The Agent Skills Directory

[SAFE]: The skill instructions do not contain any malicious patterns such as prompt injection, obfuscation, or unauthorized data access.
[EXTERNAL_DOWNLOADS]: The skill instructs the user to install @membranehq/cli via npm. This is a standard installation of the vendor's official command-line interface tool needed for the skill to function.
[COMMAND_EXECUTION]: The skill uses various membrane CLI commands to manage authentication and execute actions. These commands are documented for their intended purpose of interacting with the Membrane integration platform.
[CREDENTIALS_UNSAFE]: No hardcoded credentials or sensitive files are accessed. The skill correctly directs users to use the platform's authentication mechanism (membrane login), which is a recommended security practice for managing secrets.

product-fruits