A collection of Python scripts for working with draw.io diagrams, including code analysis, architecture modeling, AsyncAPI visualization, and diagram manipulation
This MCP server has severe quality gaps across naming, descriptions, schema clarity, and output documentation. Tool names are inconsistent and vague (e.g., 'aiicons' is not verb-based; 'asyncapiimports' combines read + import). Parameter descriptions exist but are minimal and lack detail about formats, constraints, or usage context. Input schemas are present but lack clarity on what the tools actually return. Most critically, no output schemas are documented, LLMs cannot plan downstream calls or extract the right data. The server includes 5 tools with write capabilities (autolayout, buildup, c4) but no error handling, confirmation steps, or recovery guidance documented. The example code provided shows internal usage but no tool invocations or real MCP protocol bindings, making verification difficult.
Find AI / LLM brand logos (OpenAI, Claude, Gemini, ...) as draw.io styles from lobe-icons set on unpkg CDN, with optional embedding as data URIs
Turn an AsyncAPI 2/3 spec into an event-driven architecture graph with channel, publish/subscribe operation, and message-payload schema nodes
Auto-layout a logical graph into draw.io XML using Graphviz dot for positioning nodes and emitting a .drawio file with mxGeometry coordinates
Animate a static .drawio building itself node by node into an interactive HTML player with base64-embedded PNG frames, play/pause/step/scrub controls
Generate C4 model diagrams from JSON specification into one multi-page .drawio file with drill-down links, official C4 shapes, colors, and Graphviz-based placement
No output schemas documented for any tool. LLMs cannot determine what fields to extract, plan downstream tool calls, or validate results.
Write-capable tools (autolayout, buildup, c4) lack error handling, dry-run/confirmation modes, and recovery guidance. No actionable error messages if file I/O fails or input is malformed.
Tool names are not verb-based and mix actions or use vague nouns. 'aiicons', 'asyncapiimports', 'autolayout', 'buildup', 'c4' do not clearly convey what the tool does.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 50 | 2026-07-28+ | v2 |
Input parameter descriptions are minimal and lack format constraints, ranges, or examples. E.g., 'hold' (float) has no range; 'fps' depends on 'hold' but undocumented; graph_json structure is not specified.
asyncapiimports and c4 accept file paths as input (spec, input) but do not document the file format (JSON? YAML? Both?) or link to schema/spec documentation.