A utility dashboard for OpenClaw - view and understand your agent logs in a friendly way.
- Log Viewer: View OpenClaw logs in a clean, human-readable format
- Groups repetitive messages (like polling/heartbeat) to reduce noise
- Color-coded by subsystem (WhatsApp, Agent, Memory, etc.)
- Filter by log level (Error, Warn, Info, Debug)
- Auto-refresh with live indicator
- Click to expand and see raw data
- Configurable: Change log directory and file path to match your setup
- Multi-user: Default settings work for typical OpenClaw installs, but can be customized
# Install dependencies
npm install
# Start development server (frontend + backend)
npm run devThen open:
- Frontend: http://localhost:5173
- Backend API: http://localhost:3001
By default, the app looks for logs in:
/tmp/openclaw/openclaw-YYYY-MM-DD.log
You can change this in the Settings panel (click the Settings button).
If your OpenClaw logs are in a different location:
- Click Settings in the top-right
- Update the Log Directory field
- Select the log file from the dropdown
# Build the frontend
npm run build
# Run the production server (serves both API and static files)
npm startGET /api/logs- Get parsed logs- Query params:
dir,file,limit,level
- Query params:
GET /api/logs/files- List available log files- Query params:
dir
- Query params:
GET /api/settings- Get default settings
The app parses OpenClaw's JSON log format and extracts:
- Time: When the event occurred
- Subsystem: Which component generated the log (whatsapp, agent, memory, etc.)
- Level: DEBUG, INFO, WARN, ERROR
- Message: Human-readable description of what happened
Repetitive messages (like memory polling or heartbeats) are grouped together to make the logs easier to read.
- Frontend: React + TypeScript + Vite
- Backend: Express.js
- Styling: Plain CSS (dark theme)
MIT