Server has 9 tools with complete input schemas and descriptions. Naming follows verb_noun convention (send_text, find_chats, create_instance). Descriptions are present but brief (avg ~60 chars, below the 194-char baseline). Parameters are typed with Zod and mostly described, but lack detail on constraints, formats, and error recovery. Output schemas are not documented, responses are JSON-stringified without field-level guidance. Error handling returns raw messages but lacks actionable recovery steps or categorization. No tool annotations (readOnlyHint/destructiveHint) despite clear risk levels. Missing pagination guidance for list tools.

• Document output schemas for all tools. E.g., send_text returns {success: bool, messageId: string, timestamp: ISO8601}. Include field descriptions so LLMs know what to extract.
• Expand tool descriptions to 100 - 200 chars. Add WHEN to use, prerequisites, and return summary. E.g., 'Send a WhatsApp text message. Use after finding a chat. Returns message ID and timestamp for tracking.'
• Wrap error responses with structured metadata: {error: string, category: 'retryable'|'user_fixable'|'fatal', suggestion: string}. E.g., 'Instance not found. Try create_instance() first.'
• Add tool annotations in registerTool() calls. Mark send_text/send_media with destructiveHint: true; mark find_chats/find_messages with readOnlyHint: true. Use MCP SDK's annotation support.
• Add pagination to find_chats: accept optional limit (default 20, max 100) and return total_count. Document in description: 'Returns up to 20 chats; use limit to adjust.'