Skills
skills/membranedev/application-skills/builtwith/Gen Agent Trust Hub

builtwith

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use the Membrane CLI (membrane) to manage authentication, list connections, and execute actions.\n
  • Evidence: membrane login, membrane connection list, and membrane action run commands found in SKILL.md.\n- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the @membranehq/cli tool from the official NPM registry.\n
  • Evidence: npm install -g @membranehq/cli@latest in SKILL.md.\n- [PROMPT_INJECTION]: The skill provides mechanisms to find or create actions based on natural language input, which serves as a surface for indirect prompt injection.\n
  • Ingestion points: intent "QUERY" and action create "DESCRIPTION" parameters in SKILL.md.\n
  • Boundary markers: None identified in the command examples to isolate user-provided strings.\n
  • Capability inventory: The skill can execute platform-defined actions which may include network operations and data retrieval.\n
  • Sanitization: Input validation and sanitization are handled by the Membrane CLI and platform infrastructure.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 12:02 PM
Security Audit — agent-trust-hub — builtwith