protein-design-mcp
Verified Safeby jasonkim8652
Overview
An MCP server that enables LLM agents to run end-to-end protein binder design pipelines using advanced bioinformatics tools.
Installation
docker compose run --rm -T protein-design serverEnvironment Variables
- RFDIFFUSION_PATH
- PROTEINMPNN_PATH
- COLABFOLD_PATH
- COLABFOLD_BACKEND
- ESMFOLD_MODEL
- CACHE_DIR
- MODELS_DIR
- LOG_LEVEL
- SKIP_MODEL_DOWNLOAD
Security Notes
The server primarily uses subprocess.create_subprocess_exec for running external tools (RFdiffusion, ProteinMPNN, ColabFold). While generally safer than shell=True, arguments must be carefully sanitized to prevent command injection, which appears to be handled reasonably well in the source code. Extensive network requests are made to public bioinformatics APIs (UniProt, PubMed, RCSB PDB, AlphaFold DB, ColabFold API) for data fetching. File system operations involve temporary directories and user-specified paths for PDBs, FASTA, and JSON files. No obvious hardcoded secrets were found, with configurations relying on environment variables. Containerized deployment with Docker adds a layer of isolation.
Similar Servers
zen-mcp-server
A server for coordinating and managing AI agents, likely for simulations or complex task execution, leveraging Claude LLMs.
mcp_massive
An AI agent orchestration server, likely interacting with LLMs and managing multi-agent workflows.
knowledgebase-mcp
A standardized Model Context Protocol (MCP) server that aggregates and provides AI systems with access to diverse biomedical knowledge bases and analysis tools.
open-targets-platform-mcp
This server acts as a Model Context Protocol (MCP) endpoint, enabling AI assistants to interact with the Open Targets Platform GraphQL API for drug discovery and target-disease association data.