Documentation

Everything you need to deploy and manage your AI assistant on ClawTank.

Getting Started

ClawTank deploys a personal OpenClaw AI assistant in under 1 minute. No Docker, no SSH, no terminal required.

Step 1: Choose Your AI Model

Select the AI model that powers your assistant. Currently available:

  • Claude Opus 4.6 — Anthropic's most capable model

More models (GPT-4.1, Gemini 2.5 Flash) coming soon.

Step 2: Connect Telegram

Click Telegram to open the setup guide. You'll need to create a Telegram bot:

  1. Open @BotFather in Telegram
  2. Send /newbot and follow the prompts
  3. Copy the bot token (looks like 123456:ABC-DEF...)
  4. Paste it into the token field

You can skip this step and set it up later from the dashboard.

Step 3: Sign In & Deploy

Click Sign in with Google to create your account. Your OpenClaw instance will be provisioned automatically — you'll see a progress screen with build steps completing in real-time.


Dashboard

After deployment, the dashboard is your control center.

Status & Trial

The top of the dashboard shows your container status:

  • Running — your assistant is online and ready
  • Stopped — manually stopped or awaiting restart
  • Creating — build in progress
  • Error — something went wrong during provisioning

Your free trial lasts 7 days. The remaining time is displayed as a countdown badge.

Build Progress

While your container is being created, you'll see a live progress tracker:

  1. Entrypoint started
  2. Setting up OpenClaw
  3. API keys configured
  4. Starting gateway
  5. Finalizing

This typically takes 30–60 seconds.

Gateway Connection

Once running, connect to your assistant's web interface:

  1. Copy the Gateway Token — click the copy button next to the token
  2. Open Gateway UI — click the link to open your assistant's web interface in a new tab
  3. Approve Devices — when your browser connects for the first time, come back to the dashboard and click Approve Devices to authorize the connection

Telegram Pairing

To connect your Telegram bot to your assistant:

  1. Send any message to your bot (in a group or DM)
  2. The bot will reply with a pairing code (6–8 characters)
  3. Enter the code on the dashboard and click Pair

Once paired, you can chat with your AI assistant directly in Telegram.

Container Actions

  • Stop — pause your instance (preserves data)
  • Start — resume a stopped instance
  • Restart — stop and start again
  • Rebuild — recreate the container from scratch (use if something breaks)
  • Delete — permanently remove your instance and all data

Resource Stats

The dashboard displays real-time resource usage:

  • CPU Usage — percentage with progress bar
  • Memory — current usage vs. allocated limit

Memory

The Memory page shows an interactive knowledge graph of everything your AI assistant has learned from conversations.

  • Memories are automatically extracted from your chats
  • Related facts are connected as nodes in a graph
  • You can view and manage all stored memories

The graph updates every 30 seconds. If you don't see any memories yet, start chatting with your assistant — it builds its knowledge over time.


Logs

The Logs page provides a live terminal view of your container output.

  • Live streaming — logs appear in real-time via server-sent events
  • Filter — search logs with real-time highlighting and match count
  • Auto-scroll — automatically follows new output (disables when you scroll up)
  • Download — export filtered logs as a .log file
  • Clear — remove all displayed logs

Up to 5,000 lines are kept in view. The connection reconnects automatically if interrupted.


Settings

Domain

Gateway Subdomain Your assistant is accessible at your-name.onlymake.ai. You can change the subdomain:

  • Only lowercase letters, numbers, and hyphens allowed
  • Changes take effect immediately
  • You'll need to reconnect to the Gateway UI after changing

Custom Domain Point your own domain to your assistant:

  1. Enter your domain (e.g., chat.yourdomain.com)
  2. Add a DNS A record pointing to 108.61.182.112
  3. TLS certificate is provisioned automatically via Let's Encrypt

API Keys

Configure the services your assistant can use:

Key Purpose
Anthropic API Key Powers the AI model (required)
Telegram Bot Token Connects your Telegram bot
Telegram Group ID Links a specific Telegram group
OpenAI API Key Enables multimodal capabilities
Gemini API Key Alternative AI model support
ElevenLabs API Key Voice synthesis
OpenClaw Model Override the default model

The Telegram Group ID can be auto-detected — click the detect button if your bot is already in a group.

After changing API keys, a rebuild is required for changes to take effect.

Password

If you signed up with email/password (not Google), you can change your password from this section. Minimum 8 characters.

Ready to get started?

Deploy your AI assistant in under 1 minute.

Deploy now