NoCodeAPI

Clawdbot(Moltbot): The Complete Guide to the AI Personal Assistant

Last Updated: January 28, 2026

Clawdbot (officially rebranded to Moltbot on January 27, 2026) is an open-source, self-hosted AI personal assistant that has taken the developer community by storm. Created by Peter Steinberger, founder of PSPDFKit (now Nutrient), Moltbot has garnered over 60,000 GitHub stars in just weeks, making it one of the fastest-growing open-source projects in recent memory.

Unlike traditional AI assistants like Siri or ChatGPT that wait passively for your commands or forget you between sessions, Moltbot runs on your own hardware, remembers everything about you, and can proactively reach out with reminders, briefings, and alerts. This is what makes it genuinely revolutionary in the personal AI assistant space.

“At this point I don’t even know what to call Moltbot. It is something new. After a few weeks in with it, this is the first time I have felt like I am living in the future since the launch of ChatGPT.” — @davemorin

Table of Contents

What Is Clawdbot/Moltbot?
Why the Name Changed from Clawdbot to Moltbot
Key Features and Capabilities
How to Install Moltbot
Practical Use Cases with Examples
- Email Management
- Support Automation
- System Monitoring
- Calendar and Scheduling
Cloud Deployment Options
Security Considerations
Moltbot vs Traditional AI Assistants
Community and Resources
Frequently Asked Questions

What Is Clawdbot/Moltbot?

Moltbot is a locally-hosted AI agent that transforms large language models like Claude or GPT-4 into a true 24/7 personal assistant. The project launched in January 2026 and quickly captured the imagination of developers worldwide who have been waiting for a personal AI assistant that actually works since Siri launched in 2011.

At its core, Moltbot operates on three fundamental principles that set it apart from every other AI assistant on the market.

Persistent Memory allows Moltbot to remember your conversations, preferences, and context across sessions. Your data is stored locally as Markdown files in Obsidian format, meaning your assistant learns and adapts over time rather than resetting every day. The system maintains conversation context for days, weeks, or even months.

Proactive Assistance enables Moltbot to reach out to you first through its Heartbeat Engine and cron job integration. Instead of waiting for prompts, it can send morning briefings, traffic-based reminders, health alerts from connected wearables, and notifications when something you care about happens.

Real-World Task Execution gives Moltbot the ability to actually do things for you. With proper permissions, it can manage emails, control smart home devices, execute terminal commands, browse the web, and automate complex workflows.

Why the Name Changed from Clawdbot to Moltbot {#why-the-name-changed}

On January 27, 2026, the project officially rebranded from Clawdbot to Moltbot following a trademark request from Anthropic regarding potential confusion with its Claude AI model. Peter Steinberger, the creator, revealed on X that Anthropic “forced” him to change the branding for copyright reasons.

The new name “Moltbot” represents the concept of “molting” — the process by which crustaceans shed their shells as they grow — symbolizing the project’s continuously evolving nature. The project’s beloved lobster mascot (🦞), originally named “Clawd,” is now called “Molty.”

Aspect	Old Name	New Name
GitHub Repository	clawdbot/clawdbot	moltbot/moltbot
NPM Package	clawdbot	moltbot
Official Website	clawd.bot	molt.bot
Documentation	docs.clawd.bot	docs.molt.bot
Command Line	`clawdbot`	`moltbot` (legacy command still works)

The legacy clawdbot command remains available as a compatibility shim, so existing installations continue to work.

Key Features and Capabilities

Multi-Platform Integration

Moltbot connects to virtually every messaging platform you use. You maintain the same conversation with the same memory regardless of which platform you access it from.

