A Model Context Protocol (MCP) server that discovers and interacts with bioinformatics services through embedded "breadcrumb" metadata in HTML pages.
- 🔍 Service Discovery: Parse HTML pages for embedded API instructions
- 🔗 MCP Detection: Identify services with native MCP support
- 📡 API Integration: Make direct API calls to discovered endpoints
- 🧬 Bioinformatics Focus: Designed for bioinformatics and computational biology services
# Clone the repository
git clone https://github.com/MCPmed/breadcrumbsMCP.git
cd breadcrumbsMCP
# Install dependencies
pnpm install
# Build the project
pnpm run devAnalyzes a service URL to discover its capabilities.
Parameters: url (string)
Makes direct API calls to discovered endpoints.
Parameters: base_url, endpoint, method, parameters (optional)
Lists all available endpoints from a service.
Parameters: url (string)
Services can embed discovery metadata in HTML comments using this format:
<!-- LLM-INSTRUCTIONS
{
"service": "MyBioService",
"version": "1.0.0",
"mcp_available": {
"status": true,
"url": "https://mcp.example.com",
"recommendation": "Use native MCP for best experience"
},
"fallback_api": {
"base_url": "https://api.example.com",
"endpoints": [
{
"path": "/v1/search",
"method": "GET",
"concept": "sequence_search",
"description": "Search for biological sequences",
"parameters": {
"query": {
"type": "string",
"description": "Search query",
"required": true,
"example": "BRCA1"
},
"limit": {
"type": "number",
"description": "Maximum results",
"required": false,
"default": 10
}
}
}
]
}
}
END LLM-INSTRUCTIONS -->Add to your MCP settings:
{
"mcpServers": {
"breadcrumbs-discovery": {
"command": "node",
"args": ["/path/to/breadcrumbsMCP/dist/index.js"]
}
}
}pnpm run dev # Build and run
tsc # Build onlyBSD-3-Clause