How to make two agents work if one is offline? | Friends of the Crustacean 🦞🤝 | Page 1

spare totem Jan 27, 2026, 8:32 AM

#

I have added LM Studio agent and Clawdbot was working correctly, later, I added MiniMax with API key and subscription, but Clawdbot still asks for LM Studio agent first and then fails immediately after connection is dropped. I have set models.mode: merge when defining models, but still no luck.. What's up? Any ideas?

Maybe because I am messaging in the same chat (old session)? And I should reset my context or something?

light kettleBOT Jan 27, 2026, 8:32 AM

#

To help others find answers, you can mark your question as solved via Right click solution message -> Apps -> ✅ Mark Solution

lethal horizonBOT Jan 27, 2026, 8:34 AM

#

spare totem I have added LM Studio agent and Clawdbot was working correctly, later, I added ...

Based on the docs, I can see what's happening here. The issue is likely session stickiness combined with how model selection works.

How Model Selection Works

Clawdbot selects models in this order:

Primary model (agents.defaults.model.primary)
Fallbacks (agents.defaults.model.fallbacks)
Provider auth rotation within a provider

The models.mode: merge only controls whether models.json is merged with built-in providers — it doesn't control which model gets priority.

The Problem

Your old session has the LM Studio auth profile pinned. Clawdbot reuses the same profile until:

Session reset (/new or /reset)
Compaction completes
Profile enters cooldown

#

So even though you added MiniMax, Clawdbot still tries LM Studio first because that's what the session remembers.

Solutions

Option 1: Start a New Session (Quickest)

/new

This clears the pinned profile and lets Clawdbot re-evaluate available models.

Option 2: Configure Fallbacks

Set MiniMax as primary with LM Studio as fallback:

{
  "agents": {

#

    "defaults": {
      "model": {
        "primary": "minimax/MiniMax-M2.1",
        "fallbacks": ["lmstudio/your-model-here"]
      }
    }
  }
}

Option 3: Switch in Current Session

/model minimax/MiniMax-M2.1

This overrides the primary for just that session.

#

Check Your Config

Run this to see what's actually configured:

clawdbot models status

And verify agents.defaults.model.primary points to the right model. If it's still set to LM Studio, change it to MiniMax.

The docs say: "Session stickiness keeps provider caches warm" — which is exactly what's catching you here. A fresh session should solve it! 🦐

spare totem Jan 27, 2026, 8:44 AM

#

lethal horizon So even though you added MiniMax, Clawdbot still tries LM Studio first because t...

When entering /new I got "You are not authorized to use this command."

lethal horizonBOT Jan 27, 2026, 8:45 AM

#

spare totem When entering `/new` I got "You are not authorized to use this command."

Found it! The /new command requires authorization via access groups/allowlists. The docs say:

"Slash commands and directives are only honored for authorized senders. Authorization is derived from channel allowlists/pairing plus commands.useAccessGroups."

Solutions

Option 1: Ask the Owner to Add You to the Allowlist

For Discord group chats, you need to be explicitly allowlisted in channels.discord.guilds.<guildId>.users:

{
  "channels": {
    "discord": {
      "guilds": {
        "YOUR_GUILD_ID": {

#

          "users": ["704609146166575125"],  // Your Discord user ID
          "channels": {
            "1465625381490000016": { "allow": true }
          }
        }
      }
    }
  }
}

Then restart the gateway: clawdbot gateway restart

Option 2: Switch Models Directly (No `/new` Needed)

Since you can't reset the session, just switch to MiniMax directly:

#

/model minimax/MiniMax-M2.1

Or if you have an alias:

/model minimax

This overrides the primary model for your current session without needing /new.

Option 3: Have the Owner Disable Access Group Enforcement

If this is a personal instance, they can set in clawdbot.json:

#

{
  "commands": {
    "useAccessGroups": false
  }
}

Then restart: clawdbot gateway restart

Note: This makes /new and other commands available to everyone who can message the bot.

Quickest fix for you: Ask the owner to either (a) add your Discord ID to the guild's users allowlist, or (b) set commands.useAccessGroups: false. Then /new will work! 🦐

spare totem Jan 27, 2026, 8:46 AM

#

I am using Telegram. Can you give me Telegram specific config

#

@lethal horizon

lethal horizonBOT Jan 27, 2026, 8:48 AM

#

spare totem <@1457407575476801641>

The /new command requires authorization — you're not on the allowlist for that Discord server.

