AutoClaude - Automated Claude Code Assistant

AutoClaude is an enterprise-grade automated assistant for Claude Code that provides intelligent task queuing, workflow automation, and 24/7 processing capabilities. Now featuring Universal Language Conversion with advanced Neo-rs specialization, persistent learning, multi-agent coordination, and production-ready automation systems.

💤 "Queue up 100 tasks Friday evening, wake up Monday with everything done"

🚀 NEW: Universal Language Conversion System

🌍 Multi-Language Support

Universal Converter: Intelligent conversion between 8+ programming languages
Smart Detection: Automatic source and target language identification
Pattern Learning: AI-powered conversion improvement over time
Type Mapping: Comprehensive type system translations across languages

🎯 Supported Language Pairs

C# ↔ Rust: Complete Neo-rs ecosystem support with 25+ type mappings
JavaScript ↔ TypeScript: Seamless type annotation and modern syntax
Python ↔ Rust: Performance optimization conversions
Java ↔ Kotlin: Modern JVM language migration
Extensible Framework: Easy addition of new language pairs

🧠 Advanced Memory System

Persistent Learning: SQLite-based storage with pattern recognition
Multi-Language Patterns: Cross-language conversion optimization
Type Intelligence: Dynamic type mapping learning and validation
Conversion History: Complete audit trail and performance analytics

🤖 Universal Agent System

Multi-Language Agents: Converter and Validator agents for any language pair
Task Orchestration: Intelligent workload distribution and priority management
Real-time Monitoring: Live agent status and performance tracking
Fault Tolerance: Automatic retry logic and graceful error recovery

🎯 Agent Coordination

AgentCoordinator: Centralized task management and agent orchestration
Priority Queue System: Intelligent task prioritization with deadline awareness
Parallel Processing: Concurrent execution across multiple specialized agents
Load Balancing: Dynamic workload distribution based on agent availability
Task Dependencies: Smart dependency resolution and execution ordering
Memory Integration: Shared learning across all agent operations
Error Recovery: Automatic retry logic with exponential backoff
Performance Monitoring: Real-time agent metrics and performance analytics

🔗 Universal Validation System

Multi-Language Validation: Syntax checking for all supported languages
Compatibility Verification: Ensures functional equivalence across conversions
Security Scanning: Built-in vulnerability detection for all languages
Performance Analysis: Code quality and optimization recommendations

📊 System Monitoring & Analytics

Real-time Metrics: CPU, memory, and conversion-specific performance tracking
Alert System: Configurable thresholds with VS Code notifications
Export Capabilities: Comprehensive metrics and learning data export
Performance Dashboard: System health overview and diagnostics

📖 View Complete Enhanced Features Documentation →

🆕 v3.8.0: Complete Session Isolation & Multi-Window Independence

🔒 Complete Session Isolation

Independent VS Code Windows: Each window runs completely separate Claude sessions
Process Isolation: Separate Claude CLI processes with different PIDs
Memory Isolation: No shared state or variables between different windows
Workspace-Based Sessions: Unique session identifiers per workspace
Automatic Cleanup: Resource management and cleanup for inactive sessions

🛡️ Multi-Window Safety

Zero Cross-Contamination: Changes in one window don't affect others
Crash Isolation: Issues in one window don't impact other sessions
Independent Processing: Separate message queues per window
Configuration Isolation: Workspace-specific settings and logs

🚀 Enhanced Development Experience

Parallel Development: Work on multiple projects simultaneously
Session Monitoring: New command to view session isolation details
Improved Reliability: Enhanced fault tolerance across sessions
Seamless Multi-Window: No configuration needed for multi-window support

🔧 Technical Excellence

Backward Compatibility: No breaking changes to existing functionality
Session Registry: Global tracking of all active sessions
Resource Management: Efficient cleanup and memory management
Comprehensive Logging: Session-specific logging and debugging

🧠 Intelligent Task Automation System

🧠 Natural Language Commands

Transform high-level goals into completed work with simple commands:

"make project production ready" - Fixes tests, adds docs, creates Docker config, and more
"fix all failing tests" - Analyzes failures and fixes them automatically
"create a React website with authentication" - Builds complete project from scratch
"add comprehensive documentation" - Generates README, API docs, and architecture docs

🤖 12+ Specialized Sub-Agents

Each agent is an expert in its domain:

TestFixerAgent - Fixes failing tests automatically
DocGeneratorAgent - Creates comprehensive documentation
DockerCreatorAgent - Adds Docker support to your project
WebsiteBuilderAgent - Creates beautiful, responsive websites
CodeCleanerAgent - Refactors and organizes your codebase
And many more!

