MCP Integration

Pass

Audited by Gen Agent Trust Hub on Sep 3, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious patterns or security vulnerabilities were detected. The skill is an educational resource that provides templates and guides for Model Context Protocol (MCP) integration.
  • [REMOTE_CODE_EXECUTION]: The documentation references the use of npx -y to run official MCP servers like @modelcontextprotocol/server-filesystem. This is a standard deployment pattern for the protocol.
  • [DYNAMIC_EXECUTION]: The skill describes the use of a headersHelper script to dynamically generate authentication tokens. This is a platform-supported configuration for managing short-lived credentials.
  • [INDIRECT_PROMPT_INJECTION]: The skill documents how to handle data from external services. It includes advice on validating inputs and handling errors to ensure the agent processes external tool outputs securely.
  • [CREDENTIALS_SAFE]: The skill correctly instructs users to manage secrets using environment variables (e.g., ${API_TOKEN}) and explicitly warns against hardcoding credentials in configuration files.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 3, 2026, 02:29 PM
Security Audit — agent-trust-hub — MCP Integration