Find agent skills by outcome

The measure of quality of an MCP server is NOT how well or comprehensively the server implements tools

but how well these implementations (input/output schemas

- Answers must be single

verifiable values

Create 10 human-readable questions requiring ONLY READ-ONLY

INDEPENDENT

- Performance testing: Check behavior under load

timeouts

- Questions must be READ-ONLY

INDEPENDENT

This document provides guidance on creating comprehensive evaluations for MCP servers. Evaluations test whether LLMs can effectively use your MCP server to answer realistic

complex questions using only the tools provided.

- Security testing: Validate auth

input sanitization

- Provide helpful

specific error messages with suggested next steps

Writing & Content

return { content: [{ type: text

text: result }] };

Important: Annotations are hints

not security guarantees. Clients should not make security-critical decisions based solely on annotations.

Coding & Debugging

- Store API keys in environment variables

never in code

- Single-user

single-session scenarios

Best for: Remote servers

web services

- Simple setup

no network configuration needed

Best for: Local integrations

command-line tools

### Markdown Format (`response_format=markdown`

typically default)

Data

- Return pagination metadata: Include `has_more`

`next_offset`/`next_cursor`

- Use headers

lists

- Provide tool annotations (readOnlyHint

destructiveHint

Creative

3. Be action-oriented: Start with verbs (get

list

1. Use snake_case: `search_users`

`create_project`

The name should be general

descriptive of the service being integrated

- Examples: `slack-mcp-server`

`github-mcp-server`

- Examples: `slack_mcp`

`github_mcp`