Three tools with complete input schemas and descriptions, but significant gaps in output documentation, error guidance, and parameter constraints. Tool names follow verb_noun pattern (create_, get_, download_). Descriptions are present (60-90 chars) but lack actionable context for LLM selection. No enum constraints on task_type despite known valid values. Error handling returns structured responses but lacks recovery guidance. Output schemas are not documented, LLMs cannot plan downstream calls. No tool annotations (readOnlyHint/destructiveHint). Baseline: 549 production tools average 194 chars for descriptions; these average ~75 chars.

• Add enum constraint to task_type parameter: {"enum": ["text_to_model", "image_to_model", "multiview_to_model"]} to prevent LLM hallucination.
• Document output schemas for all three tools. E.g., create_tripo3d_task returns {task_id: string, response: object}; get_tripo3d_task returns {status: string, response: object}; download_tripo3d_task returns {files: [string], response: object}.
• Expand tool descriptions to 150-250 chars. Include WHEN to use and dependencies. E.g., 'Fetch the status and details of a Tripo3D task. Call this after create_tripo3d_task to poll for completion. Returns status (pending|processing|success|failed) and output URLs.'
• Add recovery guidance to error responses. Replace 'prompt is required' with 'prompt is required for text_to_model tasks. Provide a text description of the 3D model you want to generate.'
• Add tool annotations: create_tripo3d_task and download_tripo3d_task with destructiveHint=true; get_tripo3d_task with readOnlyHint=true. This signals to LLMs which tools have side effects.