Platform	Status	Integration Type
WhatsApp	✅ Native	Baileys (WhatsApp Web protocol)
Telegram	✅ Native	Bot API via grammY
Discord	✅ Native	Bot API via discord.js
Slack	✅ Native	Bolt framework
Signal	✅ Native	signal-cli
iMessage	✅ Native	imsg CLI (macOS)
Google Chat	✅ Native	Chat API
Microsoft Teams	✅ Extension	Bot Framework
Matrix	✅ Extension	Matrix SDK
BlueBubbles	✅ Extension	BlueBubbles API

Full Computer Access

Moltbot can execute terminal commands, write and run scripts, browse the web via CDP (Chrome DevTools Protocol), control smart home devices like Philips Hue lights and air purifiers, and access your file system with the permissions you grant.

Extensible Skills System

The platform supports bundled, managed, and custom workspace skills through ClawdHub (now Moltbot Hub). Users can create new capabilities simply by asking Moltbot to build them. The community has developed over 100 skills for tasks ranging from RSS parsing to health metric tracking.

Proactive Automation Engine

The Heartbeat Engine combined with cron job integration allows Moltbot to wake itself up, check conditions, and message you when thresholds are met. Instead of asking “Is the server down?” or “Did the stock price drop?”, Moltbot monitors and alerts you automatically.

Multi-Agent Architecture

Moltbot supports agent-to-agent communication via sessions_* tools, enabling complex workflows where multiple AI agents collaborate on tasks. One agent can monitor your email while another manages your calendar and a third coordinates between them.

How to Install Moltbot

Setting up Moltbot requires some technical knowledge but can be completed in under 30 minutes.

System Requirements

Node.js version 22 or higher
macOS, Linux, or Windows (via WSL2 — native Windows is not supported)
Claude Pro/Max subscription or Anthropic/OpenAI API key (or local LLM via Ollama)

Quick Installation Steps

Step 1: Install Node.js

# Using nvm (recommended)
curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.0/install.sh | bash
source ~/.bashrc  # or ~/.zshrc
nvm install 22
nvm use 22

Step 2: Install Moltbot CLI

npm install -g moltbot@latest
# Or using the legacy package name (redirects automatically)
npm install -g clawdbot@latest

Step 3: Run the Onboarding Wizard

moltbot onboard --install-daemon

The wizard will guide you through:

Choosing local vs. remote gateway deployment
Configuring authentication (Anthropic API key recommended)
Selecting messaging channels to enable
Installing the background service (launchd/systemd)

Step 4: Connect Messaging Platforms

For WhatsApp:

moltbot channels login
# Scan the QR code via WhatsApp → Settings → Linked Devices

For Telegram, provide your bot token (obtained from BotFather):

{
  "channels": {
    "telegram": {
      "botToken": "123456:ABCDEF"
    }
  }
}

Step 5: Verify Installation

moltbot health
moltbot status --all

You should see output confirming:

Gateway listening on ws://127.0.0.1:18789
Browser control active
Model configured (e.g., anthropic/claude-opus-4-5)
Heartbeat started

Practical Use Cases with Examples

Email Management {#email-management}

Moltbot excels at automating email workflows through Gmail Pub/Sub integration and the gog skill for Google Workspace.

Setting Up Gmail Automation

First, install the Gmail skill:

moltbot skills install gog

Configure Gmail Pub/Sub triggers to enable real-time email monitoring. This allows Moltbot to react to incoming emails instantly rather than polling.

Example: Morning Email Rollup

Set up a cron job to receive a daily email summary at 8 AM:

moltbot cron add \
  --name "Morning status" \
  --cron "0 7 * * *" \
  --tz "America/Los_Angeles" \
  --session isolated \
  --message "Summarize inbox + calendar for today." \
  --deliver \
  --channel whatsapp \
  --to "+15551234567"

This creates a recurring job that runs at 7 AM Pacific, summarizes your inbox and calendar, and delivers the briefing to your WhatsApp.

Example: Inbox Zero Automation

Simply message Moltbot via Telegram or WhatsApp:

“Clear my inbox of newsletters and promotional emails. Unsubscribe from anything I haven’t opened in 3 months. Archive anything older than 30 days that I’ve read. Prioritize and flag anything from my VIP contacts.”

