Sage Veterinary Imaging
Public MCP server for veterinary diagnostic imaging — 800+ provider directory, pricing, locations, educational content, and clinical resources for dogs and cats
Install
Install and configure the MCP from https://github.com/sageveterinary/sage-mcp-public now. Follow the repository's installation instructions, ask me for anything you can't complete yourself, and verify its tools load.README
Sage MCP Public
Model Context Protocol server for Sage Veterinary Imaging's public data.
What It Serves
| Table | Records | Description |
|---|---|---|
content | 536 | Website pages, blog posts, conditions, education articles |
providers | 808 | Veterinary imaging provider directory (40 states) |
pricing | 33 | SVI imaging service pricing |
locations | 3 | SVI facility details (Round Rock TX, Spring TX, Sandy UT) |
MCP Tools
search_content— Full-text search across all website contentget_page— Get a specific page by URL slugsearch_providers— Search 800+ vet imaging facilities by location/modalityget_provider— Provider detail by slugget_pricing— SVI service pricing, filterable by categoryget_location_info— SVI location details (address, phone, modalities)get_company_info— Company info, FAQs, policiesget_service_info— Imaging service descriptions
Endpoints
| Path | Purpose |
|---|---|
/ | Public landing page (docs + connect instructions, crawlable) |
/mcp/http | Streamable HTTP transport — recommended for MCP clients |
/mcp/sse | SSE transport (legacy, still supported) |
/.well-known/mcp.json | Discovery document |
/health | Health check |
Note: mcp is pinned <2.0.0 — mcp 2.x removed mcp.server.fastmcp.
Architecture
GitHub → Railway (FastAPI + MCP) → Supabase (PostgreSQL + PostgREST)
- No PHI. Completely separate from the clinical MCP server.
- Supabase REST API for all queries (no direct DB connections needed).
- Full-text search via PostgreSQL tsvector indexes.
Deployment
Railway
- Connect this repo to Railway
- Set environment variables:
SUPABASE_URL— Supabase project URLSUPABASE_SERVICE_KEY— Service role JWTPORT— Railway sets this automatically
Content Refresh
Run the sync worker to re-scrape website content:
python -m sync.refresh
Set up as a Railway cron service for weekly refresh.
Local Development
pip install -r requirements.txt
cp .env.example .env # Add your Supabase credentials
python -m server.main
Server runs at http://localhost:8000
- Health:
GET /health - MCP:
POST /mcp
Related servers
Everythingmodelcontextprotocol91KReference / test server with prompts, resources, and toolsFetchmodelcontextprotocol91KWeb content fetching and conversion for efficient LLM usageFilesystemmodelcontextprotocol91KSecure file operations with configurable access controlsGitmodelcontextprotocol91KTools to read, search, and manipulate Git repositories
