processes the content of other agent skills as data, which is a potential surface for indirect prompt injection. This risk is mitigated by explicit instructions to treat all reviewed skill content as non-authoritative data rather than as guidance for the agent's own behavior.
[SAFE]: No malicious patterns such as credential harvesting, obfuscation, or network-based data exfiltration were found. The tool is designed for project maintenance and follows best practices for tool execution.

skill-review