101 tools, 16+ apps, 7 data sources, AI-inferred focus areas, and a unified brain that learns from your books, browsing, projects, and conversations.
Enso is a self-hosted AI platform with a Knowledge Cortex at its center. Ask a question and get a structured research dashboard, not a wall of text. Connect your data sources — Kindle library, YouTube subscriptions, browser history, email, projects — and watch the Cortex populate with hundreds of interlinked wiki pages on day one. Describe an app and watch Claude Code build it in a live terminal. Run /evolve and an AI team improves your entire installation autonomously.
101 tools. 16+ apps. 7 data sources. 1 brain that compounds. Yours forever.
| What You Want | Others | Enso |
|---|---|---|
| Research a topic | Wall of text | Interactive research board with 48+ sources, charts, AI podcast |
| Build an app | Component preview or code edits | Full app built live in Claude Code terminal |
| Track knowledge | Re-derive from scratch every time | Knowledge Cortex: AI-maintained wiki that compounds with every use |
| Improve your tools | Wait for vendor updates | AI team runs evolution sprints on YOUR installation |
| Stay current | Manual news reading | Daily intelligence briefing: AI searches web for your interests, emails a digest |
| Own your workspace | Rent SaaS ($20-200/mo) | Own source code + data + build pipeline. Fork it. Modify it. Keep it. |
# 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 PowerShellSetup handles everything interactively: API keys, Claude Code auth, remote access, app personalization, build, and launch. See SETUP.md for the full guide.
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.
Research questions become interactive dashboards. Build requests become running apps. File operations become a desktop-grade file manager. 16 built-in apps, plus unlimited custom apps from a single command.
Type /evolve and an AI team — project leader, architect, engineers, QA, marketing, sales — runs a full improvement sprint with real browser testing, code implementation, and validation. No other tool does this.
| Capability | Description |
|---|---|
| 🧠 Knowledge Cortex | The ONLY brain — 680+ interlinked wiki pages auto-populated from 7 data sources, treemap graph, web discovery, AI digest, daily intelligence briefing with email delivery |
| 🎯 Focus Areas | AI-inferred goals from your Cortex data — concrete outcomes you're working toward. Progressive refinement through evidence + conversation. Deeper intention analysis reveals the WHY behind each goal and suggests adjacent pursuits |
| 🔍 Research Engine | 48+ source analysis with structured boards, AI podcast, contradiction detection, deep research escalation via Claude Code |
| 💻 Claude Code | /code — live terminal coding with model selection (Opus/Sonnet/Haiku), extended thinking |
| ⚡ Orchestrator | Multi-agent teams with DAG-based task decomposition for complex goals |
| 🧬 Evolution | AI team sprints: 7 phases, 6 parallel agents, persona testing, fix-verify loops |
| 📸 Photo Studio | 56 film/cinematic styles, batch processing, AI analysis, EXIF editing |
| 🎬 Video Studio | AI video generation from text prompts via Seedance |
| 🌐 Remote Browser | Full Puppeteer-driven browser inside the conversation |
| 📂 File Manager | Desktop-grade with previews, search, CRUD, 14 file operations |
| 🖥️ Remote Desktop | Control remote machines with screenshots, click, type, scroll |
| 🐚 Terminal | Full PTY terminal (PowerShell/bash/zsh) with xterm.js rendering |
| 📊 Data Analyzer | CSV/JSON analysis with statistics, charts, queries |
| 📚 Kindle Library | Browse 471+ books with covers, ratings, categories, one-click research into Cortex |
| 📺 YouTube Manager | 208 subscriptions, liked videos, feed analysis — all ingested into Cortex |
| 🌐 Browser Data | History + bookmarks combined — browsing patterns and interests into Cortex |
| 📧 Email Scanner | Communication patterns, contacts, topics — feeds into Cortex |
| 📁 Projects Scanner | Tech stack detection across local codebases — project knowledge into Cortex |
| 💻 System Info | Environment analysis — installed apps, running processes |
| 🔬 Mission Planner | AI VC team discovers project opportunities tailored to your interests |
| ⏰ Scheduled Tasks | Durable cron system: run any tool or prompt on a schedule |
| 🔄 Settings Transfer | Export/import 9 categories of data across machines |
| 🛒 ClawHub | Skill store: browse, install, manage OpenClaw extensions |
Based on Karpathy's LLM Wiki pattern — the Cortex is the ONLY brain for memory, profile, and knowledge. Instead of re-deriving knowledge from scratch each time, the LLM incrementally builds and maintains a persistent, interlinked markdown knowledge base that compounds with every interaction.
