ADEPT: Agentic Discovery and Exploration Platform for Tools

ADEPT is a modular framework for building agentic scientific applications using the Model Context Protocol (MCP) and Streamlit. It demonstrates how to integrate LLMs, agentic tools, and workflows for scientific discovery, with a focus on teaching and rapid prototyping.

Key Features:

• Modular architecture: Streamlit UI, OpenWebUI backend, MCP tool servers, and Langchain agent orchestration
• Tool wrapping for scientific APIs (BLAST, UniProt, PubChem, RAG, etc.)
• Multi-agent orchestration (static plan and dynamic graph modes)
• LLM-agnostic layer (OpenAI, Azure, NVIDIA, Ollama, etc.)
• Persistent state management with ChromaDB
• Progressive learning path: 7 tutorial chapters (0-6) with increasing complexity

For a detailed walkthrough, see the Agentic Framework Tutorial. Each tutorial chapter builds upon the previous one, allowing users to learn progressively from basic concepts to advanced multi-agent orchestration.

---

Quick Start

Docker Compose (Recommended):

1. Copy .env.example to .env and edit as needed:

   cp .env.example .env
   # Edit .env for API keys and config

2. Build and run all services:

   docker compose build
   docker compose up -d

3. Access the Streamlit UI at http://localhost:8501

For OpenWebUI integration and advanced options, see the tutorial.

Local Development:

1. Install Python 3.11+ and uv:

   pip install uv
   uv venv .venv --python 3.11
   uv sync --all-extras

2. Run servers and app (in separate terminals):

   uv run run-mcp-server
   uv run run-streamlit-harness

---

Contributing

Contributions are welcome! Please open issues or pull requests for bug fixes, new tools, or documentation improvements. See local_development.md for developer setup.

All contributors must follow the Pacific Northwest National Laboratory open source guidelines and include the appropriate disclaimer and license notices.

---

Citation

If you use ADEPT in your research, please cite:

George, A., Bilbao, A., Agarwal, K., Mejia-Rodriguez, D., Samantray, S., Kim, H., Rice, P. S., Jacob, B., Baer, M., Raugei, S., Cheung, M. S., & Rigor, P. (2025). ADEPT: A Pedagogical Framework for Integrating Agentic AI with Deterministic Scientific Workflows. Zenodo. https://doi.org/10.5281/zenodo.17315801

---

Documentation & Resources

• Tutorial: Setup, Tool Development, Example Workflows
• Tool User Guide
• Sample User Queries
• Licenses
• Helm/Kubernetes Deployment Guide

---