Two tools with basic verb-noun naming (get_alerts, get_forecast). Both have short descriptions (34-40 chars, below the 50-200 char LLM-optimized baseline). Input schemas are present with type definitions and parameter descriptions, meeting minimum requirements. However, output schemas are completely undocumented, tools return free-text strings with no structured field definitions, forcing LLMs to parse unstructured text. No error handling guidance, no pagination support, no output field documentation. No tool annotations (readOnlyHint, etc.). Descriptions lack WHEN/WHY context and dependency hints. Parameter descriptions are minimal but present.

• Document output schemas explicitly. Define structured return types: get_alerts should return {alerts: [{event, area, severity, description, instructions}], count, timestamp}, not a free-text string.
• Expand descriptions to 80-150 chars with WHEN/WHY context. Example: 'Get active weather alerts for a US state. Call this to check for severe weather warnings before planning outdoor activities. Returns structured alert list with severity levels.'
• Add error handling guidance. Replace generic 'Unable to fetch' with actionable messages: 'Invalid state code. Use two-letter US state codes (CA, NY, TX). Or try search_locations() to find the correct code.'
• Add tool annotations. Mark both tools with readOnlyHint=true since they only fetch data. This signals to LLMs they are safe to retry.
• Implement pagination for get_forecast. Accept optional limit (1-10, default 5) and offset parameters. Return {periods: [...], total_periods, next_offset} to enable multi-step forecast retrieval.