How do I use Construction Reference MCP Remote Server?

1. Click on "Deploy Server". 2. Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state. 3. In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Construction Reference MCP Remote Server What's the reference price range for cement in Guatemala?" 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.

Construction Reference MCP Remote Server

by DavidDominguez-11

Overview Schema Source Related Servers Discussions Admin

Python

Remote

Construction Reference MCP Remote Server

A remote Model Context Protocol (MCP) server that provides reference information about construction materials for small projects (houses, offices, commercial spaces) in Guatemala.

Built for the CC3067 Redes course project at Universidad del Valle de Guatemala.

Disclaimer: All material prices are illustrative reference values only. Actual prices vary by supplier, quality, location, date, quantity, and delivery conditions. This is not financial, professional, or legal advice.

Features

Manual JSON-RPC 2.0 implementation (no MCP SDK).
MCP lifecycle: initialize, notifications/initialized, tools/list, tools/call.
get_material_reference tool — looks up construction materials by name or alias.
11 pre-loaded materials with price ranges in GTQ.
FastAPI used only for HTTP routing (POST /mcp, GET /health).
Docker-ready with Google Cloud Run deployment configuration.

Related MCP server: comprasal-mcp

Prerequisites

Python 3.12+
pip
Docker (optional, for container builds)
Google Cloud CLI (optional, for deployment)

Installation

# Clone the repository
git clone https://github.com/YOUR_USER/construction-reference-mcp-remote.git
cd construction-reference-mcp-remote

# Create virtual environment
py -3.12 -m venv .venv

# Activate (Windows PowerShell)
.\.venv\Scripts\Activate.ps1

# Activate (Linux/macOS)
# source .venv/bin/activate

# Install with dev dependencies
pip install -e ".[dev]"

Local Execution

python -m construction_reference_mcp_remote.main

The server starts on http://0.0.0.0:8080 by default. Override with environment variables:

PORT=9000 python -m construction_reference_mcp_remote.main

Local HTTP Test Examples

Health check

curl http://localhost:8080/health
# {"status":"ok"}

Initialize

curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "initialize",
    "id": 1,
    "params": {
      "protocolVersion": "2025-06-18",
      "capabilities": {},
      "clientInfo": {"name": "test-client", "version": "0.1.0"}
    }
  }'

List tools

curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -d '{"jsonrpc":"2.0","method":"tools/list","id":2}'

Call a tool

curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "id": 3,
    "params": {
      "name": "get_material_reference",
      "arguments": {"material_name": "cement"}
    }
  }'

Unknown material (error case)

curl -X POST http://localhost:8080/mcp \
  -H "Content-Type: application/json" \
  -d '{
    "jsonrpc": "2.0",
    "method": "tools/call",
    "id": 4,
    "params": {
      "name": "get_material_reference",
      "arguments": {"material_name": "unobtainium"}
    }
  }'

Docker

Build

docker build -t construction-reference-mcp .

Run

docker run -p 8080:8080 construction-reference-mcp

Then test with curl http://localhost:8080/health.

Cloud Run Deployment

See docs/deployment.md for complete instructions covering:

Required Google Cloud APIs
Manual deployment with gcloud
Continuous deployment from GitHub
Environment variables
Finding the service URL
Configuring the MCP host
Inspecting logs
Deleting resources

Quick Deploy

gcloud builds submit --tag gcr.io/YOUR_PROJECT_ID/construction-reference-mcp
gcloud run deploy construction-reference-mcp \
  --image gcr.io/YOUR_PROJECT_ID/construction-reference-mcp \
  --region us-central1 --platform managed --allow-unauthenticated --port 8080

MCP Endpoint

The MCP endpoint is:

POST https://<cloud-run-url>/mcp

Or locally:

POST http://localhost:8080/mcp

All requests must use Content-Type: application/json and contain valid JSON-RPC 2.0 messages.

Tool Specification

`get_material_reference`

Returns reference information for a construction material.

Parameters:

Parameter	Type	Required	Description
`material_name`	string	Yes	Name or alias of the material
`location`	string	No	Location context (default: `"Guatemala"`)

Available materials: concrete block 10 cm, concrete block 15 cm, cement, sand, gravel, reinforcing steel bar, electrical conduit, electrical cable, PVC pipe, ceramic floor tile, interior paint.

Aliases are supported (e.g., "rebar" → reinforcing steel bar, "pintura" → interior paint).

Example response:

{
  "material_name": "concrete block 15 cm",
  "category": "masonry",
  "unit": "unit",
  "reference_price_range": {
    "minimum": 8.0,
    "maximum": 11.0,
    "currency": "GTQ"
  },
  "reference_updated_at": "2026-09-01",
  "location": "Guatemala",
  "observations": [
    "Standard 15 cm concrete block for structural walls.",
    "Verify strength, dimensions, and supplier delivery conditions."
  ],
  "disclaimer": "Reference values only. Actual prices may vary by supplier, quality, location, date, quantity, and delivery conditions."
}

Error response (unknown material):

{
  "content": [{"type": "text", "text": "Material 'unobtainium' not found..."}],
  "isError": true
}

Testing

pytest -v

Tests cover:

JSON-RPC parsing and validation
Invalid JSON and unknown methods
MCP lifecycle (initialize, tools/list, tools/call)
Material aliases and case-insensitive lookup
Price-range data validity
Missing/invalid parameters
Unknown material error
Default location behavior
HTTP content-type validation
Health endpoint
Full HTTP end-to-end interactions

No Docker, Cloud Run, external APIs, or network connection required.