Moltbot will use browser automation to navigate Gmail, process emails, and report back with a summary of actions taken.

Example: Automated Email Response Drafting

“When I receive emails from clients asking about project status, draft a professional reply that references our last meeting and any relevant documents in my Google Drive.”

Moltbot uses multi-agent coordination to check email content, search your Drive, and compose contextual responses.

Real User Result: One user achieved Inbox Zero through Moltbot, which “managed bulk unsubscriptions and prioritized important messages.” Another transcribed over 1,000 WhatsApp voice messages and archived them with semantic search.

Support Automation {#support-automation}

Moltbot can function as a Slack-based auto-support system that handles up to 70% of incoming tickets autonomously.

Setting Up Slack Support Bot

Configure Slack integration with channel-specific settings:

{
  "channels": {
    "slack": {
      "enabled": true,
      "botToken": "xoxb-...",
      "appToken": "xapp-...",
      "channels": {
        "#support": {
          "allow": true,
          "requireMention": false,
          "systemPrompt": "You are a helpful support agent. Answer questions about our product using the knowledge base. Escalate complex issues to human agents by tagging @support-team."
        }
      }
    }
  }
}

Example: Auto-Reply to Common Questions

Create a dedicated support agent with specialized skills:

moltbot skills install confluence  # For knowledge base access
moltbot skills install slack       # For Slack actions

Configure the agent to monitor the #support channel and respond to common queries using your Confluence documentation.

Example: Production Bug Auto-Fix

User @henrymascot documented this real-world case:

“Set up Moltbot as a Slack auto-support system. One night, the bot detected a production bug and fixed it on its own.”

Set up webhook integration with your error monitoring service (Sentry, PagerDuty):

moltbot cron add \
  --name "Error monitor" \
  --cron "*/5 * * * *" \
  --session isolated \
  --message "Check Sentry for new errors. If critical bugs found, analyze the stack trace, propose a fix, and create a PR if the fix is straightforward." \
  --deliver \
  --channel slack \
  --to "#engineering-alerts"

Example: Ticket Triage and Routing

“Monitor #customer-support for new messages. Categorize them as billing, technical, or feature-request. For billing issues, check the customer’s account status. For technical issues, search our docs and provide an initial response. For complex issues, create a Jira ticket and notify the appropriate team.”

Moltbot can handle routine inquiries automatically while escalating complex issues to humans, reducing support team workload by up to 70%.

System Monitoring {#system-monitoring}

Moltbot’s Heartbeat Engine and cron integration enable 24/7 proactive system monitoring with instant alerts.

Setting Up Heartbeat Monitoring

The heartbeat mechanism allows Moltbot to wake itself up periodically and check conditions:

{
  "heartbeat": {
    "enabled": true,
    "interval": "30m",
    "duplicateSuppression": "24h"
  }
}

Example: Server Health Checks

Create a recurring monitoring job:

moltbot cron add \
  --name "Server health check" \
  --cron "*/15 * * * *" \
  --session isolated \
  --message "Check server status: run 'df -h' for disk space, 'free -m' for memory, 'uptime' for load average. Alert me if disk > 85%, memory > 90%, or load > 4.0." \
  --deliver \
  --channel telegram \
  --to "@yourusername"

Example: API Availability Monitoring

moltbot cron add \
  --name "API health" \
  --cron "*/5 * * * *" \
  --session isolated \
  --message "Ping our production API endpoints. If any return non-200 status or response time > 2s, alert immediately with the endpoint and error details."

Example: Log File Monitoring

“Monitor /var/log/application.log for any ERROR or CRITICAL entries. When found, extract the relevant context, identify the likely cause, and send me a summary via WhatsApp. If it’s the same error I’ve seen before, just note the recurrence count.”

Example: Resource Threshold Alerts

Moltbot can proactively reach out when conditions change:

“Watch CPU temperature via sensors command. If it exceeds 80°C, check what processes are consuming resources and send me a report. If it’s a runaway process I’ve marked as safe to kill, terminate it automatically.”

