Enabling Composable Agentic Systems for Biomedical Research
BioContextAI is an open-source initiative that bridges the gap between Large Language Models (LLMs) and specialized biomedical knowledge. We're building the infrastructure for reliable, composable AI agents that can access validated scientific resources without hallucinations or fragmented implementations.
Transform biomedical research by making AI systems more reliable, interconnected, and accessible through standardized access to domain-specific knowledge bases and tools.
We leverage the emerging Model Context Protocol standard to create composable, interoperable biomedical AI tools that can be mixed and matched across different LLM providers and applications.
🗂️ Registry - Community-driven catalog of biomedical MCP servers with rich metadata, enforcing FAIR principles and open-source standards
🖥️ Knowledgebase MCP - Production-ready MCP server providing access to essential biomedical databases
💬 Chat - Interactive interface for testing and using MCP-enabled biomedical AI assistants
🌐 Website - Documentation, tutorials, and community resources
- Find tools: Visit biocontext.ai to explore the registry and chat interface
- Integrate: Add our MCP server to your existing AI workflows by following the installation guide
- Contribute servers: Add your biomedical tools to our registry
- Extend functionality: Build upon our Knowledgebase MCP implementation
- Join development: Check our documentation and roadmap
We're building a collaborative ecosystem inspired by successful bioinformatics initiatives like scverse, BioCypher, and Bioconda. Join us in advancing biomedical AI:
- Registry contributions: Share your MCP servers with the community
- Code contributions: Improve our core infrastructure
- Documentation: Help others learn and adopt our tools
- Feedback: Report issues and suggest improvements
| Repository | Description | Language | Status |
|---|---|---|---|
| registry | Community MCP server catalog | Metadata/JSON | ✅ Active |
| knowledgebase-mcp | Biomedical MCP server | Python | ✅ Active |
| website | Main website and documentation | TypeScript/Next.js | ✅ Active |
| evaluation | Benchmarking scripts | Python | ✅ Active |
- Website: biocontext.ai
- Registry: biocontext.ai/registry
- Chat Interface: biocontext.ai/chat
- Documentation: biocontext.ai/docs
- Remote Knowledgebase MCP Server: mcp.biocontext.ai
Tba.