🚜 NEW: Parallel Agent Farm

Run multiple Claude agents simultaneously for massive productivity:

20-50+ Parallel Agents - Launch armies of AI agents working on your codebase
Smart Work Distribution - Intelligent task assignment prevents conflicts
Real-time Monitoring - Beautiful dashboard shows agent status and progress
Cooperating Agents - Advanced coordination protocol for complex tasks
34 Tech Stack Support - Pre-configured for popular frameworks and languages

💾 Cross-Session Task Persistence

Never lose work progress:

Tasks persist across VSCode sessions
Automatic work resumption after interruptions
Complete task history and context tracking
Recovery from failures with smart retry strategies

📊 Intelligent Project Understanding

AutoClaude learns your project:

Comprehensive file indexing and analysis
Symbol extraction and dependency tracking
Pattern recognition and code complexity analysis
Context-aware task execution

Learn more about the Automation System →

✨ Features

🌍 Universal Language Conversion

Multi-Language Support: Convert code between 8+ programming languages intelligently
Auto-Detection: Automatically detect source language and suggest optimal target languages
Pattern Learning: AI continuously improves conversion quality based on successful patterns
Type System Mapping: Comprehensive type translations with language-specific optimizations
Neo-rs Specialization: Advanced C# to Rust conversion with Neo blockchain compatibility
Validation & Quality: Multi-layer validation ensures converted code maintains functionality
Security Scanning: Built-in vulnerability detection across all supported languages

🚀 24/7 Automated Processing

Set It and Forget It: Queue hundreds of tasks and let AutoClaude work autonomously
Auto-Resume: Automatically resume processing when Claude usage limits reset - no manual intervention needed
Sleep Prevention: Keeps your computer awake during processing so work continues overnight
Smart Queue Management: Process multiple Claude Code tasks automatically with intelligent queueing
Batch Processing: Handle large workloads efficiently - perfect for weekend or overnight runs

🔧 Robust Process Management

Dependency Checking: Automatic detection and validation of Claude Code and Python dependencies
Cross-Platform Support: Works seamlessly on Windows, macOS, and Linux
IDE Compatibility: Full support for VS Code and Cursor
Error Recovery: Comprehensive error handling with automatic retry mechanisms
Health Monitoring: Continuous monitoring of Claude Code process health

⚙️ Advanced Configuration

Extensive Settings: Fine-tune every aspect of AutoClaude behavior
Configuration Validation: Built-in validation with helpful error messages
Development Mode: Special features and debugging tools for developers
Sleep Prevention: Keep your computer awake during long processing sessions

🔍 Script Runner & Quality Checks

Automated Code Quality Checks: Run predefined and custom scripts to ensure code quality
Production Readiness Check: Automatically detect TODOs, FIXMEs, placeholders, and incomplete implementations
Multi-Language Support: Built-in checks for JavaScript, TypeScript, Go, C++, Rust, C#, Java
Build & Test Verification: Ensure your project builds and all tests pass
Format Checking: Verify code formatting standards are met
GitHub Actions Validation: Check your CI/CD workflows for errors
Fix Loop: Automatically ask Claude to fix issues and re-run checks until all pass
Message Loop: Run individual messages in a loop with script checks until they pass
Custom Scripts: Add your own validation scripts in the .autoclaude folder

📊 Rich User Interface

Interactive Webview: Intuitive interface for managing queues and monitoring progress
Real-time Updates: Live status updates and progress tracking
History Browser: Browse and filter previous processing runs
Command Palette: Quick access to all AutoClaude commands

🚀 Quick Start

Prerequisites

Claude Code: Install Claude Code from https://www.anthropic.com/claude-code
Python 3.8+: Required for process management
VS Code 1.74.0+ or Cursor: Compatible with VS Code and Cursor

Installation

🖥️ Terminal Version (Standalone)

npm install -g autoclaude-terminal
autoclaude  # Start the terminal interface
ac          # Short alias

