Skip to content

openJiuwen-ai/jiuwenswarm

JiuwenSwarm Logo

JiuwenSwarm

Understands Your Intent, Evolves Autonomously — Swarm Collaboration for Complex Tasks

Chinese · English · Docs · Docs (EN) · Website · GitCode

License Release Python Version OS Support Coordination Engineering Skill Self-Evolution

JiuwenSwarm Interface Preview


Introduction

JiuwenSwarm is an Agent system that makes multi-agent collaboration truly work. Designed for developers and teams who need to automate complex tasks, it helps users drive multi-agent collaboration, Skill self-evolution, and tool invocation through natural language — delivering end-to-end from intent to result. Compared to similar projects, JiuwenSwarm's core differentiators are: Skill self-evolution makes capabilities grow stronger with use, swarm collaboration enables multiple Agents to specialize and coordinate on complex tasks, and multi-platform access covers mainstream IM platforms.

Why JiuwenSwarm

Capability Value
Multi-Agent Collaboration Complex tasks are no longer a single Agent's battlefield: Leader automatically decomposes tasks, assembles teams, and multiple Agents specialize and negotiate dynamically
Distributed Agent Swarm Break through single-machine compute limits: Leader/Teammate deploy across processes and machines, coordinating at scale
Swarmflow Orchestrate workflows with natural language: Leader automatically decomposes into multi-stage workflows, with Agents seamlessly handing off between stages
Skill Self-Evolution Capabilities grow stronger with use, not rigid over time: automatically detects error signals and user dissatisfaction, then optimizes Skill definitions
Skill Hub Sharing Capability assets are built once and reused everywhere: Skills can be searched, installed, combined, and remixed across developers
Auto Harness Evaluation-driven, end-to-end automated Harness optimization: not training model weights, but letting the Harness learn and optimize automatically in practice
AI Infrastructure Compatibility One system adapts to multiple inference backends: compatible with Huawei Cloud MaaS and other mainstream platforms, supports OpenAI-compatible APIs and local models
Tool Permissions & Security Every step is under your control: pre-execution approval for tools, file access whitelists, sensitive operation interception

Latest Updates

  • 2026-05-18: v0.2.0 released — JiuwenClaw officially upgraded to JiuwenSwarm, desktop app with auto-update, Swarm mode and evolution enhancements, CLI/TUI command extensions.
  • 2026-05-21: openJiuwen Roundtable livestream — open sharing on Swarm Skills Hub ecosystem co-building and Skills self-evolution.
  • 2026-06-02: v0.2.1 released — Swarm cluster mode frontend display optimization, TUI integration with Auto Harness new commands, automated evolution and scheduled task system improvements.
  • 2026-06-12: Invited to Singapore Lorong AI community event.

Installation & Launch

Desktop

Platform Download Notes
Windows Download Windows Version For Windows 10 / 11
macOS Download macOS Version For Intel / Apple Silicon

Download and follow the installer prompts to get started.

Command Line

# Install JiuwenSwarm
pip install jiuwenswarm

# Use China mirror (recommended)
pip install jiuwenswarm -i https://pypi.tuna.tsinghua.edu.cn/simple

# Initialize JiuwenSwarm (first-time setup)
jiuwenswarm-init

# Start JiuwenSwarm
jiuwenswarm-start

After launching, visit http://localhost:5173 to open the frontend.

To use TUI (terminal interface), open a new terminal after starting JiuwenSwarm:

# Install JiuwenSwarm-tui
pip install jiuwenswarm-tui

# Use China mirror (recommended)
pip install jiuwenswarm-tui -i https://pypi.tuna.tsinghua.edu.cn/simple

# Start JiuwenSwarm-tui
jiuwenswarm-tui

For detailed installation instructions, see: Install Guide

Quick Start

Configure Model

JiuwenSwarm supports multiple model platforms: Huawei Cloud MaaS, OpenAI, DeepSeek, DashScope, SiliconFlow, OpenRouter and other OpenAI-compatible APIs, as well as local model deployment.

JiuwenSwarm Model Configuration

Start a Conversation

JiuwenSwarm supports three execution modes — switch as needed:

Mode Description Best For
Plan Mode Decomposes requirements into concrete steps, executes step by step Complex tasks, need to confirm each step
Performance Mode Flexible handling, supports parallel tasks Simple tasks, fast response
Swarm Mode Leader orchestrates multiple specialized Agents in coordinated collaboration Large complex tasks, multi-role collaboration

Swarm Mode (default)

Example input:

Conduct an in-depth research on the new energy vehicle industry and generate an analysis report.

JiuwenSwarm Swarm Mode

Planning Mode

Example input:

Help me build a personal blog website — from tech selection, page design, frontend and backend development to deployment, create a complete development plan and execute step by step.

JiuwenSwarm Plan Mode

Performance Mode

Example input:

Check today's weather in Beijing, and recommend 3 books about artificial intelligence.

JiuwenSwarm Performance Mode

TUI Interface

Experience JiuwenSwarm in the terminal — ideal for headless environments or CLI-preferred users.

Example input:

Write a bubble sort algorithm for me.

JiuwenSwarm TUI Interface

For detailed operation guide, see: Quick Start

Documentation

For common usage instructions and feature documentation, see: Documentation

Roadmap

Feature Status ETA Value
Swarmflow Stateful Operators In Development 2026-07 Workflow nodes support human intervention and state memory — Humans can participate as stateful operators, approving, correcting, or taking over subsequent steps based on intermediate results
Team Mode Same-Session Feishu Integration In Development 2026-07 Feishu users join as Human Agents in the same session, collaborating with AI Agent teams — human-AI mixed squads for more precise decisions

FAQ

For solutions to common issues, see: FAQ.

Contributing

We welcome developers to contribute to JiuwenSwarm. You can contribute in the following ways:

Please read the Contributing Guide before contributing to understand the debugging workflow, code style, and commit conventions. For the contribution roadmap, see the openJiuwen Contribution Page.

Contributors

Thanks to all developers who have contributed to JiuwenSwarm: View Contributor List

Community

Channel Purpose Link
Website Product info, updates, and ecosystem Visit Website
SIG Technical roadmap, engineering practices, ecosystem building Join SIG
Swarm Skills Hub Browse, publish, and reuse JiuwenSwarm Skills Visit Swarm Skills Hub

License

This project is licensed under Apache License 2.0.

About

JiuwenSwarm is an intelligent AI Agent built on openJiuwen. It extends the powerful capabilities of large language models directly to your fingertips through various communication apps you use daily.

Resources

License

Code of conduct

Contributing

Security policy

Stars

Watchers

Forks

Packages

 
 
 

Contributors