Three well-named tools with clear, domain-specific descriptions (150 - 250 chars each). All tools have complete input schemas with proper JSON Schema structure and type declarations. soil_properties and soil_classification accept numeric lat/lon with explicit range constraints; list_soil_properties has minimal params. Descriptions explain WHAT each tool does and note the external API may be slow. However, output schemas are not documented, LLMs cannot predict response structure for downstream chaining. Error handling is basic (generic error objects). No parameter validation guidance beyond description text. Missing: enum constraints for 'properties' array items and 'depth' string (currently free-form with validation only at runtime).

• Document output schemas for all three tools. Example for soil_properties: {latitude: number, longitude: number, depth: string, count: number, properties: [{name: string, unit_measure: {target_units: string}, depths: [{label: string, values: {mean: number|null, Q0.5: number|null}}]}]}
• Convert 'properties' parameter to enum: {type: 'array', items: {type: 'string', enum: ['clay', 'sand', 'silt', 'phh2o', 'soc', 'nitrogen', 'cec', 'bdod', 'cfvo', 'ocd', 'wv0010', 'wv0033', 'wv1500']}}
• Convert 'depth' parameter to enum: {type: 'string', enum: ['0-5cm', '5-15cm', '15-30cm', '30-60cm', '60-100cm', '100-200cm'], default: '0-5cm'}
• Enhance error responses with recovery hints. E.g., 'Invalid property code: nitrogen. Valid codes: clay, sand, silt, phh2o, soc, nitrogen, cec, bdod, cfvo, ocd, wv0010, wv0033, wv1500. Call list_soil_properties() to see all available properties.'
• Add pagination to list_soil_properties: accept optional limit (1 - 100, default 20) and offset (default 0); return {total: number, limit: number, offset: number, properties: [...]}