Quick Fixes

Option A: Switch Models Directly (Works Now)

/model minimax/MiniMax-M2.1

This changes your session's model without needing /new.

Option B: Get Added to the Allowlist

The owner needs to add your Discord ID (704609146166575125) to the guild config in clawdbot.json:

#

  "channels": {
    "discord": {
      "guilds": {
        "GUILD_ID_HERE": {
          "users": ["704609146166575125"]
        }
      }
    }
  }
}

Then run clawdbot gateway restart.

Option C: Disable Access Group Enforcement (Less Secure)

#

Owner sets in clawdbot.json:

{
  "commands": {
    "useAccessGroups": false
  }
}

Then clawdbot gateway restart.

Best option for you right now: Just use /model minimax/MiniMax-M2.1 and you're done! 🦐

haughty sable Jan 27, 2026, 1:38 PM

#

@lethal horizon installed on vps using antigravity and getting no output after wake up my friend prompt in setup

lethal horizonBOT Jan 27, 2026, 1:40 PM

#

haughty sable <@1457407575476801641> installed on vps using antigravity and getting no output ...

The "Wake up, my friend!" screen means the Gateway isn't reachable or authenticated yet. Let's troubleshoot:

Quick Fixes

1. Restart the Gateway

moltbot gateway restart

2. Check Gateway Status

moltbot gateway status
moltbot status
moltbot models status

3. Follow the Logs

#

moltbot logs --follow

Look for errors about authentication, connection, or model config.

4. Run Doctor

moltbot doctor

Common VPS Issues

Antigravity on VPS:

Ensure you ran the OAuth login: moltbot models auth login --provider google-antigravity --set-default
The Gateway needs to be reachable at the IP/hostname you're accessing

Remote Access:

#

If accessing the UI remotely, make sure Tailscale/SSH tunnel is connected
The UI must point at the correct Gateway URL

Firewall:

Ensure port 18789 is open on the VPS

What does moltbot gateway status show? And are you accessing the dashboard locally on the VPS or remotely? 🦐

haughty sable Jan 27, 2026, 1:42 PM

#

the vps terminal says this:

root@ClawdBot:~# moltbot gateway restart
moltbot: command not found
root@ClawdBot:~# moltbot gateway status
moltbot status
moltbot models status
moltbot: command not found
moltbot: command not found
moltbot: command not found
root@ClawdBot:~#

lethal horizonBOT Jan 27, 2026, 1:42 PM

#

haughty sable the vps terminal says this: root@ClawdBot:~# moltbot gateway restart moltbot: c...

The CLI isn't in your PATH. How did you install it?

Quick Fix

If you used the installer script:

# Find where it was installed
which node
npm config get prefix

# Or run directly via pnpm if you used the git method
cd /root/moltbot  # wherever you cloned it
pnpm moltbot gateway restart

If npm global install:

#

# Check if it's installed globally
npm list -g moltbot

# Add npm global bin to PATH
export PATH="$(npm config get prefix)/bin:$PATH"

# Or run directly
/root/.npm-global/bin/moltbot gateway restart

What to Share

Tell me:

How did you install? (installer script, npm, git clone?)

#

Where did you clone/install it?
Run ls -la ~/ to see what's in your home directory

The most common fix is:

source ~/.bashrc  # or ~/.bash_profile
moltbot gateway restart

Or find the binary:

find /root -name "moltbot" -type f 2>/dev/null

haughty sable Jan 27, 2026, 1:45 PM

#

root@ClawdBot:~# npm list -g moltbot
/usr/lib
└── (empty)

root@ClawdBot:~# export PATH="$(npm config get prefix)/bin:$PATH"
root@ClawdBot:~# /root/.npm-global/bin/moltbot gateway restart
-bash: /root/.npm-global/bin/moltbot: No such file or directory
root@ClawdBot:~# source ~/.bashrc # or ~/.bash_profile
moltbot gateway restart
moltbot: command not found
root@ClawdBot:~# find /root -name "moltbot" -type f 2>/dev/null
root@ClawdBot:~#

lethal horizonBOT Jan 27, 2026, 1:46 PM

#

haughty sable root@ClawdBot:~# npm list -g moltbot /usr/lib └── (empty) root@ClawdBot:~# expo...

You haven't installed Moltbot yet! You need to run the installer.

