$npx skillfedfor your agent

mcp-development

Learn to develop Model Context Protocol servers with typed tools, resource endpoints, and prompt templates. This skill covers server initialization, input validation with schemas, error handling, and transport configuration for both stdio and HTTP-based deployments. Includes patterns for composable tool design and security best practices like secret redaction.

mcp-development teaches you to build MCP servers with tools, resources, prompts, and transport configuration.

AI-generated summary based on this skill's SKILL.md

2,394 849 Apache-2.0updated by rohitg00

Decision gist · record as of 2026-05-12

mcp-development teaches you to build MCP servers with tools, resources, prompts, and transport configuration. Learn to develop Model Context Protocol servers with typed tools, resource endpoints, and prompt templates. This skill covers server initialization, input validation with schemas, error handling, and transport configuration for both stdio and HTTP-based deployments. Includes patterns for composable tool design and security best practices like secret redaction.

manual: git clone https://github.com/rohitg00/awesome-claude-code-toolkit → cp -r awesome-claude-code-toolkit/skills/mcp-development ~/.claude/skills/mcp-development
skills/mcp-development/SKILL.md · version 7042d795

Use it when

  • mcp-development explains that MCP resource endpoints expose data sources to clients through URI-based access patterns.
  • mcp-development covers two main transport types: stdio for local processes and SSE (Server-Sent Events) for HTTP.

Verify before relying

Read SKILL.md below before installing (1 file). Open directory: indexed for reading, not audited.

Same gist for agents: .md · .json

Install

rohitg00/awesome-claude-code-toolkit/mcp-development · repository language: JavaScript

Open directory. Skills are indexed for reading, not audited. Review a skill's body before installing it.

Frequently asked questions

AI-generated answers based on this skill's SKILL.md and metadata

How do I build an MCP server with tools?

mcp-development teaches you to build Model Context Protocol servers by initializing a server instance, defining typed tools with Zod schemas for input validation, and implementing tool handlers. Start by creating a server object, registering tools with clear descriptions and input schemas, then handle tool calls with proper error handling. The skill covers composable tool design patterns and demonstrates real examples like database tools and API integrations.

What are MCP resource endpoints and how do I design prompt templates?

mcp-development explains that MCP resource endpoints expose data sources to clients through URI-based access patterns, while prompt templates provide reusable instruction sets. Resources can serve files, database records, or API responses. Prompt templates combine static instructions with dynamic variables. The skill shows how to implement both, including best practices for organizing resources hierarchically and designing templates that compose well with tools for flexible client workflows.

How do I configure MCP server transport and client connections?

mcp-development covers two main transport types: stdio for local processes and SSE (Server-Sent Events) for HTTP. For stdio, configure the server to read JSON-RPC messages from stdin and write responses to stdout. For SSE, set up HTTP endpoints for client connections. The skill includes configuration examples for both transports, connection lifecycle management, and how to handle client initialization and capability negotiation during setup.

How do I validate MCP tool inputs and handle errors properly?

mcp-development teaches input validation using Zod schemas defined in each tool's input_schema field. Zod automatically validates incoming arguments against your schema before the tool handler runs. For error handling, implement try-catch blocks in tool handlers, return structured error responses with clear messages, and avoid exposing sensitive details. The skill emphasizes logging errors for debugging while keeping client-facing messages user-friendly and secure.

What MCP development best practices and anti-patterns should I know?

mcp-development highlights key practices: keep tools focused and composable, use descriptive names and documentation, validate all inputs with schemas, redact secrets from responses, and test error paths. Anti-patterns to avoid include overly broad tool permissions, missing input validation, exposing internal errors to clients, and poor error messages. The skill provides a development checklist covering security, reliability, and maintainability for production MCP servers.

Can you show me a tutorial for building MCP servers?

mcp-development provides step-by-step guidance: initialize your server with the MCP SDK, define tools with typed inputs using Zod, implement handlers with error handling, add resources and prompts if needed, configure your transport (stdio or SSE), and test end-to-end. The skill includes concrete examples like database tools, API wrappers, and file servers. Follow the development checklist to ensure security, proper validation, and client compatibility before deployment.

SKILL.md

Rendered from the published skill. Quoted content, verbatim.