Wireshark Analysis Notes

When capturing traffic between the MCP host and this remote server:

TCP — Three-way handshake (SYN, SYN-ACK, ACK).
TLS — Client Hello, Server Hello, certificate exchange (Cloud Run uses HTTPS).
HTTP POST /mcp — JSON-RPC messages in the body.
JSON-RPC sequence — initialize → notifications/initialized → tools/list → tools/call.

Since Cloud Run uses HTTPS, the JSON-RPC payload is encrypted in Wireshark. To see cleartext JSON-RPC, capture locally on the loopback interface (tcp.port == 8080).

See docs/deployment.md for detailed Wireshark instructions.

Limitations

Reference prices are illustrative and do not reflect live market data.
Location parameter is informational only — prices do not change by location.
No authentication or authorization (suitable for course project scope).
No SSE/streaming — uses simple HTTP request/response.
Single-threaded material lookup from a static JSON file.

Project Structure

construction-reference-mcp-remote/
├── src/
│   └── construction_reference_mcp_remote/
│       ├── __init__.py
│       ├── main.py                    # Entry point
│       ├── config.py                  # Configuration
│       ├── protocol/
│       │   ├── errors.py              # JSON-RPC error codes
│       │   ├── jsonrpc.py             # JSON-RPC 2.0 implementation
│       │   ├── mcp_handler.py         # MCP method dispatcher
│       │   └── http_server.py         # FastAPI HTTP layer
│       ├── tools/
│       │   └── material_reference.py  # get_material_reference tool
│       └── data/
│           └── material_references.json
├── tests/
│   ├── test_jsonrpc.py
│   ├── test_material_reference.py
│   ├── test_mcp_handler.py
│   └── test_http_e2e.py
├── docs/
│   ├── guide.md
│   ├── agent-brief.md
│   └── deployment.md
├── Dockerfile
├── .dockerignore
├── cloudbuild.yaml
├── pyproject.toml
├── .env.example
└── README.md

License

MIT

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityMaintained

ResponsivenessNo issues

Resources

GitHub Repository

Need Help?

Related Servers

Related MCP Connectors

Homecastr Remote MCP A
io.github.dhardestylewis
Remote MCP endpoint for U.S. home forecasts, public benchmark data, and permit or zoning readiness.
Real Estate Finance
2
Worldcastr Remote MCP
io.github.dhardestylewis
Built-environment forecasts, public benchmarks, and permit or zoning readiness through remote MCP.
Real Estate Government Data
1
fastmcpruntime-mcp
com.clauxel.fastmcpruntime
A paid remote MCP for hosted MCP server, built to return verdicts, receipts, usage logs, and audit-r
Agent Orchestration Monitoring
2
context7docs-mcp
com.clauxel.context7docs
A paid remote MCP for Context7 MCP docs, built to return verdicts, receipts, usage logs, and audit-r
Documentation Access Monitoring
2

Related MCP Servers

Kolmo Construction
Real Estate App Automation
Kolmo-Construction
A
license
A
quality
A
maintenance
MCP server providing 12 tools for Seattle-area home remodeling: real-time cost estimation across 8 project types, contractor business info, project portfolio, blog content, and quote submission. Connects via Streamable HTTP — no auth required.
Updated 6 days ago (2026-09-16 16:53 UTC)
36
MIT
comprasal-mcp
Government Data Open Data Search
glassbit
A
license
A
quality
D
maintenance
Enables querying El Salvador's public procurement data, including awarded processes, supplier contracts, and institution details, from the live COMPRASAL API through an MCP interface.
Updated 3 months ago (2026-06-27 20:38 UTC)
8
1
MIT
CAFIR (Imóveis Rurais)
Government Data Real Estate
mcp-dir
A
license
Not graded
quality
C
maintenance
MCP server to query Brazil's rural property registry (CAFIR) using the CIB code, providing read-only consultation of rural property data via a hosted API with prepaid credits.
Updated a month ago (2026-08-15 20:25 UTC)
MIT
FGTS: Guia de Arrecadação
Government Data Finance Open Data
mcp-dir
A
license
Not graded
quality
C
maintenance
Enables consultation of FGTS (Brazilian severance guarantee fund) collection guides from official sources via a hosted, read-only MCP server. Supports natural language queries through any MCP client with pay-per-use credit.
Updated 23 days ago (2026-08-30 19:45 UTC)
MIT

Construction Reference MCP Remote Server

Features

Prerequisites

Installation

Local Execution

Local HTTP Test Examples

Health check

Initialize

List tools

Call a tool

Unknown material (error case)

Docker

Build

Run

Cloud Run Deployment

Quick Deploy

MCP Endpoint

Tool Specification

get_material_reference

Testing

Wireshark Analysis Notes

Limitations

Project Structure

License

Maintenance

Resources

Related MCP Connectors

Related MCP Servers

Kolmo Construction

comprasal-mcp

CAFIR (Imóveis Rurais)

FGTS: Guia de Arrecadação

`get_material_reference`