while comprehensive coverage gives agents flexibility to compose operations. Performance varies by client—some clients benefit from code execution that combines basic tools

Mar 20, 2026

Find agent skills by outcome

The name should be general

- Examples: `slack_mcp`

- Examples: `slack-mcp-server`

- Return `has_more`

- **stdio**: For local integrations

- **Streamable HTTP**: For remote servers

- Example: `slack_send_message`

- **Node/TypeScript**: `{service}-mcp-server` (e.g.

- **Python**: `{service}_mcp` (e.g.

- **Verifiable**: Single

3. **Question Generation**: Create 10 complex

Use evaluations to test whether LLMs can effectively use your MCP server to answer realistic

To create effective evaluations

After implementing your MCP server

- [⚡ TypeScript Guide](./reference/node_mcp_server.md) - Project structure

Prioritize comprehensive API coverage. List endpoints to implement

- [🐍 Python Guide](./reference/python_mcp_server.md) - Module organization

- **Language**: TypeScript (high-quality SDK support and good compatibility in many execution environments e.g. MCPB. Plus AI models are good at generating TypeScript code

Review the service's API documentation to identify key endpoints

- **Transport**: Streamable HTTP for remote servers

- Tool

Then fetch specific pages with `.md` suffix for markdown format (e.g.

- Transport mechanisms (streamable HTTP

Balance comprehensive API endpoint coverage with specialized workflow tools. Workflow tools can be more convenient for specific tasks

- stdio: For local integrations

- Streamable HTTP: For remote servers

- Node/TypeScript: `{service}-mcp-server` (e.g.

- Python: `{service}_mcp` (e.g.

- Verifiable: Single

3. Question Generation: Create 10 complex

- Language: TypeScript (high-quality SDK support and good compatibility in many execution environments e.g. MCPB. Plus AI models are good at generating TypeScript code

- Transport: Streamable HTTP for remote servers