📦 VS Code Extension

Install via command palette: Extensions: Install Extensions → Search "AutoClaude"
Or install from VSIX: Download from GitHub Releases
Cursor: AutoClaude works seamlessly in Cursor with the same installation process
VS Code Web: Full compatibility with vscode.dev and github.dev

Basic Usage

Start AutoClaude: Open Command Palette (Cmd/Ctrl+Shift+P) → Claude: Start Claude Assistant
Add Messages: Queue up all your tasks - add 10, 50, or 200+ messages
Start Processing: Click "Start Processing" and walk away
Live Your Life: Go eat dinner, play with kids, sleep, or enjoy your weekend
Return to Completed Work: AutoClaude handles everything automatically, even through Claude usage limits

🔍 Script Runner - Automated Quality Checks

The Script Runner is a powerful feature that ensures your code meets production standards by running automated checks and fixing issues automatically.

Overview

The Script Checks section appears directly above the message input area, providing instant access to quality validation tools. All scripts are visible and can be enabled/disabled with checkboxes.

Built-in Scripts

Production Readiness Check 🏭
- Scans for TODOs, FIXMEs, placeholders, and incomplete implementations
- Ensures no debug code or temporary solutions remain
- Validates error handling and edge cases
Build Check 🔨
- Automatically detects your project's build system (npm, go, cargo, dotnet, etc.)
- Runs the appropriate build command
- Ensures code compiles without errors
Test Check ✅
- Detects and runs your test suite
- Supports multiple test frameworks across languages
- Ensures all tests pass before proceeding
Format Check 💅
- Validates code formatting standards
- Detects formatters (prettier, gofmt, rustfmt, etc.)
- Ensures consistent code style
GitHub Actions Check 🚀
- Validates workflow YAML syntax
- Checks for deprecated actions
- Ensures CI/CD pipelines are correctly configured

Using Script Runner

Enable/Disable Scripts
- Click checkboxes to select which scripts to run
- Scripts only run when enabled
- Your selection is saved automatically
Reorder Scripts
- Drag scripts using the handle icon (☰)
- Numbers (1, 2, 3...) show execution order
- Order affects fix priority - earlier scripts are fixed first
Run Individual Scripts ▶️
- Each script has its own run button (▶️) on the right
- Click to run just that specific script
- Useful for debugging or checking specific aspects
- Results shown via VS Code notifications
Run All - "Run Checks" 🔍
- Executes all enabled scripts once
- Shows results with ✅ pass or ❌ fail status
- Displays specific errors found
- No automatic fixing - just validation
Fix Loop - "Run Loop" 🔄
- Runs all enabled scripts
- If any fail, asks Claude to fix the issues
- Re-runs scripts after fixes
- Continues until all pass or max iterations reached
- Set max iterations (1-20, default: 5)
Custom Scripts 📝
- Add your own validation scripts to .autoclaude/scripts/
- Scripts automatically appear in the UI
- Must output JSON format:
```
{
  "passed": true/false,
  "errors": ["error1", "error2"],
  "warnings": ["warning1"] // optional
}
```
- Support any language (bash, python, node, etc.)
- Examples provided in .autoclaude/scripts/examples/

Script Configuration

Scripts are configured in .autoclaude/config.json:

{
  "scripts": {
    "production-check": {
      "enabled": true,
      "order": 1
    },
    "build": {
      "enabled": true,
      "order": 2
    }
  },
  "maxIterations": 5
}

🚜 Parallel Agent Farm - Massive Scale Automation

Run multiple Claude agents simultaneously to tackle large-scale code improvements:

Getting Started with Parallel Agents

Start Parallel Agents 🚀
- Command: Claude Agent Farm: Start Parallel Agents
- Choose number of agents (1-50)
- Agents launch in tmux sessions with automatic coordination
Monitor Progress 📊
- Command: Claude Agent Farm: Show Agent Monitor
- Real-time dashboard with agent status, context usage, and work cycles
- Visual indicators for agent health and activity
Attach to Session 🖥️
- Command: Claude Agent Farm: Attach to Agent Session
- View agents working in terminal
- Direct tmux access for debugging
Clear All Context 🧹
- Command: Claude Agent Farm: Clear All Agent Context
- Broadcasts /clear to all agents simultaneously
- Useful when multiple agents are running low on context

