Smart-MCP
Allows generating complex draw.io diagrams (tables, kanbans, GANTT, shapes) via a local LLM with RAG, and provides a real-time dashboard for monitoring tests and results.
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Smart-MCPcreate a GANTT chart for project milestones"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Smart MCP Pipeline
đŻ Le dĂ©fi
Utiliser un petit modĂšle local (9B) sur un GPU grand public (RTX 3060 12Go) pour gĂ©nĂ©rer automatiquement des diagrammes draw.io complexes â tableaux, kanbans, GANTT, formes â une tĂąche oĂč mĂȘme de gros modĂšles cloud Ă©chouent.
Related MCP server: Excalidraw Architect MCP
đ§ L'approche
RAG (ChromaDB) : 26+ patterns de diagrammes stockés et réinjectés dans le prompt
Classification macro : FORME/TABLEAU/KANBAN/GANTT/AGENDA/TEXTE â spĂ©cialisation du prompt
ModĂšle local :
qwen3.5:9b-hermes(65536 tokens contexte, entiÚrement sur GPU)Coût : 0⏠(pas d'API cloud)
đ Dashboard temps rĂ©el
đ http://192.168.1.100:8050Auto-refresh 3s avec tendances et graphiques
Toggle AUTO/MANUEL : mode auto = batchs de tests, mode manuel = exécution live du pipeline
Filtres : source, macro, status, modĂšle â tous les rĂ©sultats historiques visibles
Diagrammes persistants : liens cliquables â ouvrir direct dans draw.io (GitHub)
Vue ChromaDB : patterns, types, stats en haut du dashboard
đ Diagramme persistant F1-F17
Le diagramme complet de l'architecture Smart MCP est versionné dans le repo :
13 pages : Vue Macro, Pipeline, Infrastructure, Workflows, Benchmarks, Carte Systémique, Architecture LLM, Optimisation, Kanban, Test, Banc de Test (F16), Amélioration Continue (F17)
đ§Ș RĂ©sultats
Métrique | Smart MCP (9B local) |
Coût | 0⏠/ mois |
Tests CI | 214 â local / 208 â CI |
Tableaux | â 100% |
Kanbans | â 90% |
Formes | â 100% |
Temps moyen | 5-15s |
Contexte | 65536 tokens |
VRAM | 6.6 GB |
đ Architecture
User Prompt â Classifier (LLM) â RAG ChromaDB (26 patterns)
â
LLM local (9B) â XML draw.io â Dashboard
â
Test Runner (214 tests) â CI/CD GitHub Actionsđ DĂ©marrage rapide
# 1. Installer les dépendances
pip install -r requirements.txt
# 2. Seeder ChromaDB
python -c "from brain.seed_patterns import seed; from brain.shape_templates import seed_shapes; seed(); seed_shapes()"
# 3. Lancer le dashboard
python dashboard/app.py # â http://0.0.0.0:8050
# 4. Lancer les tests (214)
python ci_tests.py # â ci_results.json
# 5. Mode CI (sans ChromaDB réelle)
CI=true CI_CHROMA_PATH=/tmp/chroma python ci_tests.py # â 208 testsđŹ CI/CD (GitHub Actions)
Workflow :
.github/workflows/ci.ymlJobs : lint (ruff) â test (208 tests sans LLM, seed ChromaDB dynamique)
Déclenché : push/PR vers main
đ Structure
Smart-MCP/
âââ brain/ # RAG, seed patterns, templates XML
â âââ rag.py # ChromaDB RAGBrain
â âââ seed_patterns.py # Seed 26 patterns
â âââ shape_templates.py # SHAPE_PATTERNS (15 formes + 5 architectures)
âââ dashboard/ # Dash dashboard + test runner
â âââ app.py # App Dash (956 lignes)
â âââ test_runner.py # Runner pipelines
âââ diagrams/ # Diagrammes .drawio persistants (13 pages)
âââ models/ # LLM client
â âââ llm_client.py # LLMClient avec classify_intent
âââ mcp_client/ # MCP draw.io client
â âââ drawio.py
âââ skills/ # Skills Hermes pour agents
â âââ smart-mcp-pipeline.md
â âââ drawio-mcp-integration.md
âââ .github/workflows/ # CI/CD
â âââ ci.yml # Lint + 208 tests
âââ tests/ # Tests unitaires
âââ smart_mcp.py # Pipeline principal
âââ smart_mcp_server.py # Serveur MCP
âââ ci_tests.py # Suite >200 tests (8 sections A-H)
âââ SOUL.md # Macro-rĂšgles de classification
âââ README.md # Ce fichierđ§ ChromaDB Visualisation
2 collections :
diagram_patterns(patterns XML) +feedback_history26 patterns : 5 architectures + 15 formes + 4 couleurs + 1 test + 1 généré
Chemin :
~/.hermes/profiles/default2/home/.smart-mcp/brain/chroma.sqlite3
đ Points clĂ©s
Classification automatique via LLM local
RAG avec ChromaDB (26+ patterns)
Dashboard temps réel avec mode auto/manuel
CI/CD GitHub Actions (208 tests sans LLM)
214 tests local / 208 CI â 100% verts
Diagramme persistant F1-F17 lié au dashboard
Mode manuel avec exécution live du pipeline
Visualisation ChromaDB (Neo4J / graphe)
Mode auto-apprentissage
Related MCP Connectors
Create, read and live-edit visual boards, Kanban plans, Gantt timelines and diagrams with AI agents.
End-to-end agent-managed company brain. Docs, diagrams, plans, databases, agents, Knowledge Graph.
Create, edit and organize hand-drawn diagrams (flowcharts, ER, architecture) from any AI agent.
Flowcharts that people and agents edit together on one live canvas, exported as Mermaid.
Related MCP Servers
- FlicenseNot gradedqualityAmaintenanceEnables AI coding assistants to turn natural-language descriptions of systems and data into editable, auto-laid-out Excalidraw diagrams across 25 types â from architecture and sequence diagrams to Gantt charts and scatter plots â plus optionally persist the architecture as a queryable, version-controlled knowledge graph. Runs fully offline with export to SVG and PNG, and diagrams can be revised afterward in natural language.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to turn natural-language descriptions of systems, flows, and data into hand-drawn-style Excalidraw diagrams across 25 typesâarchitecture, sequence, swimlane, ER, Gantt, charts and moreâwith automatic layout, architecture-aware styling of 50+ technologies, and export to SVG/PNG. It also lets users revise diagrams conversationally and optionally persist their system as a queryable, version-controlled architecture knowledge graph, all running fully offline with no API keys.-
- FlicenseNot gradedqualityAmaintenanceEnables AI assistants to turn plain-language descriptions of systems, flows, and data into 25 kinds of fully laid-out, hand-drawn Excalidraw diagrams â architecture, sequence, ER, swimlane, Gantt, charts and more â with auto-styling for 50+ technologies and revision in natural language. Runs fully offline, exports to SVG and PNG without a browser or Node.js, and can optionally persist the architecture as a queryable knowledge graph.-
- FlicenseNot gradedqualityAmaintenanceEnables users to generate and revise 25 types of hand-drawn Excalidraw diagramsâincluding architecture, flowcharts, sequence, ER, Gantt, and chartsâfrom natural-language specs without managing coordinates. It can also persist architecture as a queryable, version-controlled knowledge graph and export offline to SVG or PNG.-