Server has 6 tools with mixed quality. send_notification has a comprehensive schema with 12 parameters and clear descriptions, scoring well. However, 4 of 6 tools (request_user_input, request_llm_completion, subscribe_to_topic, get_messages) declare features the server explicitly does NOT support (elicitation=false, sampling=false). These tools are non-functional stubs that mislead LLMs into selecting them. Descriptions are present but generic (50-100 chars). Parameter descriptions lack detail on constraints, formats, and valid ranges. No output schemas documented. Error handling is absent, no recovery guidance or actionable error messages visible in code.

• Remove or implement request_user_input and request_llm_completion. If elicitation/sampling are not supported, do not register these tools. If they will be supported, implement the handlers and set elicitation=true, sampling=true.
• Document output schemas for all tools. Example for send_notification: {"success": boolean, "message_id": string, "timestamp": string (ISO 8601), "topic": string}.
• Add format and constraint details to parameter descriptions. E.g., 'delay: Duration string (e.g., 5m, 1h, 30s); max 24h' and 'topic: Required. 1-64 alphanumeric + dash/underscore characters.'
• Add error handling guidance to send_notification description: 'Returns error if topic is invalid, auth fails, or rate limit exceeded. Retry with exponential backoff for rate limits.'
• Implement pagination for get_messages: add 'limit' (default 20, max 100) and 'offset' or 'cursor' parameters; return total_count and next_cursor in response.