Advanced Features

Smart Work Distribution

Automatic chunk-based task assignment
Conflict prevention through file locking
Dynamic chunk sizing based on remaining work
Priority-based work queue management

Cooperating Agents Protocol

Agents coordinate through shared work registry
File and feature locking prevents conflicts
Stale lock detection and recovery
Business value tracking for work items

Multi-Tech Stack Support

34 pre-configured technology stacks
Auto-detection of project type
Stack-specific prompts and best practices
Custom stack configuration support

Monitoring & Recovery

Heartbeat monitoring for stuck agents
Automatic restart with exponential backoff
Settings backup and corruption recovery
Adaptive timing adjustments

Configuration

Configure parallel agents in VS Code settings:

{
  "autoclaude.parallelAgents.enabled": true,
  "autoclaude.parallelAgents.maxAgents": 50,
  "autoclaude.parallelAgents.defaultAgents": 5,
  "autoclaude.parallelAgents.staggerDelay": 10,
  "autoclaude.parallelAgents.contextThreshold": 20,
  "autoclaude.parallelAgents.autoRestart": true,
  "autoclaude.parallelAgents.coordinationEnabled": false
}

AutoClaude provides two powerful loop features that ensure your tasks are completed to perfection.

1. Script Fix Loop (Global)

The "Run Loop" button runs all enabled scripts repeatedly until they pass:

How it works:

Runs all enabled script checks
If any fail, sends errors to Claude with fix request
Claude implements fixes based on error messages
Scripts run again to verify fixes
Continues until all pass or max iterations reached

Best for:

Preparing code for production deployment
Ensuring all quality standards are met
Automated code cleanup and refactoring
Pre-commit quality assurance

Example workflow:

1. Enable: Production Check, Build, Test, Format
2. Click "Run Loop" with max iterations = 5
3. Claude fixes TODOs → Build errors → Test failures → Formatting
4. All scripts pass → Code is production-ready!

2. Message Loop (Individual Tasks)

The 🔄 button on each message runs that specific task in a quality loop. Note: Hover over a pending message to see the action buttons including the loop button (🔄).

How it works:

Processes the individual message/task
Runs enabled script checks on the changes
If checks fail, asks Claude to fix while maintaining task completion
Re-runs scripts after fixes
Ensures task is both complete AND high quality

Best for:

Complex features that need refinement
Tasks requiring multiple iterations
Ensuring individual changes meet standards
Focused quality improvement

Example workflow:

Message: "Add user authentication with JWT"
1. Click 🔄 on the message
2. Claude implements authentication
3. Scripts find: missing error handling, no tests
4. Claude adds error handling and tests
5. All checks pass → Feature is complete and production-ready!

Loop Configuration

Control loop behavior with these settings:

Max Iterations: 1-20 (default: 5)
- Lower for quick fixes
- Higher for complex refactoring
Script Selection: Enable only relevant scripts
- Faster loops with fewer scripts
- More thorough with all scripts
Smart Ordering: Drag scripts to prioritize
- Put critical checks first
- Format checks last

Pro Tips

Start Simple: Begin with 1-2 scripts, add more as needed
Watch Progress: Each iteration shows what was fixed
Trust the Process: Let Claude handle the fixes
Custom Standards: Add your own scripts for team standards
Combine Features: Queue messages + loops = fully automated development

📋 Commands

Command	Description
`Claude: Start AutoClaude`	Start the AutoClaude interface and session
`Claude: Stop AutoClaude`	Stop AutoClaude and close the session
`Claude: Add Message to Queue`	Add a new message to the processing queue
`Claude: Run Script Checks`	Run all enabled validation scripts
`Claude: Run Script Check Loop`	Run scripts and auto-fix issues in a loop

⚙️ Configuration

AutoClaude offers extensive configuration options. Access settings via File → Preferences → Settings → Extensions → AutoClaude.