Install Now

curl -fsSL https://molt.bot/install.sh | bash

This will:

Install Moltbot
Set up the CLI
Create necessary directories

After Install

Then run the onboarding wizard:

#

moltbot onboard

Or if you want to skip the wizard and just start the gateway:

moltbot gateway

For Antgravity (Google Models)

After installation, you'll need to:

Enable the plugin: moltbot plugins enable google-antigravity-auth
Login: moltbot models auth login --provider google-antigravity --set-default

#

Run the installer first and let me know if you hit any errors! 🦐

haughty sable Jan 27, 2026, 1:52 PM

#

root@ClawdBot:~# curl -fsSL https://molt.bot/install.sh | bash
curl: (28) Failed to connect to molt.bot port 443 after 134441 ms: Couldn't connect to server

#

@lethal horizon

lethal horizonBOT Jan 27, 2026, 1:59 PM

#

haughty sable <@1457407575476801641>

Ping received! 🎉

How did the install go? Any errors?

If Install Succeeded

Run the onboarding wizard:

moltbot onboard

If You Want to Skip Onboarding and Configure Manually

Create a minimal config first:

mkdir -p ~/.clawdbot

#

cat > ~/.clawdbot/clawdbot.json << 'EOF'
{
  agents: {
    defaults: {
      workspace: "~/clawd"
    }
  },
  channels: {
    discord: {
      enabled: true,
      token: "YOUR_DISCORD_BOT_TOKEN"
    }
  }
}
EOF

#

Then start the gateway:

moltbot gateway

What's Next?

Do you have a Discord bot token? (Need one to enable Discord)
Which model do you want to use? (Antigravity, OpenAI, Anthropic, MiniMax?)
Are you running this on a VPS or local machine?

Let me know where you're at and I'll guide you through! 🦐

haughty sable Jan 27, 2026, 2:01 PM

#

no i had this error
root@ClawdBot:~# curl -fsSL https://molt.bot/install.sh | bash
curl: (28) Failed to connect to molt.bot port 443 after 134441 ms: Couldn't connect to server
root@ClawdBot:~# curl -fsSL https://molt.bot/install.sh | bash
curl: (28) Failed to connect to molt.bot port 443 after 134645 ms: Couldn't connect to server
root@ClawdBot:~#

lethal horizonBOT Jan 27, 2026, 2:01 PM

#

haughty sable no i had this error root@ClawdBot:~# curl -fsSL https://molt.bot/install.sh | b...

Network issue on your VPS - can't reach molt.bot. Let's try alternatives:

Option 1: Git Clone (Direct from GitHub)

# Install Node.js first if needed
apt update && apt install -y nodejs npm

# Clone and install from source
git clone https://github.com/moltbot/moltbot.git
cd moltbot
npm install
npm run build

# Create symlink so moltbot works
ln -sf $(pwd)/bin/moltbot.js /usr/local/bin/moltbot

#


# Or run directly
node bin/moltbot.js gateway

Option 2: Try the Clawdbot Installer (Same Project)

# Same thing, different domain
curl -fsSL https://clawd.bot/install.sh | bash

Option 3: npm Install Directly

#

npm install -g moltbot

Option 4: Check DNS/Network

# Test connectivity
ping -c 3 github.com
ping -c 3 npmjs.com

# Check DNS
nslookup github.com

#

Try Option 3 first (npm install) - it's simplest!

npm install -g moltbot
moltbot gateway

Let me know what happens! 🦐

haughty sable Jan 27, 2026, 2:16 PM

#

root@ClawdBot:~# bash
root@ClawdBot:~# npm install -g moltbot
npm warn EBADENGINE Unsupported engine {
npm warn EBADENGINE package: 'moltbot@0.1.0',
npm warn EBADENGINE required: { node: '>=24', pnpm: '>=10' },
npm warn EBADENGINE current: { node: 'v22.22.0', npm: '10.9.4' }
npm warn EBADENGINE }

added 1 package in 798ms
root@ClawdBot:~# moltbot onboard
moltbot: command not found
root@ClawdBot:~#

lethal horizonBOT Jan 27, 2026, 2:17 PM

#

haughty sable root@ClawdBot:~# bash root@ClawdBot:~# npm install -g moltbot npm warn EBADENGIN...

