- Category
- AI Agents
- Rank
- No. 205Tools index
- Pricing
- Open Source
- Type
- AGENT
- Builder
- agentscope-ai
- GitHub
- 35.1k stars
- Latest release
- v2.2.1-beta.1
- Date
About
Self-hosted personal AI assistant from AgentScope, easy to deploy locally or in the cloud, with chat-app integrations and extensible capabilities.
What it does
QwenPaw is an agent workspace centered on persistent conversations, editable Markdown memory, files, and scheduled work. Users can chat through a browser console, terminal interface, desktop app, or messaging services. Agent modes, plugins, skills, and external tool connectors turn those conversations into coding, document, research, and automation workflows.
Why it's ranked here
The appeal comes from unusually broad coverage inside one coherent system. Memory, file operations, scheduling, multiple agent modes, channel delivery, and extension APIs share the same sessions. Policy gates, command guards, sandboxing, and credential encryption also address risks created by capable agents. The cost is a large operational surface and dependency set.
What's good
Memory remains inspectable rather than disappearing into an opaque store: conversations and resources become readable, editable, searchable, linked Markdown. The same agent and sessions carry across the console, terminal, and channels. Extensions can add HTTP routes, agent tools, commands, lifecycle hooks, and FastAPI routers. Tests are organized into unit, contract, integration, channel, and end-to-end categories.
Tradeoffs
The base Python package pulls in channel clients, browser automation, desktop UI, model tooling, security libraries, and a language server. That breadth can make installation and maintenance heavier than a focused assistant library. Python support is limited to versions from 3.11 through 3.13. The script installer downloads its package manager and dependencies, and the documentation warns about restricted networks, corporate firewalls, and constrained Windows environments.
How to use it well
Choose QwenPaw when one agent should retain context, work with local files, run scheduled tasks, and deliver results through several interfaces. Start with one model, one channel, and a narrow skill set, then add permissions and extensions deliberately. It does not replace a small model client or lightweight library when you only need isolated API calls without memory, scheduling, governance, or user interfaces.
Technical notes+
pyproject.toml defines a setuptools package under src, requires Python >=3.11,<3.14, exposes both qwenpaw and copaw console scripts, and declares extensive default plus optional dependencies. src/qwenpaw/cli/main.py implements a Click LazyGroup, lazily imports commands, defaults the API endpoint to 127.0.0.1:8088, and launches the TUI for a bare invocation. src/qwenpaw/pawapp/app.py provides the PawApp SDK over PluginApi, buffering routes, tools, commands, routers, and lifecycle hooks until registration. Makefile and pyproject.toml define unit, contract, integration, channel, end-to-end, coverage, and quick-test surfaces, with coverage configured against src/qwenpaw and a 50 percent floor.
Observed
- License
- Apache 2.0
- Primary language
- Python
- Packaging
- PyPI package using setuptools, with script installers for macOS, Linux, and Windows
- Python support
- Python 3.11 through 3.13
- Interfaces
- CLI, browser console, terminal UI, desktop app, messaging channels, HTTP extension routes, and MCP integration
- Platform support
- macOS, Linux, and Windows
- Testing structure
- Unit, contract, integration, channel, end-to-end, and coverage configurations are present
Read from README.md, Makefile, setup.py, pyproject.toml, src/qwenpaw/__init__.py, src/qwenpaw/__main__.py, src/qwenpaw/cli/main.py, src/qwenpaw/pawapp/app.py, src/qwenpaw/envs/__init__.py, src/qwenpaw/loop/__init__.py, src/qwenpaw/hooks/__init__.py, src/qwenpaw/modes/__init__.py, src/qwenpaw/tauri/__init__.py.
What it can do
Provide AI-powered conversational assistance
Text messages or questions from user → AI-generated responses and answers
Deploy AI assistant locally
Installation commands and local server configuration → Running AI assistant instance on local machine
Deploy AI assistant to cloud
Cloud platform credentials and deployment configuration → Running AI assistant instance on cloud infrastructure
Integrate with chat applications
Chat platform API credentials and configuration → AI assistant functionality within existing chat apps
Extend capabilities through plugins
Custom plugin code or third-party extensions → Enhanced AI assistant with additional features
Process multi-agent workflows
Complex tasks requiring multiple AI agents → Coordinated responses from AgentScope framework
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.