Real User Benefit: Instead of manually checking dashboards, users receive instant alerts to their messaging apps. The 24-hour duplicate suppression prevents alert fatigue while ensuring critical issues are surfaced.

Calendar and Scheduling {#calendar-and-scheduling}

Moltbot integrates deeply with calendar systems through the gog skill (Google Workspace) and browser automation.

Setting Up Calendar Integration

moltbot skills install gog  # Includes Gmail, Calendar, Drive, Contacts

Example: Daily Schedule Briefing

moltbot cron add \
  --name "Morning briefing" \
  --cron "0 6 * * *" \
  --tz "Europe/London" \
  --session isolated \
  --message "Good morning! Give me my day: weather forecast, calendar events, tasks due today, and top 3 priority emails. Suggest optimal times for focused work based on my meeting gaps." \
  --deliver \
  --channel whatsapp \
  --to "+44..."

Example: Smart Meeting Scheduling

Message Moltbot via any connected platform:

“Schedule a 1-hour meeting with [email protected] and [email protected] next week. Find a time that works for everyone based on their calendar availability. The topic is Q1 Planning. Send calendar invites once confirmed.”

Moltbot will:

Check your calendar for availability
Check invited participants’ free/busy status (if accessible)
Propose optimal time slots
Create the calendar event with proper details
Send calendar invitations

Example: Travel-Aware Scheduling

“I have a client meeting in downtown at 2 PM. Based on current traffic from my home, when should I leave? Add a reminder 15 minutes before I need to leave.”

Real user @davemorin uses Moltbot for “traffic-based departure alerts” — receiving proactive reminders like “Leave now for pickleball based on current traffic.”

Example: Meeting Preparation

moltbot cron add \
  --name "Meeting prep" \
  --cron "0 8 * * *" \
  --session main \
  --system-event "Check my calendar for meetings today. For each meeting, summarize relevant email threads, pull any related documents from Drive, and list action items from previous meetings with these participants."

Example: Rescheduling Automation

“Cancel all my meetings tomorrow and reschedule them to next week with the same participants. Send apologetic notifications explaining I’m unwell.”

Cloud Deployment Options

While many users run Moltbot on local machines, cloud deployment offers always-on availability.

DigitalOcean droplets work well starting at $12/month for the recommended 2GB RAM tier. Nader Dabit’s community guide provides step-by-step instructions.

Railway offers one-click deployment with usage-based pricing typically between $5-20/month. See the official Railway deployment guide.

Render provides a generous free tier suitable for testing. Documentation available at docs.molt.bot/render.

Raspberry Pi deployment is popular for low-power, always-on setups at home.

Hetzner offers cost-effective cloud servers in Germany with excellent performance.

Security Considerations

Because Moltbot has full system access, security is critical.

Default Security Model

Session Type	Execution Environment	Risk Level
Main (You)	Host machine	⚠️ High trust
Groups	Docker sandbox (optional)	✅ Isolated
Unknown DMs	Pairing required	✅ Protected

The Pairing System

Moltbot implements DM pairing to prevent unauthorized access. Unknown senders receive a short pairing code, and messages are not processed until approved:

moltbot pairing approve whatsapp 123456

Security Best Practices

Do:

Use DM pairing for unknown senders (dmPolicy: "pairing")
Run group sessions in Docker sandboxes
Set up Tailscale for secure remote access
Regularly run moltbot doctor to check configuration
Keep allowlists updated

Don’t:

Expose port 18789 to the public internet
Grant shell access without understanding risks
Install unverified skills from unknown sources
Use dmPolicy: "open" without allowlists

Prompt Injection Risk

As TechCrunch reported, investor Rahul Sood warns about “prompt injection through content” — where malicious messages could lead Moltbot to take unintended actions. Running Moltbot in a sandbox or on a separate machine with throwaway accounts mitigates this risk.

