BlockAGI is an open-source, hackable research agent designed for iterative, domain-specific research, primarily in cryptocurrency but adaptable to other fields. It leverages LangChain and OpenAI to automate information gathering, refinement, and evaluation, presenting findings through a real-time web interface.

How It Works

BlockAGI automates research tasks by employing a structured approach inspired by AutoGPT and BabyAGI. It utilizes LangChain for orchestrating LLM interactions and managing agent workflows. The agent can access live data from the internet or custom databases, process information, and generate detailed narrative reports. Its design prioritizes efficiency by working well with gpt-3.5-turbo-16k, offering a cost-effective alternative to gpt-4 for research.

Quick Start & Requirements

• Install via poetry install.
• Requires Playwright dependencies: poetry run playwright install.
• OpenAI API key is mandatory.
• Python 3.x (implied by Poetry and LangChain).
• Setup involves cloning, installing dependencies, configuring .env with an API key, and running poetry run python main.py.
• Official documentation: https://github.com/orgexyz/BlockAGI#documentation

Highlighted Details

• Efficiency: Optimized for gpt-3.5-turbo-16k, offering cost savings.
• Interactive Web UI: Real-time visualization of agent progress and decision-making.
• Focused Functionality: Dedicated to research assistance, not general-purpose task execution.
• Simplified Setup: No external datastore or file I/O required, avoiding Docker/sandboxing.

Maintenance & Community

• Contributions are encouraged via GitHub Issues and Pull Requests.
• Community engagement is fostered through a Discord server.
• Roadmap and contribution guidelines are available in the repository.

Licensing & Compatibility

• Licensed under the Apache License 2.0.
• Permissive license allows for commercial use and integration with closed-source projects.

Limitations & Caveats

The project is primarily focused on research tasks and may not be suitable for general-purpose AI agent applications. While designed for simplicity, advanced customization requires understanding LangChain and agent architecture.