Server has 4 tools with clear verb-noun naming (list_chats, read_chat, search, send_message). All tools have descriptions (avg 95 chars, within 10-1024 range). Input schemas are present with type definitions and parameter descriptions. However, output schemas are undocumented, tools return formatted strings rather than structured objects, forcing LLMs to parse unstructured text. Error handling is minimal; no recovery guidance or actionable error messages. Missing tool annotations (readOnlyHint/destructiveHint) despite clear risk profiles. Parameter descriptions are adequate but lack format constraints (e.g., chat identifier format, message length limits). No pagination support despite list_chats accepting a limit parameter.

• Return structured JSON objects instead of formatted strings. E.g., list_chats should return {dialogs: [{id, name, kind, unread_count}], total: N} instead of a formatted string. This enables reliable field extraction and downstream tool chaining.
• Add tool annotations via FastMCP: @mcp.tool(annotations={readOnlyHint: true}) for list_chats/read_chat/search; @mcp.tool(annotations={destructiveHint: true}) for send_message. Agents need explicit safety signals.
• Implement structured error responses with recovery hints. E.g., if chat not found, return {error: 'chat_not_found', message: 'Chat "xyz" not found', suggestions: ['Try list_chats() to see available chats', 'Use @username or numeric id']}.
• Document parameter constraints in descriptions. E.g., 'chat: Chat identifier (name, @username, or numeric id; max 255 chars). Example: "General" or "@username" or "123456789".'
• Add pagination support: return {items: [...], total_count: N, has_more: bool} for list_chats/read_chat/search. Include next_offset or next_cursor if results are truncated.