K-Mail-MCP has 7 tools with complete input schemas and descriptions, but quality is inconsistent. Tool names follow verb_noun convention (check_, list_, read_, reset_, set_). Descriptions are present but vary in clarity, some are verbose (check_new_mails: 180+ chars with implementation details) while others are minimal (list_accounts: 60 chars). All parameters have type and description fields. However, output schemas are not documented anywhere in the source, LLMs cannot predict what fields to expect from responses. Error handling is absent: no guidance on retryability, user-fixable vs fatal errors, or recovery paths. Security sanitization for prompt injection is implemented but not reflected in tool descriptions. STDIO transport caps protocol readiness at 50, limiting overall score despite reasonable definition structure.

• Document output schemas for all tools. E.g., check_new_mails returns {emails: [{id, from, subject, date, preview, spam_score}], total_count, next_cursor}. Include field types and descriptions.
• Add error handling guidance: 'If account_label not found, call list_accounts() to see available accounts. If IMAP connection fails, retry after 30s (retryable).'
• Refactor check_new_mails description: 'Fetch new emails from a mail account. Returns up to [limit] emails with spam filtering applied. Call list_accounts() first to get valid account labels.' (Keep it under 150 chars.)
• Add confirmation step for destructive operations: reset_last_run should return {action: 'reset_last_run', account_label, confirm_token} and require a confirm_reset_last_run(confirm_token) call to execute.
• Document parameter constraints in descriptions: 'limit: number between 1 and 100 (default 10)' instead of just 'Maximum number of emails to return'.

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

• Relies on Dynamic Client Registration (deprecated) - use Client ID Metadata Documents (CIMD)