Summary

EchoBot is an AI companion for immersive role-playing and emotional support, featuring Live2D avatars and real-time voice. It also acts as a productivity agent for coding and file management, accessible via web, QQ, and Telegram.

How It Works

EchoBot utilizes a "Decision-Roleplay-Agent" three-layer architecture for optimized performance and persona integrity. The Decision Layer uses rules and a lightweight LLM for rapid intent recognition and dispatching. The Roleplay Layer provides engaging, context-aware conversational responses, free from tool metadata. The Agent Core silently executes complex tasks with system permissions, tools, and memory, reporting results via the Roleplay Layer. This separation prevents token bloat and ensures swift, in-character replies.

Quick Start & Requirements

Installation involves pip install -r requirements.txt. Python 3.11+ is recommended. Configure LLM provider details (API key, model, base URL) in a .env file. Launch with python -m echobot app; the web UI is at http://127.0.0.1:8000/web.

Highlighted Details

• Live2D Integration: Immersive, real-time Live2D avatars with eye-tracking and customizable models/backgrounds.
• Multi-Platform Support: Web UI, QQ, and Telegram integrations.
• Productivity Suite: Local file I/O, scheduled/periodic tasks, tool/skill integration, and memory for complex agentic tasks.
• Voice Interaction: Built-in TTS (edge-tts, local kokoro-multi-lang) and ASR (local sherpa-onnx Sensevoice), with support for custom OpenAI-compatible endpoints.
• Flexible Routing: Manual switching between Auto-decision, Chat-only, and Force-Agent modes.

Maintenance & Community

The project acknowledges inspiration from several open-source projects but provides no direct links to community channels or specific maintenance details.

Licensing & Compatibility

The provided README does not specify a software license, creating ambiguity regarding usage rights, modification, and distribution, especially for commercial applications.

Limitations & Caveats

Requires significant LLM API key and endpoint configuration. Setup may challenge less technical users. The absence of explicit licensing is a critical adoption blocker.