
J.A.R.V.I.S.
J.A.R.V.I.S. is an intelligent coding assistant that leverages multiple state-of-the-art language models to help you with code generation, modifications, and technical discussions. - essenecrucix-netizen/jarvis
README
J.A.R.V.I.S. - AI Code Assistant
J.A.R.V.I.S. is an intelligent coding assistant that leverages multiple state-of-the-art language models to help you with code generation, modifications, and technical discussions.
Features
-
Multi-Model Support: Choose between different AI models for your coding needs:
- DeepSeek Coder V3
- Gemini 2.0 Flash Experimental
- Grok 2
- Qwen 2.5 Coder
- Llama 3.3 70B Instruct
- Claude 3.5 Sonnet
- GPT-4 Turbo
- GPT-4o
- o1 Preview
-
File Attachment Support:
- PDF files with text extraction
- Microsoft Word documents (.docx)
- Excel spreadsheets with sheet parsing
- Images with OCR capabilities
- Enhanced Markdown with GFM support
- All major programming languages
- Configuration files
- Text and documentation files
- File preview with syntax highlighting
- Multiple file upload support
- Progress indicators and file size display
- Type-specific icons and preview buttons
-
Real-Time Updates:
- WebSocket-based notifications
- Instant feedback for code changes
- Real-time workspace updates
- Automatic change notifications
-
Workspace Management:
- Create and manage multiple workspaces
- View workspace history
- Delete workspaces when no longer needed
- Rename workspaces
- Browse workspace file structure
-
Code Generation & Modification:
- Generate new code based on natural language prompts
- Modify existing code with AI assistance
- Preview changes before applying them
- View diffs of proposed changes
-
Interactive Chat:
- Discuss code and technical concepts
- Get explanations about existing code
- Context-aware responses based on workspace content
- Attach files for additional context
Technical Stack
-
Backend:
- Flask web framework
- Flask-SocketIO for WebSocket support
- Eventlet for async operations
-
Frontend:
- Pure JavaScript
- TailwindCSS for styling
- CodeMirror for code editing
- Socket.IO client for real-time notifications
- PDF.js for PDF processing
- Mammoth.js for Word documents
- XLSX.js for Excel files
- Tesseract.js for OCR
- Marked and Unified.js for Markdown
Installation
- Clone the repository
- Install dependencies:
pip install -r requirements.txt
- Set up your environment variables in
.env
:DEEPSEEK_API_KEY=your_deepseek_api_key GROK_API_KEY=your_grok_api_key GOOGLE_API_KEY=your_google_api_key ANTHROPIC_API_KEY=your_anthropic_api_key OPENAI_API_KEY=your_openai_api_key
Usage
- Start the server:
python app.py
- Open your browser and navigate to
http://localhost:5000
- Create a new workspace or select an existing one
- Choose your preferred AI model
- Start coding with AI assistance!
Model Capabilities
- DeepSeek Coder V3: Specialized in code generation and modification
- Gemini 2.0 Pro: Advanced code generation and natural language understanding
- Grok 2: Advanced language model for code and natural language
- Qwen 2.5 Coder: Specialized 32B model for code generation
- Llama 3.3 70B Instruct: Large context window and strong code generation capabilities
- Claude 3.5 Sonnet: Advanced reasoning and code understanding
Contributing
Contributions are welcome! Please feel free to submit pull requests.
Special Thanks
- Nikole Cardoso for her invaluable contributions and support
- Guilherme Guirro for his expertise and guidance
- Felipe Santos for his dedication and insights
Their contributions have been instrumental in making J.A.R.V.I.S. better.
Platform Compatibility
This application has been tested and confirmed working on:
- Linux (native)
- Windows Subsystem for Linux (WSL 2)
- Windows (native, no admin privileges required)
The application uses directory junctions on Windows to avoid requiring admin privileges, while maintaining symlink functionality on Unix-like systems.
License
This project is licensed under the MIT License - see the LICENSE.md file for details.
推荐服务器

e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
AIO-MCP Server
🚀 集成了 AI 搜索、RAG 和多服务(GitLab/Jira/Confluence/YouTube)的一体化 MCP 服务器,旨在增强 AI 驱动的开发工作流程。来自 Folk。
Knowledge Graph Memory Server
为 Claude 实现持久性记忆,使用本地知识图谱,允许 AI 记住用户的信息,并可在自定义位置存储,跨对话保持记忆。
https://github.com/Streen9/react-mcp
react-mcp 与 Claude Desktop 集成,能够根据用户提示创建和修改 React 应用程序。

any-chat-completions-mcp
将 Claude 与任何 OpenAI SDK 兼容的聊天完成 API 集成 - OpenAI、Perplexity、Groq、xAI、PyroPrompts 等。
Exa MCP Server
一个模型上下文协议服务器,它使像 Claude 这样的人工智能助手能够以安全和受控的方式,使用 Exa AI 搜索 API 执行实时网络搜索。
AI 图像生成服务
可以用于 Cursor 集成 MCP 服务器。
mcp-perplexity
Perplexity API 的 MCP 服务器。