kosli

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli package from the NPM registry. This is a legitimate tool provided by the platform to manage service integrations.
  • [COMMAND_EXECUTION]: Executes shell commands using the membrane CLI to perform DevOps tasks, such as tracking deployments and managing attestations. These operations are consistent with the skill's documented purpose.
  • [DATA_EXFILTRATION]: The skill adheres to security best practices by explicitly instructing the agent and user to manage credentials through the platform's connection manager, preventing the exposure of API keys.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 01:50 PM
Security Audit — agent-trust-hub — kosli