oce-mcp-integration

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill content is educational and provides a boilerplate implementation for a TypeScript MCP server using the fastmcp library.
  • [CREDENTIALS_UNSAFE]: No hardcoded secrets or API keys are present; the instructions explicitly guide developers to use environment variables for sensitive tokens.
  • [INDIRECT_PROMPT_INJECTION]: The skill defines an interface for reading external ERP data into an agent's context, creating an indirect prompt injection surface. However, this is inherent to the skill's purpose as a tool-use bridge. The provided template includes standard sanitization via encodeURIComponent and limits capabilities to specific REST API endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 01:21 AM
Security Audit — agent-trust-hub — oce-mcp-integration