Package installed but not in PATH. Let's find it:

# Find where npm installed moltbot
npm config get prefix

# Run directly from there
$(npm config get prefix)/bin/moltbot onboard

# Or add to PATH permanently
echo 'export PATH="$(npm config get prefix)/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc
moltbot onboard

Quick One-Liner

#

# Run directly without modifying PATH
$(npm config get prefix)/bin/moltbot onboard

Let me know if that works! 🦐

haughty sable Jan 27, 2026, 2:22 PM

#

root@ClawdBot:~# npm config get prefix
/usr
root@ClawdBot:~# $(npm config get prefix)/bin/moltbot onboard
bash: /usr/bin/moltbot: No such file or directory
root@ClawdBot:~# echo 'export PATH="$(npm config get prefix)/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc
moltbot onboard
moltbot: command not found
root@ClawdBot:~# $(npm config get prefix)/bin/moltbot onboard
bash: /usr/bin/moltbot: No such file or directory
root@ClawdBot:~#

#

@lethal horizon it show this root@ClawdBot:~# npm config get prefix
/usr
root@ClawdBot:~# $(npm config get prefix)/bin/moltbot onboard
bash: /usr/bin/moltbot: No such file or directory
root@ClawdBot:~# echo 'export PATH="$(npm config get prefix)/bin:$PATH"' >> ~/.bashrc
source ~/.bashrc
moltbot onboard
moltbot: command not found
root@ClawdBot:~# $(npm config get prefix)/bin/moltbot onboard
bash: /usr/bin/moltbot: No such file or directory
root@ClawdBot:~#

lethal horizonBOT Jan 27, 2026, 2:39 PM

#

That doesn't look like onboarding output. Did you:

Run moltbot onboard and it responded with that?
Or message your bot and it replied with that?

If it's option 1 (onboarding), something went wrong.

If it's option 2 (bot replied to a message), your bot is working! But that's a weird response - did you send a message about nohl.com?

To Continue Onboarding

If onboarding didn't complete, run:

$(npm config get prefix)/bin/moltbot onboard

#

To Check Gateway Status

$(npm config get prefix)/bin/moltbot status

What were you trying to do when you saw that message? 🦐

lethal horizonBOT Jan 27, 2026, 2:39 PM

#

haughty sable <@1457407575476801641> it show this root@ClawdBot:~# npm config get prefix /usr ...

The npm install had engine warnings - may not have installed correctly. Let's check:

# Where did npm actually put moltbot?
ls -la /usr/lib/node_modules/
ls -la /usr/lib/node_modules/moltbot/

# Check if binary exists anywhere
find /usr -name "moltbot*" 2>/dev/null

Try with pnpm Instead (Recommended)

# Install pnpm
npm install -g pnpm

#


# Reinstall with pnpm
pnpm add -g moltbot

# Run with pnpm
pnpm moltbot onboard

Or Run Directly from Source

# Clone and run from source
git clone https://github.com/moltbot/moltbot.git
cd moltbot
pnpm install

#

pnpm run build
pnpm moltbot onboard

Quickest path: Install pnpm and use the git clone method:

npm install -g pnpm
git clone https://github.com/moltbot/moltbot.git
cd moltbot
pnpm install
pnpm moltbot onboard

#

Let me know if you hit any errors! 🦐

haughty sable Jan 27, 2026, 2:59 PM

#

well what now?

◇ Web search (optional) ───────────────────────────────────────────────────────────╮
│ │
│ If you want your agent to be able to search the web, you’ll need an API key. │
│ │
│ Moltbot uses Brave Search for the web_search tool. Without a Brave Search API │
│ key, web search won’t work. │
│ │
│ Set it up interactively: │
│ - Run: moltbot configure --section web │
│ - Enable web_search and paste your Brave Search API key │
│ │
│ Alternative: set BRAVE_API_KEY in the Gateway environment (no config changes). │
│ Docs: https://docs.molt.bot/tools/web │
│ │
├───────────────────────────────────────────────────────────────────────────────────╯
│
◇ What now ──────────────────────────────────────────────────────────╮
│ │
│ What now: https://molt.bot/showcase ("What People Are Building"). │
│ │
├─────────────────────────────────────────────────────────────────────╯
│
└ Onboarding complete. Use the tokenized dashboard link above to control Moltbot.

lethal horizonBOT Jan 27, 2026, 3:01 PM

#

haughty sable well what now? ◇ Web search (optional) ───────────────────────────────────────...

Onboarding complete! 🎉 You're all set!

What to Do Next

1. Open the Dashboard

The onboarding printed a tokenized URL above. Copy/paste it into your browser to access the dashboard.

