Server has 6 tools with basic schemas and descriptions in Korean. All tools have input schemas with type definitions and parameter descriptions. However, descriptions are minimal (10-50 chars), lack actionable context for LLM selection, and do not explain when to use each tool vs. alternatives. No output schemas documented. The destructive tool (buy_lotto) lacks confirmation/dry-run pattern. Parameter descriptions are terse and do not specify constraints (e.g., count range 1-5 is stated but not in schema). No error handling guidance visible in tool definitions. Tool naming is clear (verb_noun pattern), but composition could be improved, check_numbers and get_winning_numbers overlap in intent.

• Expand all tool descriptions to 100-200 chars. Include: what the tool does, when to use it (vs. similar tools), and what it returns. Example: 'Retrieve the winning numbers for a specific lottery round. Call this before check_numbers to verify your ticket. Returns 6 main numbers and 1 bonus number.'
• Add output schema documentation for each tool. Specify field names, types, and what each field represents. Example for get_winning_numbers: {numbers: [int], bonus: int, round: int, date: string}.
• Implement confirmation/dry-run for buy_lotto. Add a 'dry_run' boolean parameter (default false) that returns estimated cost without charging. Require explicit confirmation before real purchase.
• Add enum constraints to schema for mode parameter: {type: 'string', enum: ['auto', 'manual', 'semi']}. Add minItems/maxItems to numbers array and min/max to count parameter.
• Add error handling guidance to buy_lotto description: 'Returns error if balance insufficient, count out of range, or numbers invalid. Errors are retryable (network) or user-fixable (balance, format).'