Queue Management

{
  "claudeAutopilot.queue.autoMaintenance": true
}

Session Management

{
  "claudeAutopilot.session.autoStart": false,
  "claudeAutopilot.session.skipPermissions": true,
  "claudeAutopilot.session.healthCheckInterval": 30000
}

Sleep Prevention

{
  "claudeAutopilot.sleepPrevention.enabled": true,
  "claudeAutopilot.sleepPrevention.method": "auto"
}

History & Logging

{
  "claudeAutopilot.history.maxRuns": 20,
  "claudeAutopilot.history.autoSave": true,
  "claudeAutopilot.logging.enabled": false,
  "claudeAutopilot.logging.level": "info"
}

Script Runner

{
  "claudeAutopilot.scriptRunner.enabled": true,
  "claudeAutopilot.scriptRunner.maxIterations": 5,
  "claudeAutopilot.scriptRunner.autoCreateFolder": true
}

🏗️ Architecture

AutoClaude follows a modular architecture with clear separation of concerns:

src/
├── core/           # Core state, types, and configuration
├── claude/         # Claude CLI integration and communication
├── queue/          # Queue management and processing
├── scripts/        # Script runner and quality checks
├── services/       # External services (health, sleep, dependencies)
├── ui/             # User interface and webview management
└── utils/          # Shared utilities and logging

Key Components

Queue Manager: Handles message queueing, processing, and operations
Claude Integration: Manages Claude Code process and communication
Dependency Checker: Validates and manages required dependencies
Configuration System: Comprehensive settings with validation
Script Runner: Automated quality checks with fix loop capability

🔒 Security & Privacy

Local Processing: All processing happens locally on your machine
No Data Collection: AutoClaude doesn't collect or transmit personal data
Secure Dependencies: Validates Claude Code and Python installations
Permission Awareness: Uses --dangerously-skip-permissions only in trusted environments

🛠️ Development

Building from Source

# Clone the repository
git clone https://github.com/r3e-network/AutoClaude.git
cd AutoClaude

# Install dependencies
npm install

# Compile TypeScript
npm run compile

# Start watch mode for development
npm run watch

# Package for distribution
vsce package

Creating a Release

To create a new release:

Update version in package.json
Update version badge in README.md
Add changelog entry in CHANGELOG.md
Commit changes: git commit -m "chore: bump version to X.X.X"
Push to main: git push origin main
Create release branch: git checkout -b vX.X.X
Push release branch: git push origin vX.X.X

The GitHub Actions workflow will automatically:

Build the extension
Create a .vsix package
Create a GitHub release with the package as an asset
Tag the release

Development Mode

Enable development mode for additional debugging features:

{
  "claudeAutopilot.developmentMode": true
}

This enables:

Debug logging and diagnostics
Configuration validation tools
Advanced queue operations

🤖 Intelligent Sub-Agent System

AutoClaude includes 12+ specialized sub-agents that work together to analyze and improve your code:

Production Readiness Agents

🏭 Production Readiness: Scans for TODOs, placeholders, and incomplete implementations
🔨 Build Check: Automatically detects and validates your build system
✅ Test Check: Ensures all tests pass and identifies missing coverage
💅 Format Check: Validates code formatting standards
🚀 GitHub Actions: Validates CI/CD workflows and configurations

Advanced Analysis Agents

🧠 Context Awareness: Deep understanding of project structure and dependencies
📋 Task Planning: Creates detailed implementation plans for complex features
🔗 Dependency Resolution: Automatic dependency management and conflict resolution
💡 Code Understanding: Provides comprehensive code analysis and insights
🧪 Integration Testing: Automated test generation and validation
⚡ Performance Optimization: Identifies and suggests performance improvements
🔒 Security Audit: Comprehensive security vulnerability scanning

How Sub-Agents Work

Enable Sub-Agent System: Set autoclaude.subAgents.enabled to true
Run Analysis: Use "🤖 Run AI Analysis Agents" command
Automated Improvement: Agents analyze your code and provide intelligent fixes
Continuous Loop: Keep running until all checks pass or max iterations reached

🎯 Use Cases

Perfect for:

Weekend Warriors: Queue up your entire week's refactoring Friday evening
Large Refactoring Projects: Process hundreds of files while you sleep
Batch Code Generation: Generate components, tests, and documentation overnight
Migration Tasks: Convert frameworks or update dependencies during family time
Quality Assurance: Run comprehensive code reviews while you're at dinner
Documentation Generation: Create docs for your entire codebase while you relax

Real-World Examples:

Converting a React class component codebase to functional components
Adding TypeScript types to a large JavaScript project
Generating API documentation from code comments
Migrating from one testing framework to another
Adding accessibility features across a web application

📚 Wiki & Documentation

Core Concepts

Queue Processing: AutoClaude maintains a persistent queue of messages that can be processed automatically. Each message represents a task or instruction for Claude Code.