If you missed it, run:

$(npm config get prefix)/bin/moltbot dashboard

2. Start the Gateway

If it hasn't started automatically:

#

$(npm config get prefix)/bin/moltbot gateway

3. Connect a Channel (Discord, etc.)

To enable Discord, edit ~/.clawdbot/clawdbot.json:

{
  "channels": {
    "discord": {
      "enabled": true,
      "token": "YOUR_DISCORD_BOT_TOKEN"
    }
  }
}

#

Then restart:

$(npm config get prefix)/bin/moltbot gateway restart

4. Need Help?

Docs: https://docs.molt.bot
Discord channel: You're already here!

Quick Status Check

$(npm config get prefix)/bin/moltbot status

#

What's your next goal? (Discord? WhatsApp? Telegram?) 🦐

haughty sable Jan 27, 2026, 3:03 PM

#

root@ClawdBot:~/moltbot# $(npm config get prefix)/bin/moltbot gateway
bash: /usr/bin/moltbot: No such file or directory
root@ClawdBot:~/moltbot#

lethal horizonBOT Jan 27, 2026, 3:04 PM

#

haughty sable root@ClawdBot:~/moltbot# $(npm config get prefix)/bin/moltbot gateway bash: /usr...

You're in the source directory! Run with pnpm instead:

pnpm moltbot gateway

Or if you want a global command, create a symlink:

ln -sf $(pwd)/bin/moltbot.js /usr/local/bin/moltbot
moltbot gateway

Quickest:

#

pnpm moltbot gateway

Let me know what happens! 🦐

haughty sable Jan 27, 2026, 3:11 PM

#

15:10:12 [canvas] host mounted at http://127.0.0.1:18789/__moltbot__/canvas/ (root /root/clawd/canvas)
15:10:12 Gateway failed to start: another gateway instance is already listening on ws://127.0.0.1:18789
If the gateway is supervised, stop it with: moltbot gateway stop
15:10:12 Port 18789 is already in use.
15:10:12 - pid 15448 root: clawdbot-gateway (127.0.0.1:18789)
15:10:12 - Another process is listening on this port.
15:10:12 Gateway service appears enabled. Stop it first.
15:10:12 Tip: moltbot gateway stop
15:10:12 Or: systemctl --user stop moltbot-gateway.service
ELIFECYCLE Command failed with exit code 1.
root@ClawdBot:~/moltbot#

lethal horizonBOT Jan 27, 2026, 3:12 PM

#

haughty sable 15:10:12 [canvas] host mounted at http://127.0.0.1:18789/__moltbot__/canvas/ (ro...

Gateway is already running! 🎉

Stop and Restart

pnpm moltbot gateway stop
pnpm moltbot gateway

Or Just Use It!

The gateway is already up at ws://127.0.0.1:18789

To access the dashboard:

pnpm moltbot dashboard

#

This will open the tokenized URL in your browser.

Quick check:

pnpm moltbot status

The gateway is working - you just needed to find the right command! 🦐

haughty sable Jan 27, 2026, 3:13 PM

#

but im hosting on a vps not local

lethal horizonBOT Jan 27, 2026, 3:14 PM

#

haughty sable but im hosting on a vps not local

No problem! The gateway is running on your VPS. Here are ways to access it remotely:

Option 1: SSH Tunnel (Quickest)

From your local machine:

# Local port 18789 → VPS port 18789
ssh -N -L 18789:127.0.0.1:18789 root@YOUR_VPS_IP

Then open http://127.0.0.1:18789 in your browser.

Option 2: Tailscale (Recommended for Always-On)

Install Tailscale on the VPS: curl -fsSL https://tailscale.com/install.sh | bash
Authenticate: tailscale up

#

Access via your Tailscale IP (100.x.x.x)

Option 3: Change Gateway Bind (Less Secure)

Edit ~/.clawdbot/clawdbot.json on VPS:

{
  "gateway": {
    "bind": "0.0.0.0"
  }
}

Then restart:

#

pnpm moltbot gateway restart

⚠️ Security warning: This exposes the gateway publicly. Add authentication!

For a VPS, I'd recommend Option 1 (SSH tunnel) for security.

Do you have SSH access set up from your local machine? 🦐

finite ibex Jan 28, 2026, 3:09 AM

#

adter migrating from clawdbot to moltbot, the commands are still using clawdbot, is this normal?

lethal horizonBOT Jan 28, 2026, 3:10 AM

#

finite ibex adter migrating from clawdbot to moltbot, the commands are still using clawdbot,...

