Vyotiq AI

An Advanced AI-Powered Coding Assistant with Autonomous Agent Capabilities

Features • Installation • Quick Start • Architecture • Contributing • License

---

🌟 Overview

Vyotiq AI is a powerful desktop application that brings enterprise-grade AI assistance directly to your local development environment. Built with Electron and React, it combines multiple Large Language Models (LLMs) with an intuitive interface—all running locally with complete privacy.

💡 Why Vyotiq-ai? Unlike browser-based AI tools, Vyotiq-ai runs on your machine with direct access to your filesystem, terminal, and git repositories. Your code never leaves your computer—only API calls to your chosen AI provider.

✨ Key Highlights

Feature	Description
🔀 Multi-Provider AI	Seamlessly switch between Claude, GPT, Gemini, DeepSeek, and OpenRouter
🤖 Autonomous Agents	Advanced task execution with tool confirmation and safety guardrails
💻 Integrated IDE	xterm.js terminal and browser automation
🧠 Context-Aware	Smart summarization, prompt caching, and context window optimization
🔍 Full-Text Search	Fast codebase search powered by Tantivy (BM25) indexing
�📊 Git Integration	Review changes with git diff and commit directly
🔒 Local-First	All data stays on your machine—complete privacy

---

📢 Recent Updates

Editor Theming, Workspace UX & Indexing Improvements (v1.7.0 - February 2026)

Feature-focused release with editor theming, workspace UX improvements, and indexing/search refinements.

Highlights:

• 🎨 Monaco Custom Themes: New dark/light Monaco editor themes tuned for Vyotiq UI
• 🗂️ Workspace Tabs: Dedicated provider for multi-workspace tab state management
• ⏳ Global Loading State: Centralized loading provider for better cross-panel feedback
• 🔍 Indexing & Search: Refined workspace indexing and embedding settings across backend/frontend integration
• 💰 Cost Insights: Token cost estimation utilities for clearer usage visibility

Multi-Agent Instruction Files Support (v1.6.0 - February 2026)

New instruction files system following 2025-2026 multi-agent specification standards.

Highlights:

• 📄 AGENTS.md Support: Full implementation of the agents.md specification
• 🔀 Multi-Format: Support for CLAUDE.md, copilot-instructions.md, GEMINI.md, .cursor/rules
• 📊 TodoProgressInline: Compact task progress display in chat input header
• ⚙️ Settings UI: New instruction files configuration panel with priority ordering

UI & Developer Experience Improvements (v1.5.0 - February 2026)

Major update with new UI components, React hooks, and developer experience improvements.

Highlights:

• 🎨 New UI Components: ErrorState, LoadingState, FeatureToggle, MessageAttachments
• 🪝 React Hooks: useAsync, useFormValidation, usePagination for cleaner code
• ⚡ Performance: Event batching, improved virtualization, optimized HMR
• 📚 Documentation: Complete MCP API reference and architecture docs
• 🔧 Build System: ESLint flat config, improved Vite configuration
• 🛡️ Settings: Schema validation and enhanced appearance options

MCP System Refactoring (v1.4.0 - January 2026)

Complete architectural overhaul of the Model Context Protocol system with new registry, enhanced UI, and improved reliability.

Features:

• 📦 MCP Registry: Dynamic server discovery and installation from community sources
• 🖥️ Enhanced UI: New modals for adding, configuring, and managing MCP servers
• 🔧 Import/Export: Backup and restore MCP configurations
• 🛡️ Crash Recovery: New journal and dead letter queue for improved reliability
• ⚡ Performance: Optimized state selection hooks and improved error handling

MCP System Stability & Type Safety (v1.3.0 - January 2026)

Major improvements to MCP integration with enhanced stability, better error handling, and complete TypeScript type coverage.

Improvements:

• 🛡️ Type Safety: Fixed risk level type mismatches, added MCP metadata to tool definitions
• 🔧 API Fixes: Corrected property access patterns in tool sync and handlers
• 💚 Health Events: Extended event system with health-related events
• 🔄 Better Sync: Improved tool synchronization between MCP servers and agent
• 📘 Complete Type Declarations: Full MCP API types in global.d.ts

MCP Server Integration (v1.2.0 - January 2026)

Added comprehensive Model Context Protocol (MCP) server integration for dynamic tool discovery and external server connections.

