plaid

Pass

Audited by Gen Agent Trust Hub on Apr 29, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: Installs the official Membrane CLI from the vendor's package registry to manage the integration. Evidence: npm install -g @membranehq/cli@latest.
  • [COMMAND_EXECUTION]: Utilizes the membrane CLI to perform operations such as logging in, connecting to Plaid, and running financial actions.
  • [COMMAND_EXECUTION]: Supports dynamic creation of new integration actions via natural language descriptions through the vendor's platform. Evidence: membrane action create.
  • [PROMPT_INJECTION]: Processes financial data from Plaid, creating a potential surface for indirect prompt injection.
  • Ingestion points: Results from membrane action run containing bank account data.
  • Boundary markers: Absent; no specific delimiters are defined to isolate external data in the tool output.
  • Capability inventory: Shell command execution using the membrane CLI to read and write data.
  • Sanitization: Absent; the agent is instructed to process data directly from the CLI tool.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 29, 2026, 08:42 PM
Security Audit — agent-trust-hub — plaid