Server has 6 tools with complete schemas and descriptions. Naming follows verb_noun convention (list_, search_, get_, mark_, send_). Descriptions are present but brief (avg 80 chars, below 194-char baseline). Parameter descriptions are mostly present but lack detail on constraints and formats. Output schemas are not documented. Error handling returns JSON with isError flag but lacks recovery guidance. No tool annotations (readOnlyHint/destructiveHint). Missing parameter descriptions for some fields (e.g., 'limit' in list_messages lacks description text, only has min/max). send_mail lacks required field validation guidance.

• Add tool annotations: readOnlyHint=true for list_*, search_*, get_*; destructiveHint=true for mark_message and send_mail. This enables agents to reason about side effects.
• Document output schemas for all tools. E.g., list_messages returns [{uid: number, subject: string, from: string, to: string, date: ISO8601, seen: boolean, flagged: boolean}]. Include chaining IDs (uid) needed by downstream tools.
• Enhance error messages with recovery guidance. E.g., 'Mailbox "INBOX" not found. Available mailboxes: [list from list_mailboxes()]. Try list_mailboxes() first.' or 'Invalid email format in to field. Expected: [email protected]'.
• Expand send_mail description to 150+ chars. State that 'to' is required, explain email format (comma-separated valid addresses), note that either 'text' or 'html' must be provided, and clarify what inReplyTo/references do (threading).
• Add parameter descriptions for numeric limits. E.g., 'limit: Maximum messages to return (1-200, default 20). Larger limits may timeout on slow connections.' This matches production baseline of 72 chars per param.