Features:

• 🔌 Server Discovery: Automatic detection of local MCP servers from known locations
• 📡 Multiple Transports: Support for stdio and HTTP-based MCP servers
• 🔧 Dynamic Tool Integration: MCP tools seamlessly available in the agent
• 📊 Real-time Status: Live server status monitoring and reconnection handling

---

🚀 Features

🤖 AI Capabilities

Multi-Provider Support

Provider	Models	Best For
Anthropic	Claude 4.5 Sonnet, Opus, Haiku	Complex coding, long context
OpenAI	GPT-5.2, GPT-5.1, GPT-4.1, o-series	Flagship reasoning, coding
Google	Gemini 3 Pro, 2.5 Pro/Flash	Multimodal, 2M context window
DeepSeek	V3.2, V3.2 Reasoner	Reasoning, cost-effective
OpenRouter	200+ models	Access to any model via single API

Intelligent Routing & Failover

• 🎯 Auto Model Selection: Automatically picks the best model for your task
• 🔄 Task-Based Routing: Routes to specialized models (coding vs. reasoning vs. general)
• 🛡️ Automatic Failover: Gracefully handles rate limits with provider fallback chains
• 📊 Health Monitoring: Real-time provider health tracking and cost management
• 💰 Cost Tracking: Per-session and aggregate cost monitoring across providers

Advanced Context Management

• 📝 Smart Summarization: Automatically compresses long conversations
• 📊 Token Optimization: Efficiently manages context windows across providers
• 📎 Attachments: Drag & drop files, images, and PDFs into conversations
• 🗄️ Prompt Caching: Provider-specific caching (Anthropic, OpenAI, Gemini, DeepSeek)
• 🔄 Tool Result Caching: Configurable TTL-based caching for tool outputs

🛠️ Development Tools

Integrated Terminal

• xterm.js Powered: Full-featured terminal with Unicode support
• Multiple Shells: bash, zsh, PowerShell, cmd.exe
• Process Control: Start, monitor, and kill processes
• AI Execution: Let the agent run commands for you
• Real-Time Streaming: Live output streaming to UI
• Persistent Sessions: Terminal state preserved across restarts

🌐 Browser Automation

• Headless Browsing: AI can research and fetch web content
• Screenshot Capture: Visual debugging and documentation
• Content Extraction: Scrape and parse web pages
• Form Filling: Automate web interactions
• Navigation Control: Back, forward, reload, tab management
• Security Status: Check URL safety and SSL status
• Console & Network: Debug tools for web automation

🔧 LSP Integration

Multi-language code intelligence via Language Server Protocol:

• Hover Information: Type info and documentation on hover
• Go to Definition: Jump to symbol definitions
• Find References: Find all usages of a symbol
• Symbol Search: Search symbols across workspace
• Diagnostics: Real-time error and warning detection
• Code Completions: Context-aware suggestions
• Code Actions: Quick fixes and refactoring
• Rename Refactoring: Safe symbol renaming

Supported languages: TypeScript, JavaScript, Python, Go, Rust, Java, C/C++, and more.

🤖 Agent System

Built-in Tools (40+)

Category	Tools
File Operations	read, write, edit, ls, grep, glob, bulkOperations
Terminal	run, checkTerminal, killTerminal
Browser	browser_fetch, browser_navigate, browser_extract, browser_snapshot, browser_screenshot, browser_click, browser_type, browser_scroll, browser_fill_form, browser_wait, browser_hover, browser_evaluate, browser_state, browser_back, browser_forward, browser_reload, browser_console, browser_network, browser_tabs, browser_security_status, browser_check_url
LSP	lsp_hover, lsp_definition, lsp_references, lsp_symbols, lsp_diagnostics, lsp_completions, lsp_code_actions, lsp_rename
Linting	readLints for code quality checks
Dynamic	createTool for runtime tool generation

Session Management

• Persistent Sessions: Conversations saved to SQLite database
• Conversation Branching: Explore alternative paths from any message
• Message Reactions: Up/down voting for feedback
• Session Search: Find conversations by content
• Auto-Cleanup: Configurable retention (default 30 days)
• Workspace Filtering: Sessions organized by workspace

Error Recovery & Self-Healing

• Error Classification: Categorizes errors for appropriate handling
• Diagnostic Engine: Analyzes root causes
• Recovery Strategies: Retry, fallback, context reduction, tool substitution
• Self-Healing Agent: Attempts automatic recovery before failing

