Key takeaways
- Open-source (AGPL-3.0) autonomous coding framework with up to 12 parallel agent terminals
- Full Kanban board interface for task management from planning through completion
- Built-in QA validation loop and AI-powered merge with automatic conflict resolution
FAQ
What is Auto-Claude?
Auto-Claude is an open-source desktop app where you describe a coding goal and AI agents autonomously handle planning, implementation, and quality validation in isolated git worktrees.
Is Auto-Claude free?
Yes, Auto-Claude is free and open source under AGPL-3.0. You need a Claude Pro/Max subscription for the underlying AI, which you pay Anthropic directly.
What platforms does Auto-Claude support?
Auto-Claude has native apps for Windows, macOS (Apple Silicon and Intel), and Linux (AppImage, Debian, Flatpak).
Executive Summary
Auto-Claude is an open-source autonomous multi-agent coding framework that turns Claude Code into a "virtual dev team." Developers describe a goal, and Auto-Claude handles planning, implementation, and QA validation through a Kanban-style interface. With support for up to 12 parallel agent terminals and automatic conflict resolution, it represents the most ambitious open-source entry in the autonomous coding space.
| Attribute | Value |
|---|---|
| Company | Open Source (AndyMik90) |
| Founded | 2025 |
| Funding | Community-driven |
| Contributors | Active GitHub community |
| License | AGPL-3.0 |
Product Overview
Auto-Claude takes the "autonomous" concept seriously. Rather than requiring developers to prompt and approve each step, it implements a full autonomous loop: describe your goal, and agents plan the work, implement it across isolated worktrees, run QA validation, and even handle merge conflicts automatically.
The Kanban interface provides visibility into agent progress across planning, implementation, and review stages. Each task moves through the board as agents work, with the ability to run up to 12 agent terminals in parallel.
Key Capabilities
| Capability | Description |
|---|---|
| Autonomous Tasks | Describe goal; agents handle planning, implementation, validation |
| Parallel Execution | Up to 12 agent terminals running simultaneously |
| Kanban Interface | Visual task management from planning through completion |
| Isolated Worktrees | All changes in git worktrees; main branch stays safe |
| Self-Validating QA | Built-in quality loop catches issues before review |
| AI-Powered Merge | Automatic conflict resolution when integrating to main |
| Memory Layer | Agents retain insights across sessions |
Product Surfaces
| Surface | Description | Availability |
|---|---|---|
| Windows | Native Electron app | GA (v2.7.5) |
| macOS (Apple Silicon) | Native Electron app | GA |
| macOS (Intel) | Native Electron app | GA |
| Linux (AppImage) | Portable Linux app | GA |
| Linux (Debian) | .deb package | GA |
| Linux (Flatpak) | Flatpak package | GA |
| CLI | Headless operation | GA |
Technical Architecture
Auto-Claude is built with an Electron frontend and Python backend for agent orchestration. It communicates with Claude Code CLI to run agents, using git worktrees for isolation. The security model includes OS sandboxing, filesystem restrictions, and a dynamic command allowlist.
Project Structure
Auto-Claude/
├── apps/
│ ├── backend/ # Python agents, specs, QA pipeline
│ └── frontend/ # Electron desktop application
├── guides/ # Documentation
├── tests/ # Test suite
└── scripts/ # Build utilities
Key Technical Details
| Aspect | Detail |
|---|---|
| Deployment | Local desktop app |
| Frontend | Electron |
| Backend | Python |
| Agent | Claude Code CLI |
| Isolation | Git worktrees |
| Open Source | Yes (AGPL-3.0) |
Strengths
- Truly autonomous — Agents work for hours without intervention; built for overnight/background execution
- Cross-platform — Full native apps for Windows, macOS, and Linux (multiple formats)
- Open source — AGPL-3.0 license; community can inspect, modify, and contribute
- Built-in QA — Self-validating quality loop catches issues before human review
- Smart merging — AI-powered conflict resolution reduces manual merge work
- GitHub/GitLab integration — Import issues, investigate with AI, create merge requests
- Linear integration — Sync tasks with Linear for team progress tracking
- Active development — Regular releases, responsive Discord community
Cautions
- Claude dependency — Requires Claude Pro/Max subscription; tied to single provider
- Autonomous risk — Running agents for hours unsupervised can produce unexpected results
- Electron overhead — Desktop app uses more resources than native alternatives
- Learning curve — Autonomous workflow differs significantly from prompt-based tools
- AGPL license — Copyleft requirements may not suit all commercial use cases
- Beta features — Some functionality (like Linear integration) may be experimental
Pricing & Licensing
| Tier | Price | Includes |
|---|---|---|
| Open Source | $0 | Full app, all features |
| Commercial License | Custom | For closed-source modifications |
Licensing model: AGPL-3.0 (free for open source; commercial licensing available)
Hidden costs: Claude Pro/Max subscription ($20-100/mo) required
Competitive Positioning
Direct Competitors
| Competitor | Differentiation |
|---|---|
| Augment Intent | Auto-Claude is fully autonomous; Intent is spec-driven with human checkpoints |
| Agentastic | Auto-Claude is autonomous execution; Agentastic is parallel manual control |
| Codex App | Auto-Claude is Claude-based autonomous; Codex App is OpenAI-based managed |
| Tembo | Auto-Claude is a desktop app; Tembo is agent orchestration infrastructure |
When to Choose Auto-Claude Over Alternatives
- Choose Auto-Claude when: You want agents working autonomously for extended periods
- Choose Intent when: You prefer structured spec-driven workflows with human checkpoints
- Choose Agentastic when: You want manual control over parallel terminal agents
Ideal Customer Profile
Best fit:
- Developers comfortable with autonomous AI execution
- Teams wanting overnight/background agent work
- Open source enthusiasts who value transparency
- Users on any platform (Windows, macOS, Linux)
Poor fit:
- Developers wanting fine-grained control over each agent step
- Teams concerned about autonomous code changes
- Users preferring managed SaaS over self-hosted
- Those needing multi-model flexibility (Claude-only)
Viability Assessment
| Factor | Assessment |
|---|---|
| Financial Health | Community-funded (sustainable) |
| Market Position | Leader (open source) |
| Innovation Pace | Rapid |
| Community/Ecosystem | Active (Discord, GitHub) |
| Long-term Outlook | Positive |
As the most feature-complete open-source option, Auto-Claude has carved out a strong position. The AGPL license ensures it remains open while allowing commercial licensing.
Bottom Line
Auto-Claude is the most ambitious open-source autonomous coding tool available. Its combination of Kanban interface, parallel execution, QA validation, and AI-powered merging creates a complete pipeline from goal description to merged code. The autonomous approach won't suit everyone, but for developers comfortable with AI working independently, it's remarkably capable.
Recommended for: Developers wanting autonomous agent execution, open-source advocates, and teams needing cross-platform support
Not recommended for: Users requiring fine-grained control, those uncomfortable with autonomous code changes, or teams needing multi-model flexibility
Outlook: Strong community momentum; likely to remain the leading open-source autonomous coding tool
Research by Ry Walker Research • methodology