Auto-Resume: When Claude Code hits usage limits, AutoClaude automatically detects this and schedules the queue to resume when limits reset.

Workspace Integration: Each VS Code workspace maintains its own queue and history, allowing you to manage multiple projects independently.

Dependency Management: AutoClaude automatically checks for and validates all required dependencies (Claude Code, Python) before starting.

Advanced Features

Batch Operations: Process multiple related tasks in sequence with consistent context and state management.

Queue Management: Full CRUD operations on queue items - add, edit, remove, duplicate, and reorder messages as needed.

History Tracking: Complete history of all processing runs with filtering and search capabilities.

Cross-Platform: Native support for Windows, macOS, and Linux with platform-specific optimizations.

🤝 Contributing

Contributions are welcome! Please open issues and pull requests on our GitHub repository.

👥 Credits

Built for the Claude Code community to enhance development productivity and automation.

📄 License

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

🐛 Known Issues & Troubleshooting

Common Issues

Claude Code Not Found

Ensure Claude Code is installed and in your PATH
Restart VS Code after installing Claude Code
Check dependency status in AutoClaude panel

Python Not Found

Install Python 3.8 or later
Ensure Python is in your PATH
On Windows, check "Add Python to PATH" during installation

Permission Errors

AutoClaude uses --dangerously-skip-permissions for automation
Only use in trusted development environments
Disable if working with sensitive data

Getting Help

Support

If you find Claude Autopilot helpful, consider:

⭐ Starring the repository
🐛 Reporting bugs or suggesting features
🤝 Contributing code or documentation

Made with ❤️ for the Claude Code community

Claude Autopilot is not affiliated with Anthropic. Claude Code is a product of Anthropic.

AutoClaude

R3E Network

AutoClaude - Automated Claude Code Assistant

🚀 NEW: Universal Language Conversion System

🌍 Multi-Language Support

🎯 Supported Language Pairs

🧠 Advanced Memory System

🤖 Universal Agent System

🎯 Agent Coordination

🔗 Universal Validation System

📊 System Monitoring & Analytics

🆕 v3.8.0: Complete Session Isolation & Multi-Window Independence

🔒 Complete Session Isolation

🛡️ Multi-Window Safety

🚀 Enhanced Development Experience

🔧 Technical Excellence

🧠 Intelligent Task Automation System

🧠 Natural Language Commands

🤖 12+ Specialized Sub-Agents

🚜 NEW: Parallel Agent Farm

💾 Cross-Session Task Persistence

📊 Intelligent Project Understanding

✨ Features

🌍 Universal Language Conversion

🚀 24/7 Automated Processing

🔧 Robust Process Management

⚙️ Advanced Configuration

🔍 Script Runner & Quality Checks

📊 Rich User Interface

🚀 Quick Start

Prerequisites

Installation

🖥️ Terminal Version (Standalone)

📦 VS Code Extension

Basic Usage

🔍 Script Runner - Automated Quality Checks

Overview

Built-in Scripts

Using Script Runner

Script Configuration

🚜 Parallel Agent Farm - Massive Scale Automation

Getting Started with Parallel Agents

Advanced Features

Configuration

🔄 Loop Features - Intelligent Task Refinement

1. Script Fix Loop (Global)

2. Message Loop (Individual Tasks)

Loop Configuration

Pro Tips

📋 Commands

⚙️ Configuration

Queue Management

Session Management

Sleep Prevention

History & Logging

Script Runner

🏗️ Architecture

Key Components

🔒 Security & Privacy

🛠️ Development

Building from Source

Creating a Release

Development Mode

🤖 Intelligent Sub-Agent System

Production Readiness Agents

Advanced Analysis Agents

How Sub-Agents Work

🎯 Use Cases

Perfect for:

Real-World Examples:

📚 Wiki & Documentation

Core Concepts

Advanced Features

🤝 Contributing

👥 Credits

📄 License

🐛 Known Issues & Troubleshooting

Common Issues

Getting Help

Support