The Agent Skills Directory

[SAFE]: The skill follows official vendor implementation patterns for EyePop.ai and Membrane. No malicious behavior or safety bypass attempts were detected.
[EXTERNAL_DOWNLOADS]: The skill recommends installing the @membranehq/cli package from npm. This is a verified vendor tool for the platform.
[COMMAND_EXECUTION]: Instructions involve using the membrane command-line tool for legitimate integration tasks such as logging in, listing actions, and running API requests.
[DATA_EXFILTRATION]: All network requests are routed through the Membrane proxy which manages authentication headers. There is no evidence of local credential theft or unauthorized data transmission.

eyepopai