mcp-best-practices（MCP 最佳实践）

MCP Best Practices

Decision reference for building production MCP servers with the TypeScript SDK. Not a tutorial - assumes you already have a working server and need to make it correct, fast, and secure.

Quick Reference

v1 imports (production today):

import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js";
import { WebStandardStreamableHTTPServerTransport } from "@modelcontextprotocol/sdk/server/webStandardStreamableHttp.js";
import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js";

v2 imports (when stable):

import { McpServer } from "@modelcontextprotocol/server";
import { WebStandardStreamableHTTPServerTransport } from "@modelcontextprotocol/server";

Server Setup

Transport Decision

Stateless Pattern (recommended for remote deployment)

Per-request server+transport creation is the canonical pattern. Maintainer @ihrpr confirms: "each transport should have an instance of MCPServer" (#343). Sharing instances leaks cross-client data (GHSA-345p-7cg4-v4c7).

app.post("/mcp", async (c) => {
  const server = new McpServer({ name: "my-server", version: "1.0.0" });
  // Register tools, resources, prompts...
  registerTools(server);

  const transport = new WebStandardStreamableHTTPServerTransport({
    sessionIdGenerator: undefined,   // stateless - no session tracking
    enableJsonResponse: true,        // JSON responses, no SSE streaming
  });

  // All tools/resources must be registered before connect() (#893)
  try {
    await server.connect(transport);
    return transport.handleRequest(c.req.raw);
  } finally {
    await transport.close();
    await server.close();
  }
});

What to hoist to module level (don't recreate per request):

• Zod schemas (they never change)
• Annotation objects ({readOnlyHint: true,...})
• Tool description strings
• Payment configs, upstream API clients

The McpServer itself must be per-request, but its constant inputs should not be.

For deep dive on transports, sessions, HTTP/2 gotchas, and K8s deployment: see references/transport-patterns.md

Framework Integration

Hono (web-standard):

import { Hono } from "hono";
const app = new Hono();
app.post("/mcp", handleMcpRequest);  // WebStandardStreamableHTTPServerTransport
app.get("/mcp", handleMcpSse);       // Optional: SSE for server notifications
app.delete("/mcp", handleMcpDelete); // Optional: session termination

Cloudflare Workers: Same pattern - WebStandardStreamableHTTPServerTransport works natively in Workers runtime.

Express/Node (v2): Use @modelcontextprotocol/express middleware with NodeStreamableHTTPServerTransport (wraps the Web Standard transport for IncomingMessage/ServerResponse).

Tool Design

Registration API

v1 (current stable) - server.tool() works but has ambiguous overloads. Prefer the config-object form when possible:

server.tool("search_docs", "Search documents", {
  query: z.string().describe("Search query"),
  max_results: z.number().optional().describe("Max results (default 20)"),
}, { readOnlyHint: true, destructiveHint: false, idempotentHint: true, openWorldHint: true },
  async ({ query, max_results }) => { /* handler */ }
);

v2 (migration target) - registerTool() with config object:

server.registerTool("search_docs", {
  title: "Document Search",
  description: "Search documents by keyword or phrase",
  inputSchema: z.object({
    query: z.string().describe("Search query"),
    max_results: z.number().optional().describe("Max results (default 20)"),
  }),
  outputSchema: z.object({
    results: z.array(z.object({ id: z.string(), text: z.string() })),
    has_more: z.boolean(),
  }),
  annotations: { readOnlyHint: true, destructiveHint: false, idempotentHint: true, openWorldHint: true },
}, async ({ query, max_results }) => {
  const result = await fetchDocs(query, max_results);
  return {
    structuredContent: result,
    content: [{ type: "text", text: JSON.stringify(result) }],
  };
});

Naming

Spec (2025-11-25): 1-128 chars, case-sensitive. Allowed: A-Za-z0-9_-.

DO: search_docs, get_user_profile, admin.tools.list DON'T: search (too generic, collides across servers), Search Docs (spaces not allowed)

Service-prefix your tools (github_*, jira_*) when multiple servers are active - LLMs confuse generic names across servers.

Schema Rules

.describe() on every field - this is what LLMs use for argument generation.

