Skip to content

Home

Jump to bottom
Paul Rigor edited this page Jun 25, 2026 · 3 revisions

ADEPT

Agentic Discovery and Exploration Platform for Tools

License: BSD-2-Clause Python 3.11+ MCP

ADEPT is an open-source platform for building secure, production-ready AI agent systems with integrated scientific tooling. Built on the Model Context Protocol (MCP) and LangGraph, it provides a complete framework for deploying LLM-powered agents across multi-cloud infrastructure.

Key Capabilities

  • 28+ Built-in Scientific Tools -- BLAST, UniProt, PubChem, AlphaFold, web search, sandboxed code execution, and ChromaDB-backed RAG pipelines
  • Multi-Agent Orchestration -- Router mode (supervisor dispatches to specialized workers) and Graph mode (LangGraph DAGs with dynamic state transitions)
  • LLM-Agnostic -- Seamless provider switching via LiteLLM: Azure OpenAI, AWS Bedrock, Anthropic, Ollama, NVIDIA NIMs, and more
  • 10 Client Interfaces -- Claude Desktop, Claude Code, ChatGPT, Roo Code, Continue.dev, Cursor, OBot.ai, AG-UI, OpenWebUI, and JupyterLab through a single API surface
  • A2A Federation -- Agent-to-Agent protocol for cross-organization agent sharing and gateway mesh communication
  • Enterprise Security -- Keycloak OIDC/JWT authentication with per-tool ACL-based RBAC and sandboxed code execution via nsjail
  • Multi-Cloud Deployment -- AWS CDK, Azure Pulumi, GCP Terraform; Helm charts for Kubernetes; Docker Compose for local development
  • Observability -- Full request tracing and token cost tracking via Langfuse integration

Architecture

ADEPT implements a three-tier secure architecture separating authentication, orchestration logic, and tool execution:

Tier Component Role
1 Agent Gateway Pure authentication proxy -- JWT validation, CORS, transparent proxying
2 Orchestration Service Core brain -- LangGraph agents, OpenAI-compatible API, state management
3 MCP Tool Servers Stateless executors -- scientific tools, HPC pipelines, sandboxed code

See Architecture Overview for the full architecture documentation.

Quick Start

# 1. Clone the repository
git clone https://github.com/pnnl/adept-agentic.git
cd adept-agentic

# 2. Configure environment
cp .env.example .env
# Edit .env with your LLM provider credentials

# 3. Start the platform
make start

Prerequisites:

  • Docker Engine 24+ with Compose v2
  • Python 3.11+ (for SDK and CLI tools)
  • At least one LLM provider configured (Azure OpenAI, AWS Bedrock, Anthropic, or Ollama)

Once running, the Agent Gateway is available at http://localhost:8083/v1 with an OpenAI-compatible API.

Documentation Sections

Section Description
Getting Started Role-based onboarding for domain scientists, platform engineers, and system administrators
Architecture Three-tier design, MCP tool system, Slurm HPC integration, multi-agent orchestration, A2A federation, and security model
Deployment Docker Compose for local development, Kubernetes/Helm for production, multi-cloud IaC
User Guides CLI reference, connectors, file management, JupyterLab, RAG workflows
CI/CD GitHub Actions pipelines, versioning conventions, and the ASOPB security benchmark
Testing Test strategy, running tests, container-based execution
Contributing Contribution guidelines and code review process
Reference Changelog, known issues, troubleshooting, roadmap

Full Documentation Site

For the complete documentation with search, navigation, and diagrams, visit the ADEPT GitHub Pages site.

Technical Report

Rigor, P. et al. "ADEPT: A Pedagogical Framework for Integrating Agentic AI with Deterministic Scientific Workflows." Pacific Northwest National Laboratory, 2026.

Read on OSTI.gov

License

BSD 2-Clause License. Copyright Battelle Memorial Institute 2026.

Home

Getting Started

Architecture

Deployment

User Guides

Developer Tools

CI/CD

Testing

Contributing

Reference

Clone this wiki locally