Grand Unification architecture: buildEnsoContext() reads from Cortex only. Memory, user profile, interests, and all accumulated knowledge live as wiki pages. No separate memory or profile systems — the Cortex is the single source of truth.
7 data source apps (each a standalone Enso app with its own UI):
- Kindle Library (
server/apps/kindle/) — 471 books with covers, ratings, highlights, categories - YouTube Manager (
server/apps/youtube/) — 208 subscriptions, liked videos, feed analysis - Browser Data (
server/apps/browser/) — history + bookmarks combined - Email Scanner — communication patterns and contacts
- Projects Scanner — tech stack detection across local codebases
- System Info — installed apps, running processes, environment
- Manual ingest — any topic via the Cortex Explorer
Auto-ingest pipeline: scan -> cache -> change detection -> Cortex pages. Per-item pages (one book, one channel) are deterministic templates with zero LLM cost. Synthesis pages use AI to find cross-source patterns.
First-run onboarding: Connect data sources during setup, populate the Cortex with 680+ pages from day one.
Explorer app: Dashboard with stats and knowledge gaps, treemap graph visualization (60+ nodes, 280+ connections), full-text search with tag cloud, article reader with backlinks and related pages.
Discovery: Enter any topic — AI searches the web, suggests 5 branches to explore, one-click ingest into the Cortex.
Daily Intelligence Briefing: Scheduled task searches the web daily for your top 10 interests, AI analyzes findings with personalized impact assessment and action items, ingests significant discoveries into the Cortex, emails you a curated digest.
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 needed. The system auto-classifies every message and chooses the right level of effort.
Frontend: React 19, Zustand 5, Tailwind CSS 4, Vite 6, Capacitor (Android)
Backend: Express, WebSocket, TypeScript 5.7, node-pty
AI: Claude Code (Opus/Sonnet/Haiku), Unified llm() layer (Gemini, OpenAI, Anthropic, DeepSeek, Ollama, OpenRouter)
Mobile: Capacitor Android APK + PWA (offline-capable)
Tools: 99 registered tools across 14 families
Project structure
src/ React frontend
├── cards/ Card renderers (DynamicUI, Terminal, Shell, Orchestration, Mission)
├── components/ Timeline, ChatInput, CardContainer, ConnectionPicker
├── store/chat.ts Zustand state (cards, streaming, connections, evolution)
└── lib/ WS client, JSX sandbox (Sucrase), EnsoUI (27 components), connections
server/ Backend
├── apps/ 16 shipped app packages incl. data source apps (kindle, youtube, browser, etc.)
└── src/
├── server.ts Express + WS server with auth
├── standalone-agent.ts Chat agent for standalone mode
├── task-router.ts 4-tier message classifier (simple/research/one-off/orchestrated)
├── orchestrator.ts Multi-agent planner (goal → task DAG)
├── orchestrator-engine.ts DAG executor with parallel agents
├── evolution.ts Self-evolution sprint system (7 phases, 6 parallel agents)
├── researcher-tools.ts Two-phase streaming research pipeline
├── wiki-tools.ts Knowledge Cortex engine (ingest, search, lint, import) — the ONLY brain
├── llm.ts Unified LLM layer — single llm() call across 6 providers
├── claude-code.ts Claude Code CLI integration (NDJSON streaming)
├── build-via-claude.ts Natural language → app build via Claude Code
├── scheduled-tasks.ts Durable cron system with task execution
├── settings-transfer.ts Cross-machine export/import (9 categories)
├── user-context-tools.ts Consent-gated desktop environment scanner
└── *-tools.ts 14 tool families (99 total tools)
shared/types.ts WebSocket protocol types (shared frontend ↔ backend)
| 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 |
We welcome contributions of all sizes — bug fixes, new app types, UI improvements, docs, translations. See CONTRIBUTING.md for guidelines.
MIT License — free forever.