🔒 Safety & Compliance

Feature	Description
✅ Tool Confirmation	Review and approve dangerous operations before execution
🛡️ Safety Boundaries	Configurable limits for file operations and commands
📋 Audit Trail	Complete logging of all agent actions
🚫 Protected Paths	Prevent access to .git, .env, node_modules, etc.
⛔ Blocked Commands	Prevent dangerous shell commands
💾 Auto-Backup	Automatic backup before file modifications
📊 Compliance Monitoring	Violation detection and reporting

🐛 Debugging & Observability

Execution Tracing

• Step-by-Step Recording: Full trace of agent execution
• Breakpoint Support: Pause on specific tools or errors
• State Inspection: View context at any point
• Payload Capture: Full request/response logging
• Trace Export: JSON, Markdown, or HTML formats
• Performance Profiling: Timing and resource usage

---

📦 Installation

Prerequisites

• Node.js 20.x or higher
• npm 10.x or higher
• Git (for version control features)
• Windows 10/11, macOS 12+, or Linux (Ubuntu 20.04+)

Quick Install

# Clone the repository
git clone https://github.com/vyotiq-ai/Vyotiq-AI.git
cd Vyotiq-AI

# Install dependencies (this may take a few minutes for native modules)
npm install

# Start the development server
npm start

⚠️ Windows Users: You may need Visual Studio Build Tools for native module compilation (node-pty, better-sqlite3).

Building for Production

# Create distributable package
npm run package

# Create platform-specific installers
npm run make        # Creates .exe (Windows), .dmg (macOS), .deb/.rpm (Linux)

📁 Built packages are output to the out/ directory.

---

⚡ Quick Start

1. Configure AI Provider

1. Open Settings (gear icon or Ctrl + ,)
2. Navigate to AI Providers
3. Add your API key for at least one provider:
   • Anthropic (Claude)
   • OpenAI (GPT-4)
   • Google AI (Gemini)
   • DeepSeek (DeepSeek)
   • OpenRouter (200+ models)

2. Open a Workspace

1. Click Open Folder in the sidebar
2. Select your project directory
3. The file tree will populate automatically

3. Start Chatting

Type your request in the chat input:

"Create a React component for a user profile with local storage persistence"

"Fix the TypeScript errors in src/utils/parser.ts"

"Explain how the authentication flow works in this codebase"

4. Review & Apply Changes

• Review AI-generated code changes before committing
• Use the Undo History panel to revert any changes

---

⌨️ Keyboard Shortcuts

Action	Windows/Linux	macOS
Command Palette	Ctrl + Shift + P	Cmd + Shift + P
Quick Open File	Ctrl + P	Cmd + P
Go to Symbol	Ctrl + Shift + O	Cmd + Shift + O
Find in Files	Ctrl + Shift + F	Cmd + Shift + F
Open Settings	Ctrl + ,	Cmd + ,
Toggle Terminal	Ctrl + `	Cmd + `
New Chat Session	Ctrl + N	Cmd + N
Save File	Ctrl + S	Cmd + S
Undo	Ctrl + Z	Cmd + Z
Redo	Ctrl + Y	Cmd + Shift + Z

---

🏗️ Architecture

vyotiq/
├── src/
│   ├── main/                     # Electron main process
│   │   ├── agent/                # AI agent system (core)
│   │   │   ├── orchestrator.ts   # Main agent coordinator
│   │   │   ├── sessionManager.ts # Conversation persistence
│   │   │   ├── runExecutor.ts    # Agent loop execution
│   │   │   ├── providers/        # LLM provider integrations
│   │   │   ├── context/          # Context management
│   │   │   ├── cache/            # Caching systems
│   │   │   ├── compliance/       # Safety & compliance
│   │   │   ├── recovery/         # Error recovery
│   │   │   └── debugging/        # Execution tracing
│   │   ├── tools/                # Tool system
│   │   │   ├── implementations/  # Built-in tools
│   │   │   ├── factory/          # Dynamic tool creation
│   │   │   ├── executor/         # Tool execution engine
│   │   │   └── registry/         # Tool registry
│   │   ├── browser/              # Browser automation
│   │   ├── lsp/                  # Language Server Protocol
│   │   ├── workspaces/           # Workspace management
│   │   ├── ipc.ts                # IPC handlers
│   │   └── logger.ts             # Logging system
│   │
│   ├── renderer/                 # React frontend
│   │   ├── features/             # Feature modules
│   │   │   ├── chat/             # Chat interface
│   │   │   ├── terminal/         # Terminal emulator
│   │   │   ├── browser/          # Browser panel
│   │   │   ├── settings/         # Settings panel
│   │   │   ├── fileTree/         # File tree explorer
│   │   │   ├── sessions/         # Session management
│   │   │   ├── undo/             # Undo history
│   │   │   └── workspaces/       # Workspace switcher
│   │   ├── state/                # State management
│   │   │   ├── AgentProvider.tsx # Agent state context
│   │   │   └── UIProvider.tsx    # UI state context
│   │   ├── hooks/                # React hooks
│   │   └── components/           # Shared UI components
│   │
│   ├── shared/                   # Shared types & utilities
│   │   └── types.ts              # Type definitions
│   │
│   ├── main.ts                   # Electron entry point
│   └── preload.ts                # Preload script
│
├── docs/                         # Documentation
└── package.json

