One Platform. All Phases. Every Device.
统一平台·全开发阶段·跨全设备
AutoDev Xiuper is the AI-native development platform built on Kotlin Multiplatform, covering all 7 phases of SDLC ( Requirements → Development → Review → Testing → Data → Deployment → Operations) and supporting 8+ platforms (IDEA, VSCode, CLI, Web, Desktop, Android, iOS, Server).
- IntelliJ IDEA Plugin: JetBrains Marketplace
- VSCode Extension: Visual Studio Marketplace
- CLI Tool:
npm install -g @xiuper/cli - Web Version: https://web.xiuper.com/
- Desktop & Android: Release Pages
- iOS: You need to build it from source code.
| Module | Platform | Status | Description |
|---|---|---|---|
| mpp-idea | IntelliJ IDEA | ✅ Production | Jewel UI, Agent toolwindow, code review, remote agent |
| mpp-vscode | VSCode | ✅ Production | Xiuper Agent |
| mpp-ui (Desktop) | macOS/Windows/Linux | ✅ Production | Compose Multiplatform desktop app |
| mpp-ui (CLI) | Terminal (Node.js) | ✅ Production | Terminal UI (React/Ink), local/server mode |
| mpp-ui (Android) | Android | ✅ Production | Native Android app |
| mpp-web (Web) | Web | ✅ Production | Web app |
| mpp-server | Server | ✅ Production | JVM (Ktor) |
| mpp-ios | iOS | 🚧 Production Ready | Native iOS app (SwiftUI + Compose) |
Xiuper Edition represents a major milestone in AI-powered development:
- One Platform: Unified Kotlin Multiplatform architecture - write once, deploy everywhere
- All Phases: 7 specialized agents covering complete software development lifecycle
- Requirements → Development → Review → Testing → Data → Deployment → Operations
- Every Device: Native support for 8+ platforms with zero compromise on performance
- IDE: IntelliJ IDEA, VSCode
- Desktop: macOS, Windows, Linux (Compose Multiplatform)
- Mobile: Android, iOS (Native + Compose)
- Terminal: CLI (Node.js with React/Ink)
- Web: Modern web app
- Server: Remote agent server (Ktor)
- Multi-LLM Support: OpenAI, Anthropic, Google, DeepSeek, Ollama, and more
- DevIns Language: Executable AI Agent scripting language for workflow automation
- MCP Protocol: Extensible tool ecosystem via Model Context Protocol
- Code Intelligence: TreeSitter-based parsing for Java, Kotlin, Python, JS, TS, Go, Rust, C#
- Global Ready: Full internationalization (Chinese/English)
AutoDev Xiuper includes 7 specialized AI agents mapped to the complete Software Development Lifecycle (SDLC):
| Agent | SDLC Phase | Description | Capabilities | Status |
|---|---|---|---|---|
| Knowledge | Requirements | Requirements understanding and knowledge construction with AI-native document reading and analysis | DocQL / Context Engineering | ✅ Stable |
| Coding | Development | Autonomous coding agent with complete file system, shell, and tool access capabilities | MCP / SubAgents / DevIns DSL | ✅ Stable |
| Review | Code Review | Professional code review analyzing code quality, security, performance, and best practices | Linter / Summary / AutoFix | ✅ Stable |
| Testing | Testing | Automated testing agent that generates test cases, executes tests, and analyzes coverage | E2E / Self-healing / Coverage | 🚧 Coming |
| ChatDB | Data | Database conversation agent supporting Text-to-SQL and natural language data queries | Schema Linking / Multi-DB / Query | ✅ Stable |
| WebEdit | Deployment | Web editing agent for browsing pages, selecting DOM elements, and interacting with web content | Inspect / Chat / Mapping | 🔄 Beta |
| Ops | Operations | Operations monitoring agent for log analysis, performance monitoring, and alert handling | Logs / Metrics / Alerts | 🚧 Coming |
Each agent is designed to handle specific phases of the development lifecycle, providing comprehensive AI assistance from requirements gathering to production operations.
In addition to SDLC agents, AutoDev Xiuper includes specialized agents for specific use cases:
| Agent | Purpose | Description | Capabilities | Status |
|---|---|---|---|---|
| Artifact | Quick Demos | Generate self-contained, executable artifacts (HTML/JS, React, Node.js, Python) inspired by Claude's Artifacts | Interactive preview / Auto-fix / Multi-format support | ✅ Stable |
Artifact Agent focuses on creating complete, runnable artifacts without file system or shell access. Perfect for:
- Quick prototyping and demos
- Interactive web applications
- Data visualizations
- Python scripts with dependencies
- SVG graphics and Mermaid diagrams
SubAgents are specialized micro-agents that can be invoked by the main Coding Agent to handle specific tasks. They follow the "Agent as Tool" architecture pattern:
| SubAgent | Purpose | Key Features | Platform Support |
|---|---|---|---|
| NanoDSL Agent | Generate AI-native UI code from natural language descriptions | Token-efficient DSL / Component generation / State management / HTTP requests | All platforms |
| PlotDSL Agent | Generate statistical charts and data visualizations from natural language | ggplot2-inspired syntax / Multiple chart types / Themes / Lets-Plot rendering | JVM Desktop & Android |
| Chart Agent | Generate chart configurations for ComposeCharts library | Pie/Line/Column/Row charts / Data analysis / Cross-platform rendering | All platforms |
| Analysis Agent | Intelligently analyze and summarize any type of content (logs, errors, JSON, code, etc.) | Content type detection / Smart summarization / Metadata extraction | All platforms |
| Codebase Investigator | Investigate codebase structure, patterns, dependencies, and architectural issues | Architecture analysis / Pattern detection / Dependency mapping / Issue identification | All platforms |
| Domain Dict Agent | Generate domain dictionaries from codebase analysis for better context understanding | Hot file detection / Class/method extraction / Domain term identification | All platforms |
| Error Recovery Agent | Analyze errors and suggest fixes with self-healing capabilities | Error pattern recognition / Fix suggestion / Auto-retry logic | All platforms |
| SQL Revise Agent | Revise and optimize SQL queries based on schema and execution feedback | Schema-aware correction / Query optimization / Syntax validation | All platforms |
SubAgents enable modular, composable AI workflows where complex tasks are broken down into specialized sub-tasks, each handled by an expert agent.
This code is distributed under the MPL 2.0 license. See LICENSE in this directory.
