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
Risk Level
SAFE
Analyzed
Apr 22, 2026, 06:51 AM