Auto Claude

Autonomous multi-agent coding framework that plans, builds, and validates software for you.

---

Download

Stable Release

Platform	Download
Windows	Auto-Claude-2.7.5-win32-x64.exe
macOS (Apple Silicon)	Auto-Claude-2.7.5-darwin-arm64.dmg
macOS (Intel)	Auto-Claude-2.7.5-darwin-x64.dmg
Linux	Auto-Claude-2.7.5-linux-x86_64.AppImage
Linux (Debian)	Auto-Claude-2.7.5-linux-amd64.deb
Linux (Flatpak)	Auto-Claude-2.7.5-linux-x86_64.flatpak

Beta Release

⚠️ Beta releases may contain bugs and breaking changes. View all releases

Platform	Download
Windows	Auto-Claude-2.7.2-beta.10-win32-x64.exe
macOS (Apple Silicon)	Auto-Claude-2.7.2-beta.10-darwin-arm64.dmg
macOS (Intel)	Auto-Claude-2.7.2-beta.10-darwin-x64.dmg
Linux	Auto-Claude-2.7.2-beta.10-linux-x86_64.AppImage
Linux (Debian)	Auto-Claude-2.7.2-beta.10-linux-amd64.deb
Linux (Flatpak)	Auto-Claude-2.7.2-beta.10-linux-x86_64.flatpak

All releases include SHA256 checksums and VirusTotal scan results for security verification.

---

Requirements

• Claude Pro/Max subscription - Get one here
• Claude Code CLI - npm install -g @anthropic-ai/claude-code
• Git repository - Your project must be initialized as a git repo

---

Quick Start

1. Download and install the app for your platform
2. Open your project - Select a git repository folder
3. Connect Claude - The app will guide you through OAuth setup
4. Create a task - Describe what you want to build
5. Watch it work - Agents plan, code, and validate autonomously

---

Features

Feature	Description
Autonomous Tasks	Describe your goal; agents handle planning, implementation, and validation
Parallel Execution	Run multiple builds simultaneously with up to 12 agent terminals
Isolated Workspaces	All changes happen in git worktrees - your main branch stays safe
Self-Validating QA	Built-in quality assurance loop catches issues before you review
AI-Powered Merge	Automatic conflict resolution when integrating back to main
Memory Layer	Agents retain insights across sessions for smarter builds
GitHub/GitLab Integration	Import issues, investigate with AI, create merge requests
Linear Integration	Sync tasks with Linear for team progress tracking
Cross-Platform	Native desktop apps for Windows, macOS, and Linux
Auto-Updates	App updates automatically when new versions are released

---

Interface

Kanban Board

Visual task management from planning through completion. Create tasks and monitor agent progress in real-time.

Agent Terminals

AI-powered terminals with one-click task context injection. Spawn multiple agents for parallel work.

Roadmap

AI-assisted feature planning with competitor analysis and audience targeting.

Additional Features

• Insights - Chat interface for exploring your codebase
• Ideation - Discover improvements, performance issues, and vulnerabilities
• Changelog - Generate release notes from completed tasks

---

Project Structure

Auto-Claude/
├── apps/
│   ├── backend/     # Python agents, specs, QA pipeline
│   └── frontend/    # Electron desktop application
├── guides/          # Additional documentation
├── tests/           # Test suite
└── scripts/         # Build utilities

---

CLI Usage

For headless operation, CI/CD integration, or terminal-only workflows:

cd apps/backend

# Create a spec interactively
python spec_runner.py --interactive

# Run autonomous build
python run.py --spec 001

# Review and merge
python run.py --spec 001 --review
python run.py --spec 001 --merge

See guides/CLI-USAGE.md for complete CLI documentation.

---

Korean Fork Modifications (Auto-Claude_ko)

This fork adds dynamic language support for AI agent responses with enhanced security. Changes are designed for extensibility and community contributions.

Modified Files

Backend - Language Support

apps/backend/prompts_pkg/prompt_generator.py

• Purpose: Core language instruction generation with dynamic detection
• Changes:
  • Added get_supported_languages() - Scans i18n/locales/ folder to auto-detect available languages
  • Added get_user_language_instruction() - Generates language-specific instructions for AI prompts
  • Implements prompt injection prevention with sanitization (removes special characters, enforces 50-char limit)
  • Supports CJK languages (Korean, Japanese, Chinese) via Unicode regex
• How it works: Reads AUTO_CLAUDE_USER_LANGUAGE and AUTO_CLAUDE_USER_LANGUAGE_NAME environment variables from frontend, validates against detected languages, sanitizes input, and prepends language instruction to agent prompts

