Skip to main content
Most AI assistants wait for you to have an idea and prompt them. NanoClaw Pro flips that. NanoClaw Pro is a fork of NanoClaw with two things added by default: a proactive agent that reaches out to you every morning and afternoon, and persistent memory via qmd so it gets smarter over time.

​
Key Differentiators

Proactive Check-ins

Your assistant checks in with you twice daily—morning and afternoon—to help you plan, follow up, and stay on track.

Persistent Memory

Uses qmd for local semantic search. Every conversation builds a knowledge base about your projects, preferences, and life.

Dynamic Reminders

Mention something time-sensitive and it automatically creates a reminder. No manual setup required.

Same Core Philosophy

Everything else is identical to NanoClaw—same container isolation, same channel support, same security model.

​
Proactive Check-ins

NanoClaw Pro reaches out to you automatically, built in by default.

​
Morning Check-in (9am)

Every morning at 9am, you’ll receive: 
gm! three questions:
1. what did you work on yesterday?
2. what are you working on today?
3. what's one thing you're grateful for?
This helps you:
  • Reflect on yesterday’s progress
  • Set intentions for today
  • Start the day with gratitude

​
Afternoon Check-in (4pm)

Every afternoon at 4pm, your assistant:
  • Surfaces anything you mentioned that needs follow-up
  • Offers to write emails or messages
  • Queues tasks for tomorrow
The check-in times are customizable. You can adjust them to match your schedule by modifying the proactive agent skill.

​
Persistent Memory System

NanoClaw Pro uses qmd—a local semantic search engine—to remember context across sessions.

​
How It Works

  1. Every conversation is indexed: As you chat, qmd builds a searchable knowledge base
  2. Semantic search before check-ins: Before each check-in, your assistant searches past conversations for relevant context
  3. Personalized interactions: Questions and suggestions are tailored based on what it knows about your projects, preferences, and life

​
What Gets Remembered

Details about your work projects, codebases, and technical preferences
Communication style, work hours, tools you use, and how you like to work
Birthdays, appointments, deadlines, and other time-sensitive information
Past discussions, decisions made, and context from previous sessions

​
Privacy

All memory is stored locally on your machine. Nothing is sent to external servers except when the assistant needs to respond using Claude’s API.

​
Dynamic Reminders

Just mention something time-sensitive in conversation:
“My friend’s birthday is Monday”
NanoClaw Pro automatically:
  • Extracts the time-sensitive information
  • Creates a reminder for the appropriate time
  • Follows up when the time comes
No manual reminder creation needed. Just talk naturally.

​
What’s the Same

Everything else about NanoClaw remains unchanged:
  • Container isolation: Agents run in Linux containers with filesystem isolation
  • Multi-channel support: WhatsApp, Telegram, Discord, Slack, Gmail
  • Scheduled tasks: Recurring jobs that run Claude and message you back
  • Security model: OS-level isolation, not just application-level permissions
  • Skills-based architecture: Customize by adding skills, not by configuring options
  • AI-native approach: No config files, just modify the code

Learn More About NanoClaw

Read about the core NanoClaw philosophy and architecture

​
Quick Start

Ready to try NanoClaw Pro? Here’s how to get started: 
git clone https://github.com/andrewsiah/nanoclaw-pro.git
cd nanoclaw-pro
claude
Then run /setup to configure. After setup, apply the proactive agent: 
/proactive-agent
That’s it. The proactive check-ins and memory system will be live.

Build docs developers (and LLMs) love

Get started for freeTalk to us