Three well-named tools with clear, descriptive text (avg 180 chars). All have proper JSON Schema with typed parameters and descriptions. However, output schemas are not documented, LLMs cannot predict response structure for planning. Error handling returns structured objects but lacks recovery guidance. No pagination despite tools returning up to 50-100 results. Parameter descriptions are good but lack format constraints (e.g., group enum values not formalized in schema).

• Document output schema for each tool. Example for get_satellite: {type: 'object', properties: {name: {type: 'string'}, norad_id: {type: 'number'}, epoch: {type: 'string'}, inclination_deg: {type: 'number'}, ...}}
• Convert group parameter to enum: {type: 'string', enum: ['stations', 'starlink', 'gps-ops', 'weather', 'science', 'geo', 'active']}
• Add pagination to search_by_name and get_group: limit (1-100, default 20) and offset (default 0) parameters; return {count, total, satellites, has_more}
• Enhance error messages with recovery hints: 'Satellite NORAD ID 99999 not found. Try search_by_name() to find similar objects.' Distinguish 'not_found' from 'api_error' from 'invalid_input'.
• Remove example values from descriptions; rely on schema constraints (enum, pattern, minLength) to guide LLM input.