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 typed parameters and descriptions, but output schemas are completely undocumented, tools return plain strings with no structured format. No error handling guidance, no pagination support, no parameter constraints (e.g., state code validation). Naming is clear but descriptions lack context on when to use each tool or what the output contains.

• Expand tool descriptions to 100-150 chars. Include WHAT the tool does, WHEN to use it (e.g., 'Use this to check for active severe weather warnings'), and WHAT it returns (e.g., 'Returns a list of active alerts with event type, severity, and instructions').
• Define output schemas explicitly. Return structured JSON with typed fields: {alerts: [{event: string, area: string, severity: string, description: string, instructions: string}]} instead of plain text. Document this in the tool definition.
• Add enum constraint to state parameter in get_alerts. Validate against a list of valid US state codes (CA, NY, TX, etc.) or document the format requirement in the description.
• Implement pagination for get_alerts. Add limit (default 20, max 100) and offset parameters. Return {alerts: [...], total: number, has_more: boolean} to enable multi-page result handling.
• Enhance error messages with recovery hints. Instead of 'Unable to fetch alerts', return 'No alerts found for state code {state}. Verify the state code is a valid 2-letter US code (e.g., CA, NY).' or 'API timeout, try again in a few seconds.'