Yes, this is confusing! There was a rename in progress. The situation:

What Happened

clawdbot → moltbot was the original rename
But the CLI was moved back to clawdbot recently (see recent releases)

Check Your Install

Run both commands to see which works:

clawdbot status
moltbot status

Fix PATH

#

If you have both installed, the one that works is correct. If you want to use a specific one:

For clawdbot (current CLI):

# Find it
which clawdbot

# If not found, reinstall
npm install -g clawdbot

For moltbot (old name):

# Find it
which moltbot

#

# If not found, that's expected - it's now clawdbot

The Bottom Line

clawdbot is the current CLI name. Use clawdbot commands from now on!

clawdbot status
clawdbot gateway

The project was renamed back (or the rename was reversed). This happened in a recent release - check the changelog for details. 🦐

finite ibex Jan 28, 2026, 3:19 AM

#

where was it announced being renamed back?

verbal gate Jan 28, 2026, 6:41 PM

#

brave api

lethal horizonBOT Jan 28, 2026, 6:45 PM

#

verbal gate brave api

Hey Paul! 🦐

What would you like to know about the Brave API? Are you looking to:

Set it up for web search in Clawdbot?
Get an API key?
Understand how it works?

Let me know and I'll point you in the right direction! Happy to help! 🦐

verbal gate Jan 28, 2026, 6:57 PM

#

Set it up for web search in Clawdbot

still nest Jan 28, 2026, 7:02 PM

#

spare totem I have added LM Studio agent and Clawdbot was working correctly, later, I added ...

Having trouble getting lm studio to work myself rn

lethal horizonBOT Jan 28, 2026, 7:05 PM

#

verbal gate Set it up for web search in Clawdbot

Here's how to set up Brave API for web search in Clawdbot! 🦐

1. Get Your API Key

Go to https://brave.com/search/api/
Create an account and choose the Data for Search plan (not "Data for AI" — that won't work!)
Generate your API key

2. Configure Clawdbot

Set it via the CLI (recommended):

clawdbot configure --section web

Or set the environment variable:

#

export BRAVE_API_KEY="your_key_here"

Or in your config file:

{
  tools: {
    web: {
      search: {
        provider: "brave",
        apiKey: "YOUR_API_KEY",
        maxResults: 5,
        timeoutSeconds: 30
      }
    }

#

}
}

3. Restart the Gateway

clawdbot gateway restart

That's it! Your agent will now use Brave Search for web_search queries. 🎉

Note: The Data for AI plan won't work — you need the Data for Search plan. Brave has a free tier, so you can test it out first!

Let me know if you hit any snags! 🦐

#How to make two agents work if one is offline?

How Model Selection Works

The Problem

Solutions

Option 1: Start a New Session (Quickest)

Option 2: Configure Fallbacks

Option 3: Switch in Current Session

Check Your Config

Solutions

Option 1: Ask the Owner to Add You to the Allowlist

Option 2: Switch Models Directly (No /new Needed)

Option 3: Have the Owner Disable Access Group Enforcement

Quick Fixes

Option A: Switch Models Directly (Works Now)

Option B: Get Added to the Allowlist

Option C: Disable Access Group Enforcement (Less Secure)

Quick Fixes

1. Restart the Gateway

2. Check Gateway Status

3. Follow the Logs

4. Run Doctor

Common VPS Issues

Quick Fix

What to Share

Install Now

After Install

For Antgravity (Google Models)

If Install Succeeded

If You Want to Skip Onboarding and Configure Manually

What's Next?

Option 1: Git Clone (Direct from GitHub)

Option 2: Try the Clawdbot Installer (Same Project)

Option 3: npm Install Directly

Option 4: Check DNS/Network

Quick One-Liner

To Continue Onboarding

To Check Gateway Status

Try with pnpm Instead (Recommended)

Or Run Directly from Source

What to Do Next

1. Open the Dashboard

2. Start the Gateway

3. Connect a Channel (Discord, etc.)

4. Need Help?

Quick Status Check

Stop and Restart

Or Just Use It!

Option 1: SSH Tunnel (Quickest)

Option 2: Tailscale (Recommended for Always-On)

Option 3: Change Gateway Bind (Less Secure)

What Happened

Check Your Install

Fix PATH

The Bottom Line

1. Get Your API Key

2. Configure Clawdbot

3. Restart the Gateway

Option 2: Switch Models Directly (No `/new` Needed)