Enso
Health Warn
- License — License: MIT
- Description — Repository has a description
- Active repo — Last push 0 days ago
- Low visibility — Only 5 GitHub stars
Code Fail
- exec() — Shell command execution in dist/assets/index-CgJtiJsK.js
- network request — Outbound network request in dist/assets/index-CgJtiJsK.js
- network request — Outbound network request in dist/sw.js
- network request — Outbound network request in public/sw.js
- fs module — File system access in scripts/install.sh
- process.env — Environment variable access in scripts/personalize.cjs
- fs module — File system access in scripts/personalize.cjs
Permissions Pass
- Permissions — No dangerous permissions requested
This tool is a self-hosted, open-source AI sandbox that acts as a personalized workspace. It allows users to generate interactive applications, manage research dashboards, and execute multi-agent tasks to autonomously build and evolve software.
Security Assessment
Overall Risk: High. The audit raised a critical failure regarding shell command execution within the application's distributed JavaScript files. While this aligns with the tool's core promise of autonomously running code and self-evolving, it introduces significant security risks if deployed in sensitive environments. The scan also detected multiple outbound network requests across the application and service workers, which could potentially exfiltrate data. Additionally, the personalization script accesses both the file system and environment variables (which is expected for an interactive setup script that requests API keys, but requires strict oversight). No hardcoded secrets were detected, and no dangerous system permissions are requested.
Quality Assessment
The project is very new and currently has extremely low community visibility, evidenced by only 5 GitHub stars. However, it appears to be under active development, with the last code push occurring today. It benefits from clear documentation and a standard MIT license. Due to the lack of community vetting and a small user base, the overall trust level is quite low. This means that severe bugs or underlying security vulnerabilities in the autonomously executing code may not yet be caught or patched by the community.
Verdict
Use with caution — the autonomous code execution capabilities and low community footprint require strict isolation and a thorough manual code review before deploying.
An AI sandbox where every user owns the factory. One-command setup personalizes the entire app — Claude Code redesigns the UI for your role, builds a custom mobile APK, and the platform self-evolves sprint after sprint. AI agent teams discover, build, and ship complete solutions.
Enso — An AI sandbox that ships
Every answer is an interactive app. Every sprint makes it smarter. You own the entire factory.
What is Enso?
Enso is an open-source AI platform that doesn't just chat — it builds. Ask a question
and get a structured research dashboard, not a wall of text. Describe an app and watch
Claude Code build it in a live terminal. Run /evolve and an AI team improves your entire
installation autonomously.
Self-hosted. Self-evolving. Yours.
Why Enso?
| What You Want | Others | Enso |
|---|---|---|
| Research a topic | Wall of text (ChatGPT) | Interactive research board with 48+ sources, charts, AI podcast |
| Build an app | Component preview (v0) or code edits (Cursor) | Full app built live in Claude Code terminal |
| Improve your tools | Wait for vendor updates | AI team runs evolution sprints on YOUR installation |
| Own your workspace | Rent SaaS ($20–200/mo) | Own source code, fork it, modify it, keep it forever |
| Personalize | Config files, themes | Claude Code recompiles the entire app for your role |
Get Started
# 1. Clone
git clone https://github.com/Proxy2021/Enso.git && cd Enso
# 2. Setup (installs deps, configures API keys, personalizes your app)
./setup # macOS / Linux
.\setup.ps1 # Windows PowerShell
Setup handles everything interactively: API keys, Claude Code auth, remote access, app personalization, build, and launch. See SETUP.md for the full guide.
Three Pillars
🏭 Own the Factory
Self-hosted. Open source. Your code, your data, your API keys. No subscriptions. No vendor
lock-in. During setup, Claude Code personalizes the entire app based on who you are — a
developer gets "Forge," an investor gets "Signal," a researcher gets "Nexus."
See PERSONALIZATION-SHOWCASE.md for examples.
📱 Every Answer Is an App
Research questions become interactive dashboards. Build requests become running apps.
File operations become a desktop-grade file manager. 15+ built-in app types, plus custom
apps from a single command.
🧬 AI That Evolves Itself
Type /evolve and an AI team — project leader, architect, engineers, QA — runs a full
improvement sprint. Sprint scores: 4.5 → 7.5 over 4 cycles. No other developer tool does this.
