Server has 15 tools with mixed quality. Naming is generally verb-based (getMarketMetrics, placeOrder, submitAuctionBid) but some are overly generic (runSimulation, runEnhancedSimulation). Descriptions exist for all tools but many are brief (10-50 chars) and lack actionable context. Input schemas are present but incomplete: many parameters lack type definitions (e.g., 'req' in runSimulation/runEnhancedSimulation is typed as 'object' with no nested schema), and several tools have empty input schemas (getMarketMetrics, registerUser, getAuctionStatus, getSimulationStrategies). Output schemas are not documented. Error handling is minimal, no recovery guidance or error classification visible in tool definitions. Security concerns: placeOrder and submitAuctionBid accept userId as string without validation hints. No pagination support visible despite tools returning potentially large datasets (getOrderBook, getUserOrders, getAuctionResults).

• Add complete nested JSON Schema definitions for 'req' parameters in runSimulation and runEnhancedSimulation. Document all fields (merchantCount, steps, weatherEnabled, executeReal, noiseLevel, etc.) with types and descriptions.
• Expand tool descriptions to 50-200 characters. Include WHEN to call (e.g., 'Call this after placing an order to check if it was filled') and WHAT to do with the result (e.g., 'Returns clearing price and per-user allocations; use to reconcile positions').
• Document output schemas for all tools. Specify field names, types, and whether results are paginated. E.g., getOrderBook should return {orders: [{id, price, qty, side}], total: number, depth: number}.
• Add pagination parameters (limit, offset/cursor) to tools returning lists (getOrderBook, getUserOrders, getAuctionResults).
• Rename runSimulation → simulate_simple_trading and runEnhancedSimulation → simulate_multi_merchant_market to clarify intent and reduce LLM disambiguation overhead.