Telegram MCP has 12 tools with complete input schemas and descriptions. Naming follows verb_noun convention (list_, get_, send_, etc.). Descriptions are present but brief (avg ~60 chars, below the 194-char baseline). Parameters are typed and mostly described, but lack constraints (enums, ranges, patterns). Output schemas are partially documented via _format_message helper but not formally declared. Error handling is minimal, no recovery guidance or actionable error messages. Missing tool annotations (readOnlyHint, destructiveHint) despite clear risk levels. No pagination support for list_dialogs or get_messages despite potential for large result sets.

• Add tool annotations: @mcp.tool(readOnlyHint=True) for list_dialogs, get_messages, search_messages, get_chat_info; @mcp.tool(destructiveHint=True) for delete_messages; @mcp.tool(idempotentHint=True) for send_message (if idempotent via message dedup).
• Expand descriptions to 100-150 chars. Example: 'List your Telegram chats, channels, and groups. Use this to discover available conversations before fetching messages or sending replies.' Include when to use vs similar tools.
• Add pagination: include offset/limit in list_dialogs and get_messages; return total_count or next_offset in responses. Cap results at 50 by default.
• Document output schemas in docstrings. Example: 'Returns list of dicts with keys: id (int), text (str), date (ISO 8601), sender_id (int), sender_name (str), has_media (bool), media_type (str).'
• Add error handling with recovery hints. Wrap API calls in try/except; return structured errors like '{"error": "Chat not found", "suggestion": "Try search_messages() or list_dialogs() to find the chat ID."}'.

Inferred effective spec: <=2025-11-25.

• Relies on Logging (deprecated) - log to stderr or use OpenTelemetry