apps/backend/prompts_pkg/prompts.py

• Purpose: Inject language instructions into QA agent prompts
• Changes:
  • Imports get_user_language_instruction() from prompt_generator
  • Prepends language instruction to get_qa_reviewer_prompt() - QA Reviewer sees user's preferred language
  • Prepends language instruction to get_qa_fixer_prompt() - QA Fixer responds in user's language
• How it works: When QA agents run, they receive localized prompts if user has set a non-English language preference

apps/backend/spec/pipeline/agent_runner.py

• Purpose: Apply language preference to Spec creation agents
• Changes:
  • Imports get_user_language_instruction() from prompt_generator
  • Adds language instruction at the top of agent prompts in run() method
  • Logs language instruction length for debugging
• How it works: Spec agents (Gatherer, Writer, Critic, etc.) receive language instructions and generate specifications in user's preferred language

apps/backend/runners/roadmap/executor.py

• Purpose: Apply language preference to Roadmap generation agents
• Changes:
  • Imports get_user_language_instruction() from prompt_generator
  • Adds language instruction at the top of agent prompts in AgentExecutor.run_agent() method
  • Logs language instruction length for debugging
• How it works: Roadmap agents (Discovery, Features) receive language instructions and generate roadmaps in user's preferred language

apps/backend/ideation/generator.py

• Purpose: Apply language preference to Ideation generation agents
• Changes:
  • Imports get_user_language_instruction() from prompt_generator
  • Adds language instruction in IdeationGenerator.run_agent() method
  • Adds language instruction in run_recovery_agent() method for error recovery
• How it works: Ideation agents (Code Improvements, UI/UX, Security, etc.) receive language instructions and generate ideas in user's preferred language

Frontend - Language Transmission

apps/frontend/src/main/agent/agent-process.ts

• Purpose: Pass user's language preference from settings to backend agents
• Changes:
  • Imports AVAILABLE_LANGUAGES from i18n constants for validation
  • In setupProcessEnvironment() method (for task execution):
    • Reads settings.language from user preferences
    • Validates language against AVAILABLE_LANGUAGES list (frontend validation layer)
    • Creates languageEnv object with environment variables:
      • AUTO_CLAUDE_USER_LANGUAGE - Language code (e.g., "ko", "fr")
      • AUTO_CLAUDE_USER_LANGUAGE_NAME - Display name from i18n config (e.g., "한국어", "Français")
    • Adds languageEnv to process environment with proper priority (after gitBashEnv, before claudeCliEnv)
    • Includes error handling with try-catch and console logging for debugging
  • In getCombinedEnv() method (for roadmap/ideation generation):
    • Reads settings.language from user preferences
    • Creates languageEnv object with same environment variables as above
    • Returns combined environment with language settings included
    • Ensures roadmap and ideation agents receive language preferences
• How it works:
  • Task execution uses spawnProcess() → setupProcessEnvironment() → language settings passed
  • Roadmap/ideation uses spawnRoadmapProcess()/spawnIdeationProcess() → getCombinedEnv() → language settings passed
  • Both paths now support user's language preference, bridging the gap between frontend UI settings and backend AI agents

Added Files

tests/test_language_detection.py

• Purpose: Comprehensive test suite for language detection and security validation
• Features:
  • Tests dynamic language detection from i18n folder
  • Validates English default behavior (no instruction)
  • Verifies French and Korean support (if folders exist)
  • Security tests:
    • Unsupported language rejection
    • Prompt injection prevention (malicious newlines, special characters)
    • Length limit enforcement (50 chars)
• How to use: python tests/test_language_detection.py
• Output: 7 tests covering functionality and security (all pass)

Key Features

1. Backend Dynamic Language Detection (Security Layer)

• Purpose: Validates user language input against allowed languages to prevent prompt injection attacks
• How it works: Backend scans apps/backend/prompts_pkg/locales/ folder to auto-detect available languages
• Security validation: User-provided language code is checked against this whitelist before generating AI prompts
• Defense in depth: Even if frontend is bypassed (e.g., CLI usage, environment variable manipulation), backend rejects unauthorized languages

2. Security (Prompt Injection Prevention)

• Allowlist validation: Only languages with locales folders are accepted by backend
• Character sanitization: Removes dangerous characters (newlines, control chars, markdown injection)
• Length limit: Language names capped at 50 characters
• Multiple validation layers:
  • Frontend: Validates against AVAILABLE_LANGUAGES in constants/i18n.ts when user selects language in UI
  • Backend: Validates against get_supported_languages() when generating AI prompts
• Why both layers?: Frontend can be bypassed (CLI, direct API calls), so backend provides final security enforcement

3. Extensibility

To add a new language (requires 3 steps):

Step 1: Create translation files

mkdir -p apps/frontend/src/shared/i18n/locales/ja
# Add translation JSON files: common.json, settings.json, etc.

