deepseek
Pass
Audited by Gen Agent Trust Hub on Apr 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill implements a standard integration with DeepSeek using the official Membrane CLI (@membranehq/cli). All identified resources, including the homepage and repository, belong to the verified vendor ('membranedev').
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the official Membrane CLI from the NPM registry. This is a legitimate dependency for the skill's operation within the vendor's ecosystem.
- [CREDENTIALS_UNSAFE]: The instructions follow security best practices by leveraging Membrane's internal connection management, explicitly advising against asking users for raw API keys or tokens.
Audit Metadata