📝 Memory Viewer for OpenClaw

A beautiful, dark-themed web UI for browsing and editing OpenClaw AI agent memory files. Built specifically for OpenClaw agents that store context in Markdown files.

Why Memory Viewer for OpenClaw?

OpenClaw agents store their memory in Markdown files (MEMORY.md, memory/*.md). Memory Viewer provides a dedicated web interface to:

• Browse memory files in a collapsible tree
• 搜索 across all agent memories instantly
• Edit files directly in the browser
• Monitor agent system status and memory usage
• Connect to multiple OpenClaw agents from a single UI

Features

• 📁 File Tree Sidebar — Navigate all .md files in a collapsible tree
• 📖 Markdown Rendering — GitHub-flavored Markdown with syntax highlighting, tables, and more
• ✏️ In-Browser Editing — Edit files directly with Ctrl+S to save, with optimistic locking for conflict detection
• 🔍 Full-Text 搜索 — 搜索 across all memory files instantly (Ctrl+K)
• 📊 System 仪表盘 — Server uptime, memory usage, load averages, and today's memory summary
• 🔄 Live Reload — Files auto-refresh when changed on disk (via WebSocket), with 10s polling fallback
• 📱 PWA Support — Installable as a standalone app with offline caching
• 🔗 Deep Linking — Hash-based routing (#/file/path) for bookmarkable file URLs
• 📊 Mermaid Diagrams — Render flowcharts and diagrams from fenced code blocks
• 🚗 Large Screen Optimized — Touch-friendly UI for car displays (Tesla) and large screens
• 🌙 Dark/Light Theme — Toggle between themes, designed for always-on dashboards
• 📱 Responsive — Works on mobile with a slide-out sidebar
• 🌐 Multi-bot Connections — Connect to multiple OpenClaw agent workspaces from a single UI

Quick Start

# Clone
git clone https://github.com/silicondawn/memory-viewer.git
cd memory-viewer

# Install
npm install

# Run (starts both API server and Vite dev server)
npm run dev

Then open http://localhost:5173 in your browser.

OpenClaw Integration

Memory Viewer works seamlessly with OpenClaw agents. To connect to your OpenClaw agent:

1. Make sure your OpenClaw agent is running and accessible
2. In Memory Viewer, click the network icon in the top-right
3. Add your agent's workspace path (e.g., /home/user/clawd)
4. Start browsing and editing your agent's memory files

Deployment

Memory Viewer can be deployed as a standalone service:

# Build for production
npm run build

# Start production server
npm start

The server runs on port 8901 by default. You can expose it via Cloudflare Tunnel, Nginx, or any reverse proxy.

Docker Deployment

Quick Start (Pre-built Image)

Use the pre-built image from GitHub Container Registry:

# Run directly with docker
docker run -d \
  -p 8901:8901 \
  -v ~/.openclaw/workspace:/app/workspace:ro \
  --name memory-viewer \
  ghcr.io/silicondawn/memory-viewer:latest

# Or use docker-compose
docker-compose up -d

Open http://localhost:8901 in your browser.

Build from Source

# Clone the repository
git clone https://github.com/silicondawn/memory-viewer.git
cd memory-viewer

# Build and run
docker-compose up -d --build

Docker Configuration

Variable	Default	Description
PORT	8901	Container port (fixed in image)
WORKSPACE_DIR	/app/workspace	Directory inside container for .md files
STATIC_DIR	/app/dist	Built frontend assets

Customizing the Mount Path

Edit docker-compose.yml to point to your actual OpenClaw workspace:

volumes:
  - ~/.openclaw/workspace:/app/workspace:ro
  # Windows: C:/Users/YourName/.openclaw/workspace:/app/workspace:ro

The :ro flag mounts the directory as read-only (recommended for safety).

Building Custom Images

Use the provided build script:

# Build with default tag (latest)
./scripts/build-docker.sh

# Build with specific tag
./scripts/build-docker.sh v1.2.0

# Build and push to registry
PUSH=true ./scripts/build-docker.sh v1.2.0

License

MIT © Silicon Dawn