CoolFace
Apppublic

ghh1125/scikit-bio

sourceHugging Facemitupdated 6mo agoView on Hugging Face
0likes
App README

scikit-bio MCP Service

This deployment package exposes selected scikit-bio capabilities via FastMCP. It supports local stdio usage and HTTP transport for Docker/HuggingFace Spaces.

Included MCP Tools

  • —health
  • —list_modules
  • —list_symbols
  • —call_function
  • —create_instance
  • —dna_reverse_complement
  • —dna_transcribe
  • —neighbor_joining_tree

Detailed usage is documented in scikit-bio/mcp_output/README_MCP.md.

Local stdio Usage

Run the local stdio entry point (for Claude Desktop / CLI):

bash
cd scikit-bio/mcp_output/mcp_plugin
python main.py

Or run via transport launcher:

bash
cd scikit-bio/mcp_output
MCP_TRANSPORT=stdio python start_mcp.py

HTTP Usage (Docker / Spaces)

The Docker image starts FastMCP directly using:

bash
python scikit-bio/mcp_output/start_mcp.py

with:

  • —MCP_TRANSPORT=http
  • —MCP_PORT=7860

Clients should connect to:

  • —https://<your-host>/mcp

Supplementary Info API

app.py provides optional local info endpoints only:

  • —GET /
  • —GET /health
  • —GET /tools

This app is not the Docker entry point.