MCP Development

MCP Server with Tools

```typescript import { McpServer } from "@modelcontextprotocol/sdk/server/mcp.js"; import { StdioServerTransport } from "@modelcontextprotocol/sdk/server/stdio.js"; import { z } from "zod";

const server = new McpServer({ name: "project-tools", version: "1.0.0", });

server.tool( "search_files", "Search for files matching a glob pattern in the project directory", { pattern: z.string().describe("Glob pattern (e.g., '*/.ts')"), directory: z.string().optional().describe("Base directory to search from"), }, async ({ pattern, directory }) => { const files = await glob(pattern, { cwd: directory ?? process.cwd() }); return { content: [ { type: "text", text: files.length > 0 ? files.join("\n") : No files found matching ${pattern}, }, ], }; } );

server.tool( "run_query", "Execute a read-only SQL query against the application database", { query: z.string().describe("SQL

(truncated - see the full file via the links below)

File tree — 1 file
skills/mcp-development/SKILL.md

Let your AI agent find skills like this

Example. Real query, live index.

You found this page by searching. An agent finds it by wishing: SkillFed indexes 56,283 agent skills by what they can do, searchable in plain language.

wish › “Learn how to develop and build MCP servers with tools”

Give your agent the search over MCP, or paste the wish link into any chat. No install? Search from any chat →

Related skills

Remote Mcp Server With Auth
by Aradotso · Aradotso/mcp-skills

A production-ready template for building remote MCP servers with GitHub OAuth authentication on Cloudflare Workers. It includes modular tool architecture, role-based access control, dual transport protocols (HTTP and SSE), PostgreSQL integration with SQL injection protection, and optional Sentry monitoring.

no license declared → metadata onlyupdated Jul 2026
★ 5repo stars
mcp-builder
by Pixel-Process-UG · Pixel-Process-UG/superkit-agents

Create MCP servers that expose capabilities to AI clients through a structured development process covering tool definition, resource management, prompt templates, and transport configuration. The skill guides you through design decisions, implementation patterns, security hardening, and testing to deploy servers ready for CLI and web environments.

MITupdated Mar 2026
★ 1repo stars
Mcp Builder
by TravisLeeeeee · TravisLeeeeee/awesome-openclaw-personas

Mcp Builder guides you through designing and implementing Model Context Protocol servers that agents can reliably use. Learn to create tool interfaces with clear descriptions and typed parameters, handle errors gracefully, expose resources and prompts, and test the full agent-to-tool loop.

no license declared → metadata onlyupdated Apr 2026
★ 105repo stars
mcp-developer
by Jeffallan · Jeffallan/claude-skills

mcp-developer is a specialist skill for constructing Model Context Protocol servers and clients that bridge AI systems with external tools and data. It guides you through project initialization, schema design with Zod or Pydantic, tool and resource implementation, transport configuration (stdio/HTTP/SSE), and protocol validation using the MCP inspector. Use it to scaffold complete implementations in TypeScript or Python, debug JSON-RPC compliance, and deploy production-ready integrations.

MITupdated May 2026
★ 10,759repo stars
mcp-builder
by jnMetaCode · jnMetaCode/superpowers-zh

mcp-builder teaches systematic design, implementation, testing, and deployment of Model Context Protocol servers. Learn to define tools for AI-driven operations, resources for read-only data access, and prompts for guided workflows, with patterns for input validation, error handling, security, and debugging across TypeScript and Python.

MITdocs in Chineseupdated Jul 2026
★ 7,277repo stars
mcp-server-builder
by patricio0312rev · patricio0312rev/skills

MCP Server Builder guides you through creating Model Context Protocol servers that extend Claude's capabilities with custom tools, resources, and prompts. The skill covers the full workflow from defining your server's purpose and choosing a transport layer through designing tool schemas, adding optional resources, and deploying to Claude Desktop or the API. It includes TypeScript project setup, basic server structure with request handlers, tool definition patterns with schema validation, and resource management examples.

MITupdated Jan 2026
★ 52repo stars

More skills mcp-builder (Apache-2.0) · mcp-builder (Apache-2.0) · Essential Mcp Servers Reference (NOASSERTION) · Mcp Security Hub (NOASSERTION)

Tags
protocol-implementationserver-architecturetool-compositionschema-validationresource-exposuretransport-layerprompt-engineeringerror-handling