Skip to main content
NanoClaw Pro checks in on you — you don’t have to check in on it. Get your proactive AI assistant running with morning check-ins, afternoon follow-ups, and persistent memory in just a few steps.
1

Clone the repository

git clone https://github.com/andrewsiah/nanoclaw-pro.git
cd nanoclaw-pro
2

Open in Claude Code

claude
This will open the project in Claude Code’s interactive development environment.
3

Run setup

In Claude Code, run:
/setup
The setup will guide you through:
  • Installing Node.js 20+ (if needed)
  • Installing dependencies
  • Setting up Docker or Apple Container
  • Authenticating with Claude
  • Configuring your messaging channel
  • Starting the background service
4

Enable the proactive agent

After setup completes, activate the proactive check-ins:
/proactive-agent
This enables:
  • Morning check-ins (9am): Three questions about yesterday, today, and gratitude
  • Afternoon follow-ups (4pm): Surfaces items needing follow-up, helps with emails/messages
  • Dynamic reminders: Automatically creates reminders when you mention time-sensitive items
  • Persistent memory: Builds knowledge about your projects, preferences, and life using qmd
5

Send your first message

In your registered chat (WhatsApp, Telegram, Discord, Slack, or Gmail), send:
@Andy hello!
Replace @Andy with your assistant’s name if you customized it during setup.Your assistant will respond, and you can start having conversations. Try:
  • @Andy remember I prefer dark mode (saves to memory)
  • @Andy search for AI news from Hacker News
  • @Andy every Monday at 9am, send me a weekly planning prompt

What’s Next?

Installation Guide

Detailed setup instructions and troubleshooting

First Conversation

Learn how to interact with your assistant

Customization

Customize behavior, add channels, and tailor to your needs

Scheduled Tasks

Set up recurring tasks and reminders

Key Features

  • Proactive check-ins: Built-in morning and afternoon prompts
  • Persistent memory: Remembers context across all conversations using qmd
  • Dynamic reminders: Automatically catches time-sensitive mentions
  • Container isolation: Safe Bash access - all agents run in isolated Linux containers
  • Multi-channel: WhatsApp, Telegram, Discord, Slack, or Gmail support
  • Scheduled tasks: Recurring jobs with full agent capabilities
  • Agent swarms: First personal assistant to support collaborative agent teams

Build docs developers (and LLMs) love