CtrlK
BlogDocsLog inGet started
Tessl Logo

mcp-builder

Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).

77

2.55x
Quality

66%

Does it follow best practices?

Impact

97%

2.55x

Average score across 3 eval scenarios

SecuritybySnyk

Advisory

Suggest reviewing before use

Optimize this skill with Tessl

npx tessl skill review --optimize ./plugins/all-skills/skills/mcp-builder/SKILL.md
SKILL.md
Quality
Evals
Security

Evaluation results

98%

81%

Build a Python MCP Server for the JSONPlaceholder API

Python MCP server implementation

Criteria
Without context
With context

Server naming convention

0%

100%

Tool naming with service prefix

0%

100%

FastMCP with @mcp.tool decorator

100%

100%

Tool annotations in decorator

0%

100%

Pydantic v2 model_config

0%

100%

CHARACTER_LIMIT constant

0%

100%

Pagination metadata returned

0%

100%

Dual response formats

0%

100%

Shared utility functions

60%

100%

Async/await for I/O

0%

100%

Actionable error messages

12%

100%

Stderr for logging

50%

50%

100%

62%

Build a TypeScript MCP Server for the DummyJSON API

TypeScript MCP server implementation

Criteria
Without context
With context

Server naming convention

100%

100%

McpServer import path

0%

100%

registerTool usage

0%

100%

Project directory structure

50%

100%

Zod schemas with .strict()

0%

100%

TypeScript strict mode

100%

100%

No 'any' types

100%

100%

Tool annotations set

0%

100%

CHARACTER_LIMIT constant

0%

100%

Pagination metadata returned

0%

100%

Tool names include service prefix

0%

100%

Build succeeds

100%

100%

94%

33%

Create an Evaluation Suite for an Open Library MCP Server

MCP server evaluation file creation

Criteria
Without context
With context

Correct XML structure

25%

100%

Exactly 10 Q&A pairs

0%

100%

Simple scalar answers

100%

100%

Questions require multi-step reasoning

100%

100%

Questions avoid trivial keyword search

0%

100%

Read-only question framing

100%

100%

Question independence

100%

100%

Answer format specification

50%

75%

Topic diversity

100%

50%

Stable answers

66%

100%

Repository
davepoon/buildwithclaude
Evaluated
Agent
Claude Code
Model
Claude Sonnet 4.6

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.