Loading ToolBench
Turn any idea, blog post, article, or X post into hand-drawn "Tinku" illustrations that explain the key concepts — an MCP App with an interactive viewer for your AI host; bring-your-own image API key (OpenAI/Gemini).
Doodleworks MCP has well-structured tool definitions with comprehensive schemas and clear descriptions. All 4 tools have detailed input schemas with proper types and descriptions. However, output schemas are not explicitly documented in the visible code, and error handling guidance is minimal. Tool names follow verb_noun convention (create_, get_, regenerate_, plan_). Descriptions are substantive (100-250 chars) and explain WHEN to use each tool. Parameters are well-typed with enums for constrained values (spine, aspect, resolution, quality). The main gaps are: (1) no visible output schema documentation for any tool, (2) no error recovery guidance in descriptions, (3) no per-parameter validation rules stated in descriptions (e.g., topic length limits, count range), (4) missing idempotency/retry guidance for stateful operations.
Create a set of hand-drawn Tinku illustrations from a topic, article, or concept. Returns illustration metadata and a setId; the viewer fetches each rendered image via get_illustration.
Fetch a single rendered illustration by setId and index. Returns the illustration with dataUri once rendering is complete, or a status indicating pending/error.
Plan an illustration set from a topic: scope the set, identify cognitive anchors, and generate illustration prompts. Returns a structured plan ready for create_illustrations.
Re-render a single illustration with optional prompt/style overrides. Useful for iterating on a specific illustration without regenerating the entire set.
Output schemas not documented. Tool descriptions explain inputs but do not specify what fields/structure the LLM should expect in responses. This forces LLMs to guess at response structure and risks incorrect downstream tool chaining.
No error recovery guidance in tool descriptions. Descriptions do not explain what errors can occur, how to recover, or what to try next. E.g., create_illustrations may fail if API quota is exceeded, but the description offers no recovery hint.
Parameter constraints not stated in descriptions. E.g., 'topic' parameter has no stated length limit, 'count' has no min/max range, 'illustrations' array has no size limit. LLMs will pass unbounded values without guidance.
Inferred effective spec: 2025-06-18+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | B | 71 | 2025-06-18+ | v2 |
Idempotency not documented. create_illustrations and regenerate_illustration are stateful operations that modify server state. Descriptions do not clarify whether repeated calls with identical inputs produce the same setId or new ones, affecting retry safety.
No confirmation/dry-run pattern for destructive operations. regenerate_illustration overwrites cached illustrations without a preview or confirmation step. Agents cannot safely iterate without risking loss of prior work.