Two tools with basic verb-noun naming (get_alerts, get_forecast). Both have descriptions (50-60 chars, below the 194-char baseline) and visible input schemas with typed parameters. However, descriptions lack context on WHEN to use each tool, prerequisites, or what the output contains. No output schema documentation. Parameters have descriptions but lack constraints (e.g., state code format, lat/lon bounds). Error handling returns plain strings with no recovery guidance. No tool annotations (readOnlyHint, etc.). Overall: functional but minimal, typical of introductory MCP servers.

• Add output schema documentation. Define return type as structured object: {alerts: [{event, area, severity, description, instructions}]} or {forecast: [{name, temperature, windSpeed, windDirection, detailedForecast}]}. This enables LLMs to extract fields and chain calls.
• Expand tool descriptions to 100-150 chars. Example: 'Get active weather alerts for a US state. Use this to check for severe weather warnings, tornado watches, or flood alerts. Returns event type, affected area, severity level, and recommended actions.'
• Add enum constraint to 'state' parameter. Define valid values as 2-letter USPS codes (AL, AK, AZ, ..., WY) or accept both codes and state names with server-side normalization.
• Add bounds to latitude (-90 to 90) and longitude (-180 to 180) parameters. Document in descriptions: 'Latitude must be between -90 and 90 degrees.'
• Enhance error messages with recovery guidance. Example: 'Invalid state code. Valid codes are 2-letter USPS abbreviations (CA, NY, TX, etc.). Did you mean: CA?' This enables LLM self-correction.