agentsclimarketplace
Sign in

Mcp builder

Skill aizech/bernhard-zechmann-skills/skills/engineering/mcp-builder

Create an MCP server that exposes external APIs or services as well-designed tools. Use when building a Model Context Protocol server in Python or TypeScript.From its SKILL.md

Install
npx -y skills add aizech/bernhard-zechmann-skills --skill mcp-builder

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

One thing to look at

  • 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.

What its file declares

Copied from the file, not written here

The file declares its own license as MIT. That is the author’s claim about this one file, and it is not the same thing as the license GitHub reports for the repository, which is listed with the other numbers below.

SKILL.md

1.9 KB, 391 tokens by cl100k_base, as published. Nobody here has run it

MCP Builder

Create MCP servers that let agents interact with external services through clean tools.

Process

1. Research

  • Review the MCP specification at https://modelcontextprotocol.io/sitemap.xml.
  • Fetch the TypeScript or Python SDK README from GitHub.
  • Read the target API docs to identify endpoints, auth, pagination, and data models.

2. Plan

  • List the tools to implement. Prefer comprehensive API coverage over one-off workflow tools.
  • Use consistent, action-oriented names with prefixes, e.g. github_create_issue, github_list_repos.
  • Decide transport: streamable HTTP for remote, stdio for local.

3. Implement

TypeScript (default)

  • Use the MCP SDK with Zod for input schemas.
  • Define outputSchema and structuredContent where supported.
  • Annotate tools with readOnlyHint, destructiveHint, idempotentHint, openWorldHint.

Python

  • Use FastMCP with Pydantic models.
  • Add @mcp.tool() decorators.

Shared rules:

  • Async I/O for network calls.
  • Actionable error messages.
  • Pagination on list operations.
  • Return both text and structured data when possible.

4. Test

  • Build and run the server.
  • Test with the MCP Inspector:
npx @modelcontextprotocol/inspector

5. Evaluate

Create 10 read-only, complex, realistic questions that require multiple tool calls. Verify answers yourself. Store as evals.xml or evals/evals.json.

Reference

  • references/mcp-best-practices.md — server and tool conventions
  • references/node-mcp-server.md — TypeScript patterns
  • references/python-mcp-server.md — Python patterns
  • references/evaluation.md — evaluation guidelines

What ships with it: 10 files

111.3 KB alongside SKILL.md, 2 of them executable

evals/

references/

scripts/

Keep looking

Skills are one crate of 326,764. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.