For complete Zod-to-JSON-Schema conversion rules, what breaks silently, outputSchema/structuredContent patterns: see references/tool-schema-guide.md

Critical bugs:

• z.union() / z.discriminatedUnion() silently produce empty schemas on v1.x (#1643, fixed on main 2026-03-30). Use flat z.object() with z.enum() discriminator field instead until v1 ships the fix.
• Plain JSON Schema objects silently dropped before v1.28.0. Fixed in v1.28 - now throws at registration (#1596).
• z.transform() stripped during conversion - JSON Schema can't represent transforms (#702).
• Client-side AJV strict-mode rejection: Zod v4 z.object() produces JSON Schema with additionalProperties: false. The SDK's client validates structuredContent against outputSchema with AJV strict mode and rejects extra fields. Server-side .parse() strips extras silently, but the original structuredContent is sent to the client unchanged - so the server thinks it's fine and the client errors. Fix: explicitly .parse() upstream data before assigning to structuredContent, or use .passthrough() on schemas that intentionally pass through extra fields.

Annotations

All are optional hints (untrusted from untrusted servers per spec):

Set them accurately - clients use them for consent prompts and auto-approval decisions.

Open SEPs expanding annotations:

• #1913 Trust and Sensitivity - data classification hints
• #1984 Comprehensive annotations for governance/UX
• #1561 unsafeOutputHint - output may contain untrusted content
• #1560 secretHint - tool handles secrets/credentials
• #1487 trustedHint - server attestation of tool trustworthiness

The "Lethal Trifecta": Combining (1) access to private data + (2) exposure to untrusted content + (3) external communication ability creates data theft conditions. Researchers demonstrated this with a malicious calendar event, an MCP calendar server, and a code execution tool. Design tool sets to avoid granting all three simultaneously.

Evaluation framework for new annotation proposals:

1. What client behavior changes? (No concrete action = don't add it)
2. Does it require trust to be useful? (If yes, doesn't help against untrusted servers)
3. Could _meta handle it? (Namespaced metadata better for single-deployment needs)
4. Does it help reason about tool combinations?
5. Is it a hint or contract? (Contracts belong in auth/transport/runtime layer)

Error Handling

Two distinct mechanisms with different LLM visibility:

Per SEP-1303 (merged into spec 2025-11-25): input validation errors MUST be tool execution errors, not protocol errors. The LLM needs to see "date must be in the future" to self-correct.

// DO: Tool execution error - LLM can self-correct
return {
  isError: true,
  content: [{ type: "text", text: "Date must be in the future. Current date: 2026-03-25" }],
};

// DON'T: Protocol error for validation - LLM can't see this
throw new McpError(ErrorCode.InvalidParams, "Invalid date");

Known SDK behavior: When the SDK converts an McpError thrown from a tool handler into a CallToolResult, the error.data field is dropped. If you embed structured data in McpError's data field, it may not reach the client. The x402/MPP MCP ecosystem standardized on isError: true tool results with structuredContent for this reason. (One exception: code -32042 "Payment Required" survives McpServer end-to-end with error.data intact - see references/error-handling.md.)

For full error taxonomy, code examples, and payment error patterns: see references/error-handling.md

Resources and Instructions

Server Instructions

Set in the initialization response - acts as a system-level hint to the LLM about how to use your server:

const server = new McpServer({
  name: "docs-api",
  version: "1.0.0",
  instructions: "Knowledge base API. Use search_docs for full-text search, get_doc for retrieval by ID. All tools are read-only.",
});

Resource Registration

Expose documentation or structured data via docs:// URI scheme:

server.resource("search-operators", "docs://search-operators", {
  title: "Search Operators Guide",
  description: "Supported search operators and syntax",
  mimeType: "text/markdown",
}, async () => ({
  contents: [{ uri: "docs://search-operators", text: operatorsMarkdown }],
}));

Performance

Module-Level Caching

The McpServer must be per-request, but everything else can be shared:

// Module-level (created once)
const SCHEMAS = {
  search: z.object({ query: z.string().describe("Search query") }),
  fetch: z.object({ id: z.string().describe("Resource ID") }),
};
const READ_ONLY_ANNOTATIONS = {
  readOnlyHint: true, destructiveHint: false, idempotentHint: true, openWorldHint: true,
} as const;

// Per-request (created each time)
function createMcpServer(ctx: Context) {
  const server = new McpServer({ name: "my-server", version: "1.0.0" });
  server.tool("search", "Search", SCHEMAS.search, READ_ONLY_ANNOTATIONS, handler);
  return server;
}

Token Bloat Mitigation

Tool definitions consume context window before any conversation starts. GitHub MCP: 20,444 tokens for 80 tools (SEP-1576).

