AgentForge is a low-code framework tailored for the rapid development, testing, and iteration of AI-powered autonomous agents and Cognitive Architectures. Compatible with a range of LLM models — currently supporting OpenAI, Google's Gemini, Anthropic's Claude, and Ollama or LMStudio for local LLMs — it offers the flexibility to run different models for different agents based on your specific needs.
Whether you're a newbie looking for a user-friendly entry point or a seasoned developer aiming to build complex cognitive architectures, this framework has you covered.
Our database-agnostic framework is designed for seamless extensibility. While ChromaDB is our go-to database, integration with other databases is straight-forward, making it an ideal playground and solid foundation for various AI projects.
In summary, AgentForge is your beta-testing ground and future-proof hub for crafting intelligent, model-agnostic, and database-flexible autonomous agents.
Easily Build Agents or Cognitive Architectures (Multi-Agent Scripts) with the following AgentForge functionality:
- Customizable Agents
- Custom Tools & Actions
- Dynamic Prompt Templates
- Knowledge Graph Functionality
- LLM Agnostic Agents (Each Agent can call different LLMs if needed)
- On-The-Fly Prompt Editing
- OpenAI, Google & Anthropic API Support
- Open-Source Model Support (Ollama,LMStudio)
Welcome to the AgentForge framework documentation. This comprehensive guide will support you whether you're just getting started or diving deep into custom configurations and advanced features. Here, you'll find detailed insights into the various components that make up our system.
- Getting Started with AgentForge: Begin your journey with a straightforward setup guide, covering everything from initial installation to usage.
-
Agents: Dive deep into the agents' world. Learn how they operate, respond, and can be customized.
-
Modules: Explore multi-agent scripts, the hierarchies above agents. Understand how Modules coordinate various agents and manage the flow of information to achieve specific goals.
-
Knowledge Graphs: Discover how knowledge graphs form the backbone of advance knowledge creation and retrieval within AgentForge, empowering modules and agents with a rich, contextual data foundation.
-
LLM API Integration: Understand how AgentForge connects with various Large Language Model (LLM) APIs.
-
Settings: Delve into the model, storage, and system configurations – tweak the behavior of the system.
-
Personas: Personas encapsulate information accessible to the agents. Acting as a resource of knowledge for the system/agent, they are not limited to defining agents' personalities but can include any kind of information that could be utilized by the agents as needed.
-
Tools & Actions: The system's toolbox. Understand the tools available and how they can be choreographed into actionable sequences.
-
Utilities: Explore the array of utility functions and tools that supercharge the system's capabilities.
Note: Our documentation is a living entity, continuously evolving. Some links or features may still be under development. We appreciate your patience and welcome your feedback to improve the documentation.
Feel free to open issues or submit pull requests with improvements or bug fixes. Your contributions are welcome!
We're on the lookout for a UI/UX collaborator who's passionate about open-source and wants to join the team to help develop a front-end for this framework. This isn't a job offer, but rather an invitation to be a part of something cool. Interested? We'd love to chat! (See the Contact Us section below for details.)
If you're keen on contributing or just want to reach out to us, here's how to get in touch:
- Email: [email protected]
- Discord: Feel Free to drop by our Discord Server
This project is licensed under the GNU General Public License. See LICENSE for more details.