Built-in Capabilities
| Capability | Description |
|---|---|
| 🔍 Research Engine | 48+ source analysis with structured boards, AI podcast, PDF export |
| 💻 Claude Code | /code — live terminal coding with model selection (Opus/Sonnet/Haiku) |
| ⚡ Orchestrator | Multi-agent teams with dependency graphs for complex tasks |
| 📸 Photo Studio | 28 film/cinematic styles, batch processing, EXIF editing |
| 🌐 Remote Browser | Full browser inside the conversation |
| 📂 File Manager | Desktop-grade with previews and CRUD |
| 🖥️ Remote Desktop | Control remote machines from within Enso |
| 🔬 Mission Planner | AI discovers project opportunities tailored to your interests |
| ✨ Card Evolution | Click Evolve on any card — AI team turns it into a polished app |
| 📝 Summarizer | One-click text + AI podcast summaries for any card |
How It Works
Your message
│
▼
┌─────────────────────────────────────────────────────┐
│ Task Router (auto-classification) │
│ │
│ "What's the weather?" ──→ Simple → Agent chat │
│ "Fix this bug" ──→ One-off → Claude Code │
│ "Build a CRM system" ──→ Complex → Multi-agent │
│ orchestration with DAG │
└──────────────────────────────┬──────────────────────┘
│
▼
Tool Result + Template
│
▼
Interactive React App (instant)
No commands. No configuration. The system figures out the right level of effort.
Tech Stack
Frontend: React 19, Zustand 5, Tailwind CSS 4, Vite 6, Capacitor (Android)
Backend: Express, WebSocket, TypeScript, node-pty
AI: Claude Code, Multi-provider LLM (Claude, Gemini, GPT, DeepSeek, Ollama, OpenRouter)
Mobile: Capacitor Android APK + PWA (offline-capable)
Architecture
Project structuresrc/ React frontend
├── cards/ Card renderers (DynamicUI, Terminal, Shell, Orchestration, Mission)
├── components/ Timeline, ChatInput, CardContainer, ConnectionPicker, MemoryPanel
├── store/chat.ts Zustand state (cards, streaming, connections, evolution)
└── lib/ WS client, JSX sandbox (Sucrase), EnsoUI (17 components), connections
server/ OpenClaw channel plugin (backend)
├── apps/ Shipped app packages (app.json + template.jsx + executors/)
└── src/
├── server.ts Express + WS server with auth
├── channel.ts OpenClaw ChannelPlugin implementation
├── inbound.ts Message routing (browser → OpenClaw dispatch)
├── outbound/ Response delivery, card actions, enhancement, context
├── llm-provider.ts Multi-provider LLM abstraction (callChatLLM)
├── task-router.ts 3-tier message classifier (simple/one-off/orchestrated)
├── orchestrator.ts Multi-agent planner (goal → task DAG) with inline targeting
├── orchestrator-engine.ts DAG executor with parallel agents
├── card-evolution.ts Card-type-specific evolution via orchestration
├── card-summarizer.ts Universal content extraction + text/podcast generation
├── researcher-tools.ts Two-phase streaming research pipeline
├── build-via-claude.ts Natural language → app build via Claude Code
├── claude-code.ts Claude Code CLI integration (NDJSON streaming)
└── tool-factory.ts Template refinement, auto-heal executor
shared/types.ts WebSocket protocol types (shared frontend ↔ backend)
Documentation
| Document | Contents |
|---|---|
| SETUP.md | One-command setup, personalization, prerequisites, troubleshooting |
| PERSONALIZATION-SHOWCASE.md | 6 persona apps with screenshots |
| CLAUDE-REFERENCE.md | App building API, ExecutorContext, template rules |
| CLAUDE.md | Full architecture reference for AI-assisted development |
| PROJECTS.md | Project import, AI team generation, evolution sprints |
| CONTRIBUTING.md | How to contribute to Enso |
Contributing
We welcome contributions of all sizes — bug fixes, new app types, UI improvements, docs, translations. See CONTRIBUTING.md for guidelines.
License
MIT License — free forever.
Reviews (0)
Sign in to leave a review.
Leave a reviewNo results found