Critical failures across all dimensions. Tool definitions exist only as names in a fixture file (smoke_list_tools.py) with no visible schema, parameter, or output documentation in the provided source. The single example tool (user_tool_example) has a minimal schema with one parameter but lacks comprehensive documentation. No error handling guidance, no output schemas, no parameter descriptions beyond names. Descriptions are generic placeholders (e.g., 'Analyze a dataset in the catalog'). This is a template/fixture server with incomplete tool definitions, tool implementations are not visible in the provided code.

• Provide complete input schemas for all 35 tools with type definitions (string, number, boolean, array, object), required/optional flags, and constraints (enums, min/max, patterns). Currently 34 tools have zero schema documentation.
• Expand tool descriptions from generic placeholders to 100-200 character LLM-optimized text. Include: what the tool does, when to use it vs similar tools, prerequisites, and expected output structure. Example: 'List all datasets in the catalog. Returns dataset ID, name, row count, and creation date. Use this to discover available data before calling catalog_analyze_dataset.'
• Document output schemas for all tools. Specify field names, types, and whether results are paginated. For list tools (catalog_list_datasets, modeling_list_projects), include limit, offset/cursor, and total_count fields.
• Add parameter descriptions to user_tool_example and all other tools. Each parameter needs a description explaining its purpose, valid values, and format. Example: 'argument1 (string, required): The dataset ID to analyze. Format: alphanumeric, 1-50 chars.'
• Implement error handling guidance in tool descriptions. For each tool, document common failure modes and recovery steps. Example: 'If dataset not found, call catalog_list_datasets() to discover available IDs. If API timeout, retry with exponential backoff.'