Three simple tools with basic descriptions and schemas. All tools have explicit registrations with Zod schemas and descriptions visible in source. However, descriptions are minimal (10-20 chars baseline), parameter descriptions lack depth, and no output schemas are documented. No error handling guidance, no parameter constraints (enums, ranges), and no composition patterns. Naming is acceptable but generic. This is a starter template, not production-grade.

• Expand tool descriptions to 100-200 chars. Example: 'echo' → 'Echo back the provided message to verify the MCP server is reachable and responding correctly. Useful for testing connectivity and latency.'
• Document output schemas explicitly. Add a comment or JSDoc block showing the response structure: { content: [{ type: 'text', text: string }] }.
• Add parameter constraints: max length for 'message' and 'text' (e.g., 10000 chars), and describe expected behavior for edge cases (empty strings, very long input).
• Add error handling guidance: 'If the message exceeds 10000 characters, the tool will truncate and return a warning.' This helps LLMs understand failure modes.
• Consider adding a 'limit' parameter to word_count to cap analysis on very large texts, preventing token waste on massive inputs.