Schema

Strict Boundary Validation: Schema Definitions

💎 Value Proposition

MySQL MCP is a production-ready integration engineered for AI agents. It minimizes LLM token consumption by up to 90% via sandboxed Code Mode. It scales reliably through built-in connection pooling. It secures database access using strict OAuth 2.1 validation.

Reliable Data Processing: Handling Invalid Inputs

We treat schema validation as a strict data boundary. We strictly validate every input parameter, ensuring reliable workflows.

Enforce Reliability: Enterprise Validation Rules

We strictly enforce these schemas at the MCP boundary for secure integrations:

1. Strict Error Responses: Failing tools return structured error responses. Errors extend MySQLMcpError and use formatHandlerError(). This empowers autonomous workflows. The schema is: { success: false, error: string, code: string, category: string, suggestion: string, recoverable: boolean }
2. Deterministic Table Queries: Table queries return { exists: false, table: string } for missing tables. This enables graceful fallback handling in AI workflows.
3. Sandboxed File Operations: We bound file interactions using ALLOWED_IO_ROOTS. This prevents directory traversal and ensures isolated, secure execution.
4. Standard Schema Interoperability: Tools use Zod's Standard Schema implementation. This supports sophisticated aliases and type coercion while maintaining JSON Schema compatibility.
5. Static Schemas: All schemas must be defined statically as constants at the module level.
6. Environment Validation: Enforce the validation of process.env at startup using Zod.

Build with Confidence: Integration Guide

See the repository source for Zod schemas (src/adapters/mysql/schemas/). We recommend using Zod's .safeParse(input) to handle validation errors gracefully without exceptions.

For runtime schema introspection, see the Code-Mode documentation. It lets you discover API parameters programmatically using the .schema() introspection method.