Tech Stack

Technology	Version	Purpose
Electron	39.2	Desktop application framework
React	19.2	UI library with hooks
TypeScript	5.9	Type-safe JavaScript
Vite	7.2	Lightning-fast build tool
Tailwind CSS	4.1	Utility-first styling
xterm.js	5.5	Terminal emulator
better-sqlite3	12.5	Local database
node-pty	1.1	Terminal process management

---

🔧 Configuration

Settings

All settings are accessible via the Settings panel (Ctrl + ,):

• AI Providers: Configure API keys and default models
• Terminal: Shell path, font family, cursor style
• Safety: Tool confirmation, file limits, protected paths, blocked commands
• Cache: Prompt caching, tool result caching, context caching
• Debug: Verbose logging, trace export, breakpoints
• Advanced: Context limits, summarization settings

---

🐛 Troubleshooting

❌ "Module not found" or native module errors

# Windows (PowerShell)
Remove-Item -Recurse -Force node_modules, package-lock.json
npm install

# macOS/Linux
rm -rf node_modules package-lock.json
npm install

If issues persist with native modules:

npm rebuild
# or specifically for node-pty:
npx electron-rebuild -f -w node-pty

❌ Terminal not working

• Windows: Install Visual Studio Build Tools with "Desktop development with C++" workload
• macOS: Run xcode-select --install for build tools
• Linux: Install build-essential package
• Ensure your shell (bash, zsh, powershell) is in PATH

❌ API key errors

• Verify keys are entered correctly (no extra spaces)
• Check API key permissions and usage quotas on provider dashboard
• Try a different provider to isolate the issue

❌ Performance issues

• Enable context compression for long conversations
• Reduce tool result cache size in Settings
• Restart the app to clear caches

---

🤝 Contributing

We welcome contributions! See CONTRIBUTING.md for guidelines.

Quick Start

1. Fork the repository
2. Clone your fork: git clone https://github.com/YOUR_USERNAME/Vyotiq-AI.git
3. Install dependencies: npm install
4. Create a feature branch: git checkout -b feature/amazing-feature
5. Make your changes and test: npm test && npm run lint
6. Commit: git commit -m "feat: add amazing feature"
7. Push and open a Pull Request

---

📜 License

This project is licensed under the MIT License - see the LICENSE file for details.

For third-party dependency licenses, see THIRD_PARTY_NOTICES.md.

---

⚠️ AI Usage Disclaimer

Vyotiq is an AI-assisted coding tool that integrates with third-party AI services. Please be aware:

AI-Generated Content

• No Guarantee of Accuracy: AI-generated code may contain errors, bugs, or security vulnerabilities. Always review and test before production use.
• User Responsibility: You are solely responsible for any code you accept, modify, or deploy.
• Not a Substitute for Expertise: AI assistance should complement, not replace, your professional judgment.

Third-Party AI Services

When using Vyotiq-ai, your prompts and code context are sent to your chosen AI provider(s). Comply with each provider's terms of service:

• Anthropic Terms (Claude)
• OpenAI Terms (GPT)
• Google AI Terms (Gemini)
• DeepSeek Terms
• OpenRouter Terms

Data Privacy

• Local-First Architecture: All data stored locally on your machine
• AI API Calls: Only context you provide in chat is sent to AI providers
• No Telemetry: No tracking or analytics sent to Vyotiq-ai servers

---

🙏 Acknowledgments

• Electron - Desktop framework
• xterm.js - Terminal emulator
• Anthropic, OpenAI, Google, DeepSeek, OpenRouter - AI providers
• All our contributors and supporters!

---

Built with ❤️ by the Vyotiq team

Report Bug • Request Feature • Discussions

⭐ Star us on GitHub — it motivates us a lot!