Moltbot vs Traditional AI Assistants

Feature	Moltbot	ChatGPT	Siri	Google Assistant
Self-hosted	✅ Yes	❌ No	❌ No	❌ No
Proactive messaging	✅ Yes	❌ No	⚠️ Limited	⚠️ Limited
Multi-channel	✅ 10+ platforms	❌ Web/app only	❌ Apple only	❌ Google only
Browser control	✅ Full CDP	❌ No	❌ No	❌ No
Custom skills	✅ Unlimited	⚠️ GPTs only	❌ No	⚠️ Actions
Local LLM support	✅ Ollama/LM Studio	❌ No	❌ No	❌ No
Persistent memory	✅ Weeks/months	⚠️ Limited	❌ No	⚠️ Limited
Open source	✅ MIT License	❌ Proprietary	❌ Proprietary	❌ Proprietary
Privacy	✅ Full control	❌ Cloud-based	⚠️ Apple servers	❌ Google servers

Community and Resources

Project Statistics (January 2026)

GitHub Stars: 60,000+
Discord Members: 8,900+
Contributors: 130+
Skills on ClawdHub: 100+

Official Resources

Website: molt.bot (formerly clawd.bot)
Documentation: docs.molt.bot
GitHub: github.com/moltbot/moltbot
Discord: discord.gg/clawd
Skills Marketplace: ClawdHub
Showcase: docs.molt.bot/start/showcase

Key Documentation Pages

Recommended Reading

Frequently Asked Questions

Is Moltbot free?

Moltbot itself is free and open-source (MIT License). However, you need to pay for AI model API costs (Anthropic Claude API or OpenAI), which can range from $25-150/month depending on usage. Claude Max subscriptions offer unlimited usage within limits and work with Moltbot via OAuth.

Can I run Moltbot with local LLMs?

Yes. Moltbot supports Ollama, LM Studio, Harbor, and any OpenAI-compatible API endpoint. Best local model: GLM-4.7-Flash for tool calling capabilities.

How much technical knowledge do I need?

The onboarding wizard (moltbot onboard) makes basic setup accessible. However, advanced features like multi-agent routing, custom skills, and security hardening require intermediate technical knowledge.

Is it safe to give Moltbot shell access?

The developer describes it as “spicy.” Use Docker sandboxing for non-main sessions, carefully review commands, start with read-only tools, and never expose Moltbot to untrusted users without sandboxing.

Can Moltbot really negotiate car purchases?

Yes. User AJ Stuyvenberg documented saving $4,200 on a $56,000 car. Moltbot searched Reddit for pricing data, contacted dealers via email, and negotiated using hardball tactics.

What’s the difference between cron jobs and heartbeat?

Cron jobs run at specific scheduled times (like Unix cron). The heartbeat mechanism runs at regular intervals and can be triggered immediately. Use cron for scheduled tasks and heartbeat for continuous monitoring with wake-on-demand.

Conclusion

Moltbot represents a significant step toward the AI assistant we have been promised for over a decade. By running locally with persistent memory and proactive capabilities, it delivers personalization and functionality that cloud-based alternatives simply cannot match.

The tradeoff is complexity and responsibility. This is not a consumer-ready product that anyone can install without technical knowledge. And as TechCrunch noted, “things could turn ugly fast” if users approach it as carelessly as ChatGPT.

But for developers and power users willing to invest the setup time and maintain security best practices, Moltbot offers a glimpse of what personal AI assistance will eventually become for everyone.

“It will actually be the thing that nukes a ton of startups, not ChatGPT as people meme about. The fact that it’s hackable and hostable on-prem will make sure tech like this DOMINATES conventional SaaS.” — @rovensky

References and Further Reading

This comprehensive guide was compiled from official documentation, TechCrunch reporting, community resources, and hands-on user experiences to provide accurate, actionable information about Moltbot for prospective users. Last updated January 28, 2026.

More tutorials