unlaunch

Pass

Audited by Gen Agent Trust Hub on May 3, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Recommends the installation of the @membranehq/cli package from the official NPM registry. This is a vendor-provided tool necessary for the skill's functionality.
  • [COMMAND_EXECUTION]: Uses the membrane CLI for operations such as authentication, action discovery, and running feature flag actions. These commands are restricted to the context of the Membrane platform.
  • [DATA_EXFILTRATION]: The skill follows secure patterns by delegating authentication to the Membrane platform and explicitly instructing the agent not to request or handle user API keys or tokens.
  • [SAFE]: Analysis of the instructions and CLI usage patterns shows no evidence of malicious intent, obfuscation, or unauthorized data access. The operations described are consistent with the skill's stated purpose.
Audit Metadata
Risk Level
SAFE
Analyzed
May 3, 2026, 01:02 AM