Strategies:

1. 5-15 tools per server - community sweet spot. Split beyond that.
2. Outcome-oriented tools - bundle multi-step operations into single tools (e.g., track_order(email) not get_user + list_orders + get_status).
3. Response granularity - return curated results, not raw API dumps. 800-token user object vs 20-token summary.
4. outputSchema + structuredContent - lets clients process data programmatically without LLM parsing overhead.
5. Dynamic tool loading - register only relevant tool subsets based on request context (e.g., ?tools=search,fetch query parameter).

No-Parameter Tools

For tools with no inputs, use explicit empty schema:

inputSchema: { type: "object" as const, additionalProperties: false }

Security

Top Threats (real-world incidents, 2025-2026)

Server-Side Requirements (spec normative)

• Validate all inputs at tool boundaries
• Implement access controls per user/session
• Rate limit tool invocations
• Sanitize outputs before returning to client
• Validate Origin header - respond 403 for invalid origins (2025-11-25 requirement)
• Require MCP-Protocol-Version header on all requests after initialization (spec 2025-06-18+)
• Bind local servers to localhost (127.0.0.1) only

Auth (OAuth 2.1)

MCP normatively requires OAuth 2.1 (draft-ietf-oauth-v2-1-13). The spec states: "Authorization servers MUST implement OAuth 2.1." PKCE is mandatory, implicit flow is removed. Always build against OAuth 2.1 - not 2.0.

MCP servers are OAuth 2.1 Resource Servers. Clients MUST include Resource Indicators (RFC 8707) binding tokens to specific servers. Key requirements:

• Validate audience - reject tokens not issued for your server (token passthrough is explicitly forbidden)
• PKCE mandatory - use S256 code challenge method
• Short-lived tokens - reduce blast radius of leaked credentials
• Scope minimization - start with minimal scopes, elevate incrementally via WWW-Authenticate challenges
• Don't implement token validation yourself - use tested libraries (Keycloak, Auth0, etc.)
• Don't log credentials - never log Authorization headers, tokens, or secrets

For full security attack/mitigation patterns and auth implementation details: see references/security-auth.md

Known SDK Bugs

V2 Migration

For comprehensive migration guide with all breaking changes and before/after code: see references/v2-migration.md

Key breaking changes:

1. Package split: @modelcontextprotocol/sdk -> @modelcontextprotocol/server + /client + /core
2. ESM only, Node.js 20+
3. Zod v4 required (or any Standard Schema library)
4. McpError -> ProtocolError (from @modelcontextprotocol/core)
5. extra parameter -> structured ctx with ctx.mcpReq
6. server.tool() -> registerTool() (config object, not positional args)
7. SSE server transport removed (clients can still connect to legacy SSE servers)
8. @modelcontextprotocol/hono and @modelcontextprotocol/express middleware packages
9. DNS rebinding protection enabled by default for localhost servers

v1.x gets 6 more months of support after v2 stable ships. No rush, but write new code with v2 patterns in mind.

Extensions

MCP extensions are optional, strictly additive capabilities on top of the core protocol. Both sides negotiate support during initialization via extensions in capabilities.

Identifiers: {vendor-prefix}/{extension-name}. Official: io.modelcontextprotocol/*. Third-party: reversed domain (e.g., com.example/my-ext).

Official Extensions

Client support: Claude (web + Desktop), ChatGPT, VS Code Copilot, Goose, Postman, MCPJam all support MCP Apps. Auth extensions not yet widely adopted.

For MCP Apps architecture, ext-apps SDK, and build patterns: see references/mcp-apps.md For extensions system, auth extensions, and MCP Registry: see references/extensions-registry.md

Server Capabilities Beyond Tools