Blender Model Context Protocol (MCP) Server - High Mode. Provides 64+ handlers for Blender automation including modeling, animation, rendering, AI tools, physics simulation, and external integrations.
Critical failures across naming, descriptions, and schema quality. Tool 1 (manage_ai_tools) violates single-responsibility principle with 30+ enum actions crammed into one tool, making it impossible for LLMs to reason about intent. Descriptions are generic and lack actionable guidance. Tool 2 (manage_tool_groups) has minimal description (67 chars) and vague parameter descriptions. Neither tool documents output schemas. No error handling guidance. No parameter validation rules stated. The 'action' enum pattern is an anti-pattern, each action should be a separate tool (e.g., smart_uv_unwrap, auto_retopology, etc.). Baseline: 90% of A+ tools have 50-200 char descriptions with clear WHAT/WHEN/WHY; these fail that bar.
Super-Tool for AI-powered mesh optimization, retopology, decimation, and intelligent workflow automation. Supports EXPLAIN, DRY_RUN, and CONFIDENCE queries.
Workflow-based tool grouping for easier discovery. Groups 64 handlers into logical workflows including Scene Setup, Modeling, Retopology, Materials, Character, Animation, Rendering, Game Export, Physics, and AI/External tools.
manage_ai_tools violates single-responsibility principle. 30+ actions (SMART_UV_UNWRAP, AUTO_RETOPOLOGY, SMART_DECIMATE, etc.) crammed into one tool with a single 'action' enum. LLMs cannot infer intent from a generic 'manage_ai_tools' name, they must parse a massive enum to find the right action. This is an anti-pattern. Each action should be a separate tool: smart_uv_unwrap, auto_retopology, smart_decimate, etc.
No output schemas documented for either tool. LLMs cannot plan downstream calls or extract required fields without knowing what the response contains. Baseline: 100% of A+ tools document return types.
manage_ai_tools description (67 chars) is too generic: 'Super-Tool for AI-powered mesh optimization...'. Does not explain WHEN to use it vs other tools, what it returns, or prerequisites. Baseline: 194 chars average, 50-200 chars for LLM-optimized descriptions.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 35 | 2026-07-28+ | v2 |
manage_tool_groups parameters lack descriptions. 'group_name' has no guidance on valid values or format. 'query' is undescribed. 'workflow_name' is undescribed. Baseline: 100% of A+ tools have descriptions for every parameter.
No error handling guidance. If an action fails (e.g., mesh is invalid for retopology), what does the tool return? How should the LLM recover? No recovery guides, error classification, or actionable error messages documented.