token

Pass

Audited by Gen Agent Trust Hub on Apr 28, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes shell commands using the membrane CLI to manage connections, discover available actions, and execute platform workflows.
  • [EXTERNAL_DOWNLOADS]: Installs and utilizes the @membranehq/cli package from the npm registry. This is the official tool provided by the platform for managing integrations.
  • [PROMPT_INJECTION]: The skill incorporates natural language strings, such as search queries and action descriptions, directly into CLI command arguments. This creates a surface for indirect prompt injection if the inputs are derived from untrusted external data.
  • Ingestion points: Inputs for the --intent parameter in membrane action list and the description field in membrane action create (found in SKILL.md).
  • Boundary markers: No delimiters or instructions are used to distinguish untrusted data from the command context.
  • Capability inventory: The skill can execute arbitrary actions on the Membrane platform and perform network operations via the CLI.
  • Sanitization: There is no evidence of input validation or escaping for the natural language parameters passed to the shell.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 28, 2026, 10:53 PM
Security Audit — agent-trust-hub — token