Skills
Navigate
skills/agentient/vibekit/mcp-resource-exposure-patterns/Gen Agent Trust Hub

mcp-resource-exposure-patterns

Pass

Audited by Gen Agent Trust Hub on Mar 31, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill provides instructional content and code templates for Model Context Protocol (MCP) resources, focusing on secure implementation practices and proper architectural design.
  • [SAFE]: The code snippets include explicit security validation patterns, such as path traversal prevention using Path.resolve() and verification against allowed_roots using is_relative_to.
  • [SAFE]: The skill correctly differentiates between read-only 'Resources' and action-oriented 'Tools', encouraging developers to follow the principle of least privilege.
  • [SAFE]: Dependencies are limited to standard Python libraries or well-known, reputable packages like pydantic and aiofiles.
  • [SAFE]: External links reference official protocol specifications (modelcontextprotocol.io), established documentation (IETF RFCs), and the official GitHub repository for the aiofiles library.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 31, 2026, 09:08 AM
Security Audit — agent-trust-hub — mcp-resource-exposure-patterns