mcp-builder · dkyazzentwatwa — Claude Code skill: design MCP servers
mcp-builder — a skill for planning and building MCP servers with tools, schemas, error handling, and evaluation. Use when creating or refactoring MCP integrations.
Apache-2.0 198 installs ★ 136 updated 2026-04-08
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a claude-codeWhat mcp-builder does
This skill teaches agents to structure MCP servers around user workflows rather than raw API endpoints. The approach starts by reading target API documentation and identifying end-to-end scenarios the agent must complete, then designs a focused tool surface with clear identifiers and corrective error messages.
The skill includes design principles for every server, implementation patterns for Python and TypeScript, and helper scripts for testing connections and evaluating server behavior. Infrastructure comes first: authentication, request helpers, pagination, data truncation, and output formatting.
Output is concise and high-signal by default. Human-readable identifiers replace opaque IDs. Error messages tell the agent what to try next rather than just reporting failure. The whole approach assumes limited context and large datasets.
What it helps with
- Design MCP server tools tied to agent workflows
- Write tool schemas and docstrings that guide correct automation usage
- Turn API errors into corrective messages agents can act on
- Evaluate an MCP server against realistic tasks before launch
- Build shared infrastructure: auth, request helpers, pagination, formatting
Example prompts for the agent
- Design an MCP server for Stripe: identify workflows and build a tool inventory
- Write schemas for S3 file tools with clear error messages and retry guidance
- Evaluate the MCP server: agent creates a payment and sends a notification
How to install mcp-builder
1Skills CLI (Vercel)
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a claude-code2Manually: put the files in place
~/.claude/skills/mcp-builder/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/mcp-builder/
1Skills CLI (Vercel)
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a codex2Manually: put the files in place
.agents/skills/mcp-builder/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a cursor2Manually: put the files in place
.cursor/skills/mcp-builder/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install dkyazzentwatwa/chatgpt-skills mcp-builder2Manually: put the files in place
.github/skills/mcp-builder/SKILL.md1Skills CLI (Vercel)
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a gemini-cli2Platform CLI
gemini skills install https://github.com/dkyazzentwatwa/chatgpt-skills --path mcp-builderManually: put the files in place
.devin/skills/mcp-builder/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about mcp-builder
How do I install mcp-builder in Claude Code?
Run npx skills add dkyazzentwatwa/chatgpt-skills --skill mcp-builder -a claude-code, then open ~/.claude/skills/mcp-builder/SKILL.md.
What does mcp-builder do?
It guides agent-friendly MCP server design around workflows, not endpoints. Includes templates for tool schemas, error handling, shared auth and request infrastructure, and evaluation prompts.
What languages does the skill cover?
Python and TypeScript, with separate guides for implementation patterns in each.
Source text
SKILL.md · Apache-2.0
--- name: mcp-builder description: Plan and build MCP servers with agent-friendly tools, schemas, error handling, and evaluation. Use when creating or refactoring MCP integrations. --- # MCP Builder Build MCP servers around user workflows, not raw API endpoints. ## Workflow 1. Read the target API docs and identify the workflows an agent must complete end to end. 2. Design a small tool surface with high-signal outputs, clear identifiers, and actionable errors. 3. Implement shared infrastructure first: auth, request helpers, pagination, truncation, and formatting. 4. Add tool schemas and docstrings that make correct usage obvious. 5. Evaluate the server with realistic tasks before expanding scope. ## Principles - Prefer workflow tools over thin endpoint wrappers. - Return concise, high-signal responses by default. - Use human-readable identifiers whenever possible. - Make error messages corrective: tell the agent what to try next. - Design for limited context and large datasets. ## Resources - `references/mcp_best_practices.md` for design principles that apply to every server. - `references/python_mcp_server.md` for Python implementation patterns. - `references/node_mcp_server.md` for TypeScript implementation patterns. - `scripts/connections.py` and `scripts/evaluation.py` as repo-local helpers. ## Deliverables - A concrete tool inventory tied to user workflows. - Strict input/output schemas. - Evaluation prompts or scripts that confirm the server is usable by an agent.
© dkyazzentwatwa · Apache-2.0 · dkyazzentwatwa/chatgpt-skills
Similar skills
All skills for Agent setup: memory, context, orchestration →mcp-builder by other authors comes first.