Three tools with basic schemas and descriptions, but significant gaps in parameter documentation, output schema clarity, and error handling guidance. Tool names follow verb_noun convention (sample_, get_, download_), but descriptions lack context on when to use each tool and what downstream operations require. Parameters have types but lack constraints, ranges, and format specifications. Output schemas are not formally documented. No error recovery guidance. Inferred tool definitions from test file rather than explicit MCP registration.

• Document output schemas for all three tools. Specify that sample_objaverse_uids returns {uids: string[], count: int, license: object}, get_objaverse_annotations returns {annotations: object, count: int, license: object}, and download_objaverse_objects returns {cached_files: object, files: object, license: object}.
• Add parameter constraints: limit (1 - 10000, default 20), seed (optional, any integer for reproducibility), download_processes (1 - 16, default 1). Include these in parameter descriptions, not just schema.
• Enhance tool descriptions with dependency hints and use cases. E.g., 'Call this first to discover available UIDs before fetching annotations or downloading files. Returns a list of Objaverse object identifiers.'
• Replace generic error messages with actionable recovery guidance. E.g., 'Failed to load Objaverse module. Install via: pip install objaverse' or 'Invalid UID format. UIDs must be non-empty strings.'
• Locate and review the actual MCP server registration code. Ensure tools are registered with explicit schemas, descriptions, and input/output type declarations in the MCP server initialization, not inferred from test files.