mcp-builder

Plan, build, review, and evaluate MCP servers with Python FastMCP or TypeScript SDK.

Updated Apr 16, 2026

Search Tags:#workflow #testing #sdk #mcp #server #tools

Ready to run

npx skills add https://github.com/ejoliet/claude-skills --skill mcp-builder-ejoliet

💬 Prompt for ChatGPT / Claude Chat▼

Please help me install this Agent Skill.
Skill: mcp-builder
Source: https://github.com/ejoliet/claude-skills/tree/main/mcp-builder
Command: npx skills add https://github.com/ejoliet/claude-skills --skill mcp-builder-ejoliet

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Guides engineers to plan, implement, review, and evaluate MCP servers using FastMCP (Python) or the MCP TypeScript SDK, enabling consistent architecture and tooling across projects.

Core Features & Use Cases

Four-phase MCP development methodology (research & planning, implementation, review & refine, evaluation)
Cross-language server support with Python FastMCP and TypeScript SDK
Infrastructure-first design: shared utilities, authentication, error handling, and pagination
Tool schema design and MCP annotations to drive safe, structured tool workflows
Evaluation planning with realistic, repeatable questions to validate end-to-end workflows

Quick Start

Bootstrapping an MCP server using the four-phase workflow, choose Python or TypeScript SDK, scaffold tools, and begin iterative testing.

Frequently Asked Questions about mcp-builder

High-intent search queries and answers about installing and using this skill.

FAQPage Schema

How do I build an MCP server end-to-end from scratch?▼

To build an MCP server end-to-end, follow a repeatable four-phase workflow: research and planning, implementation, review and refine, and evaluation. This methodology scaffolds tools using either Python FastMCP or the TypeScript SDK to ensure consistent architecture.

What is the best way to structure MCP tool schemas and annotations?▼

The best way to structure MCP tool schemas is to define clear input/output schemas and apply explicit MCP annotations. This infrastructure-first design drives safe, structured tool workflows and guides tool behavior and error handling across projects.

Does the MCP TypeScript SDK work alongside Python FastMCP for server development?▼

Yes, the MCP TypeScript SDK and Python FastMCP are both supported for cross-language server development. The workflow provides consistent infrastructure utilities, authentication, and error handling across both SDKs to expand existing toolsets.

How do I evaluate MCP server workflows with realistic questions?▼

To evaluate MCP server workflows, apply the evaluation phase from the four-phase methodology. It involves planning realistic, repeatable questions that validate multi-tool agent backends and end-to-end MCP tooling behavior across contexts.

What infrastructure utilities do I need to set up before implementing an MCP server?▼

Before implementing an MCP server, you need shared infrastructure utilities for authentication, error handling, and pagination. Establishing this infrastructure-first design ensures consistent architecture and structured tool workflows during the building phase.

Can I use this MCP server workflow to expand an existing toolset?▼

Yes, you can use this four-phase MCP development workflow to expand existing toolsets. It supports adding new tools to current servers by applying consistent schema design, MCP annotations, and evaluation planning to validate the expanded multi-tool backend.