The Agent Skills Directory

[SAFE]: The skill follows security best practices by delegating authentication and credential management to the Membrane platform rather than requesting or storing sensitive API keys locally.
[EXTERNAL_DOWNLOADS]: The skill recommends installing the @membranehq/cli package from the official NPM registry. This is a legitimate tool provided by the vendor for managing integrations.
[COMMAND_EXECUTION]: The skill uses standard CLI commands (membrane login, membrane connect, membrane action run) to interact with the service. These operations are within the expected scope of an integration skill and do not involve unauthorized system modifications or privilege escalation.
[DATA_EXFILTRATION]: No patterns of unauthorized data exfiltration were detected. Network communication is directed to the vendor's established infrastructure (getmembrane.com) for the purpose of executing accounting tasks.

mip-fund-accounting