Step 2: Register in frontend constants Edit apps/frontend/src/shared/constants/i18n.ts:

export type SupportedLanguage = 'en' | 'fr' | 'ko' | 'ja';  // Add 'ja'

export const AVAILABLE_LANGUAGES = [
  { value: 'en' as const, label: 'English', nativeLabel: 'English' },
  { value: 'fr' as const, label: 'French', nativeLabel: 'Français' },
  { value: 'ko' as const, label: 'Korean', nativeLabel: '한국어' },
  { value: 'ja' as const, label: 'Japanese', nativeLabel: '日本語' }  // Add this
] as const;

Step 3: Register in i18next resources Edit apps/frontend/src/shared/i18n/index.ts:

// Add imports
import jaCommon from './locales/ja/common.json';
import jaSettings from './locales/ja/settings.json';
// ... (import all 11 translation files)

export const resources = {
  en: { /* ... */ },
  fr: { /* ... */ },
  ko: { /* ... */ },
  ja: {  // Add this
    common: jaCommon,
    settings: jaSettings,
    // ... (add all namespaces)
  }
} as const;

Backend auto-detection (no changes needed):

• prompt_generator.py - Auto-scans folder (security validation only)
• prompts.py - Uses dynamic function (no hardcoding)
• agent_runner.py - Uses dynamic function (no hardcoding)

4. Supported Languages (Current)

• English (en) - Default, no instruction needed
• French (fr) - Full support with accent characters
• Korean (ko) - CJK support (requires locales/ko/ folder)
• Japanese (ja) - CJK support (requires locales/ja/ folder)
• Chinese (zh) - CJK support (requires locales/zh/ folder)

Testing

Run the test suite to verify:

python tests/test_language_detection.py

Expected output:

[PASS] Detected languages: ['en', 'fr']
[PASS] English returns empty string (correct)
[PASS] French language instruction generated
[PASS] Malicious newlines removed
[PASS] Prompt injection sanitized
[PASS] Length limited to 50 chars
[PASS] All tests completed!

Troubleshooting

Language patch not working (still getting English responses)

Symptom: You've installed Auto-Claude_ko but AI responses are still in English even though Korean is selected in settings.

Cause: If you previously had Auto-Claude installed in a different location, your autoBuildPath setting may point to the old (unpatched) installation.

How Auto-Claude works:

• Settings are stored in ~/AppData/Roaming/auto-claude/settings.json (Windows) or ~/.config/auto-claude/settings.json (macOS/Linux)
• autoBuildPath tells the app where to find the backend Python code
• When set to an old installation, the app uses that folder's code instead of the current Git repo

Solution:

1. Open Settings → General
2. Check the "Auto-Build Source Path" setting
3. If it points to an old Auto-Claude folder:
   • Click "Browse" and select the current Auto-Claude_ko/apps/backend folder
   • Or delete the old folder entirely and reinstall
4. Restart the app for changes to take effect

Verification:

• After changing paths, run a roadmap or ideation task
• AI responses should now appear in Korean
• If still in English, check the Settings path again and ensure you're using the Auto-Claude_ko repository

Design Philosophy

Convention over Configuration: Following industry standards (i18next, Django, VS Code), language support is driven by folder structure rather than configuration files. This reduces maintenance burden and enables non-developers to contribute translations.

Security by Default: All user input (language names) is sanitized before injection into AI prompts. Multiple validation layers ensure malicious input cannot compromise agent behavior.

---

Development

Want to build from source or contribute? See CONTRIBUTING.md for complete development setup instructions.

For Linux-specific builds (Flatpak, AppImage), see guides/linux.md.

---

Security

Auto Claude uses a three-layer security model:

1. OS Sandbox - Bash commands run in isolation
2. Filesystem Restrictions - Operations limited to project directory
3. Dynamic Command Allowlist - Only approved commands based on detected project stack

All releases are:

• Scanned with VirusTotal before publishing
• Include SHA256 checksums for verification
• Code-signed where applicable (macOS)

---

Available Scripts

Command	Description
npm run install:all	Install backend and frontend dependencies
npm start	Build and run the desktop app
npm run dev	Run in development mode with hot reload
npm run package	Package for current platform
npm run package:mac	Package for macOS
npm run package:win	Package for Windows
npm run package:linux	Package for Linux
npm run package:flatpak	Package as Flatpak (see guides/linux.md)
npm run lint	Run linter
npm test	Run frontend tests
npm run test:backend	Run backend tests

---

Contributing

We welcome contributions! Please read CONTRIBUTING.md for:

• Development setup instructions
• Code style guidelines
• Testing requirements
• Pull request process

---

Community

• Discord - Join our community
• Issues - Report bugs or request features
• Discussions - Ask questions

---

License

AGPL-3.0 - GNU Affero General Public License v3.0

Auto Claude is free to use. If you modify and distribute it, or run it as a service, your code must also be open source under AGPL-3.0.

Commercial licensing available for closed-source use cases.

---

Star History