Summary This project offers a unified, voice-controlled orchestrator for coordinating three distinct AI agents: OpenAI for voice interaction, Claude Code for software development, and Gemini for web automation. It targets developers and power users seeking to leverage multi-agent systems for complex AI-driven workflows, providing a powerful framework for integrated AI capabilities.

How It Works

The system uses a voice-first orchestration model where the OpenAI agent interprets commands and dispatches tasks via a defined toolset. Claude Code agents handle coding and file operations within a configurable working directory, while Gemini agents perform web automation using Playwright. A key feature is real-time, zero-setup multi-agent observability, streaming events from all agents to a local dashboard. Session management is supported through registry files for state persistence.

Quick Start & Requirements

Requires Python 3.11+, Astral uv, and API keys for OpenAI, Anthropic, and Google. Installation involves cloning the repo, configuring API keys in .env, and running playwright install. The primary execution command is uv run big_three_realtime_agents.py --voice.

Highlighted Details

• Multi-Agent Observability: Zero-setup, real-time event streaming from all agents to a local dashboard (http://localhost:3000).
• Integrated Agent Types: Combines voice (OpenAI), coding (Claude Code), and browsing (Gemini) capabilities.
• Tool-Based Dispatch: Comprehensive tools for agent lifecycle management and task execution.
• Cost Monitoring: report_costs() tool tracks API usage.

Maintenance & Community

Associated with "IndyDevDan" and "Tactical Agentic Coding." Specific community channels, roadmaps, or notable contributors/sponsorships are not detailed in the provided README.

Licensing & Compatibility

The license type and compatibility notes for commercial use are not specified in the provided README content.

Limitations & Caveats

The core orchestrator script is over 3000 lines, potentially impacting maintainability. Numerous improvements are listed (error handling, persistence, testing), indicating an active development phase. Operation requires multiple paid API keys.