The Agent Skills Directory

[SAFE] (SAFE): Analysis of the skill instructions and example TypeScript code confirms the absence of malicious intent. The logic focuses on local service discovery and user communication.
[DATA_EXFILTRATION] (INFO): The example code utilizes the fetch API to query http://localhost:1234. As this targets the local loopback address for a standard model server (LMStudio), it does not constitute unauthorized data exfiltration.
[DATA_EXPOSURE] (INFO): Metadata mentions a prior session involving 'Environment variable debugging.' While printing computed paths can lead to information disclosure, the current skill implementation does not include this functionality and only provides instructions for graceful degradation.

graceful-degradation