graphy

Pass

Audited by Gen Agent Trust Hub on Sep 20, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs the agent to install the Membrane CLI tool (@membranehq/cli) from the npm registry. This is a vendor-controlled utility necessary for interaction with the Graphy service.
  • [COMMAND_EXECUTION]: The instruction set includes running shell commands via the membrane CLI for authentication, connection setup, and data retrieval.
  • [INDIRECT_PROMPT_INJECTION]: The skill accepts natural language intents and JSON input that are passed as arguments to CLI tools, representing a standard injection surface for tool-using agents.
  • Ingestion points: Intent strings and JSON data parameters in SKILL.md command templates.
  • Boundary markers: None explicitly defined.
  • Capability inventory: Shell command execution and network proxying via Membrane.
  • Sanitization: Relies on the agent harness or utility to handle input validation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 20, 2026, 04:00 AM
Security Audit — agent-trust-hub — graphy