Skills
skills/membranedev/application-skills/google-cloud-dataflow/Gen Agent Trust Hub

google-cloud-dataflow

Pass

Audited by Gen Agent Trust Hub on Apr 30, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill utilizes the membrane CLI to manage connections and execute actions within the Google Cloud Dataflow environment.
  • [EXTERNAL_DOWNLOADS]: Installs the @membranehq/cli package from the npm registry. This is an official, vendor-provided tool required for the skill's operation.
  • [PROMPT_INJECTION]: Indirect prompt injection surface exists through the use of natural language in CLI commands for action discovery and creation. This is a primary feature of the platform and is managed via the vendor's service infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 30, 2026, 05:37 PM
Security Audit — agent-trust-hub — google-cloud-dataflow