agentscope-ai/QwenPaw

Your Personal AI Assistant; easy to install, deploy on your own machine or on the cloud; supports multiple chat apps with easily extensible capabilities.

What it solves

QwenPaw is a personal AI assistant designed to be deployed locally or in the cloud, providing a secure, extensible, and multi-channel interface for automating tasks, managing personal knowledge, and interacting with various software tools.

How it works

It operates as an "Agent OS" architecture using a three-pillar system for each agent: Resources (disk-based), Governance (security policies), and a Sandbox (cross-platform). It utilizes a three-layer memory system (live context, verbatim history, and a self-evolving personal knowledge base via ReMe) to ensure information is not lost or merely summarized. The system supports local models (QwenPaw-Flash) or cloud providers via a protocol-neutral connector layer (MCP/A2A/ACP).

Who it’s for

It is intended for users who want a private, self-hosted AI assistant that can automate workflows, process documents, manage code, and be reachable across multiple messaging platforms (e.g., Discord, Telegram, WeChat).

Highlights

  • Self-evolving Memory: Uses ReMe to turn conversations and resources into a searchable, linked Markdown knowledge base.
  • Security-First: Includes a kernel-level sandbox, tool and file guards, and skill scanners to block dangerous commands.
  • Multi-Channel Access: Single instance connectivity to DingTalk, Lark, WeChat, Discord, Telegram, and iMessage.
  • Extensible Architecture: Supports a plugin marketplace, MCP integration for external tools, and custom skill-based workflows.
  • Local-First Deployment: Can run entirely locally without API keys or cloud dependencies using QwenPaw-Flash models.

Related

  • Project
    QwenAudio/qwen-audio-agent

    A realtime voice runtime that enables AI agents to maintain continuous, full-duplex conversations while simultaneously executing background tasks and tool calls.

  • Project
    SAIL-Research-Lab/cheetahclaws

    A Python-native agent harness for AI coding assistants that supports any LLM provider or local model, offering a hackable alternative to closed-source agent frameworks.

  • Project
    pocketpaw/pocketpaw

    A self-hosted AI agent that runs locally on your machine, providing a native desktop app and integration with multiple messaging platforms like Discord and Slack.

  • Project
    Rfym21/Qwen2API

    Qwen‑Proxy (Qwen2API) is a self‑hosted Bun server that converts Alibaba’s Qwen chat and CLI services into an OpenAI‑compatible (and partially Anthropic‑compatible) HTTP API. It supports chat, multimodal generation, function calling, per‑account proxy rotation, a web UI, and can be run via Docker, Docker‑Compose, or a single binary.

  • Project
    zhayujie/CowAgent

    CowAgent is an open‑source, self‑hosted AI assistant that plans tasks, runs tools, keeps a personal knowledge base, and can operate as a team of cooperating agents. It supports many LLM providers, multimodal input, dozens of chat channels (Telegram, Slack, WeChat, etc.), a plug‑in skill system, and a three‑tier memory with nightly “Deep Dream” distillation. Install via a one‑line script or Docker, manage everything through a web console, and extend the system through the public Skill Hub or custom tool integrations.