hermes-architectural-mcp
Converts architectural PDF plans into dimension-verified millimetre geometry, IFC models, and CPU-rendered views, with built-in validation for boundaries, areas, and overlaps.
README
Hermes Architectural MCP
An open-source, Docker-first MCP service that converts existing architectural plans into dimension-verified millimetre geometry, IFC models, and consistent CPU-rendered views for Hermes Studio.
What it guarantees
The service separates visual interpretation from deterministic geometry. It will not export IFC or render a project until its explicit PlanSpec passes boundary, expected-area, containment, and overlap checks.
PDF dimensions are treated as candidates only. Exact geometry must be represented in millimetres and validated. This prevents a 100 m2 plan from being presented as a visually enlarged but dimensionally unsupported design.
Capabilities
- inspect vector or raster PDF plans and render a review preview;
- detect dimension text candidates without trusting them automatically;
- validate site boundaries, spaces, total area, containment, and overlaps;
- reject visual-only geometry and boundaries resized merely to match a stated area;
- export validated walls as IFC using IfcOpenShell;
- queue CPU Blender render jobs so long renders do not hold the MCP connection open;
- persist source files, reports, IFC,
.blend, and PNG outputs; - expose internal download URLs so Hermes can copy deliverables into its workspace;
- expose tools through MCP Streamable HTTP at
/mcp.
This is an early engineering release. It does not replace architectural, structural, life-safety, accessibility, or local-code review by licensed professionals.
Deploy with Dockhand
The provided compose.yaml joins an existing external Docker network named proxy and does not publish a host port.
-
Create or update a Dockhand stack using
compose.yaml. -
Confirm the
proxynetwork already exists. -
Deploy the stack.
-
From another container on
proxy, verify:curl http://architectural-mcp:8000/health -
In Hermes Studio, add an HTTP MCP server with:
http://architectural-mcp:8000/mcp -
Import
hermes-skill/SKILL.mdas a Hermes skill.
The container stores projects in the named volume architectural_projects.
Internal PDF upload
The MCP tool accepts base64 PDFs, but local files are more efficiently uploaded over the internal Docker network:
curl --fail --silent --show-error \
-H 'Content-Type: application/pdf' \
--data-binary '@/workspace/plan.pdf' \
'http://architectural-mcp:8000/api/projects/my-project/plan-pdf'
No upload route is published to the host in the supplied stack.
Development
uv sync --extra dev
uv run ruff check src tests
uv run pytest -q
docker build -t hermes-architectural-mcp:local .
MCP tools
service_infoanalyze_plan_pdfcreate_or_update_planvalidate_projectexport_project_ifcrender_project_viewsget_render_joblist_project_outputs
See examples/restaurant-100m2.json for the PlanSpec shape.
Container publishing
Pushes to main run tests and publish linux/amd64 images to:
ghcr.io/twuijri/hermes-architectural-mcp:latest
License
MIT
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.