Skip to content
View in the app

A better way to browse. Learn more.

Unraid

A full-screen app on your home screen with push notifications, badges and more.

To install this app on iOS and iPadOS
  1. Tap the Share icon in Safari
  2. Scroll the menu and tap Add to Home Screen.
  3. Tap Add in the top-right corner.
To install this app on Android
  1. Tap the 3-dot menu (⋮) in the top-right corner of the browser.
  2. Tap Add to Home screen or Install app.
  3. Confirm by tapping Install.

[Support] OpenClaw — AI Personal Assistant

Featured Replies

I got everything up and running but i having an issue on getting telegram to work.

I need to run openclaw pairing approve telegram with the code and i cannot figure out how to make it work

anyone get telegram working?

  • Replies 201
  • Views 90.9k
  • Created
  • Last Reply

Top Posters In This Topic

Most Popular Posts

  • Sharing my local ollama settings. "models": { "providers": { "ollama": { "baseUrl": "http://xxx.xxx.xxx.xxx:11434/v1", "apiKey": "ollama", "api": "openai-responses

  • PikkonMG
    PikkonMG

    If your on chrome or like me edge then edge://flags/#unsafely-treat-insecure-origin-as-secure Enable Insecure origins treated as secure and add your local ip on unraid server running openclaw I.E http

  • Can't seem to find the template on community apps. Was this published yet?

Posted Images

U can use
node dist/index.js
instead of the openclaw command for everything.

So dont use
openclaw onboard
in the console, but use:
node dist/index.js onboard

Looks like all commands work this way.

Great work!


The template seems to be lacking the field for a Venice.AI api key. I like to support privacy respecting companies.

Thanks!

Hi there, been able to run and have fun with first instance, but now I want to create a second one and I can't understand how the port mapping works. 18789 is already taken so I try to use 18790 and different arrangements in the 3 fields that let me put a port.

Any help welcome. Thanks :)

Was able to make it work
Web ui: http://[IP]:[PORT:18790]/

Control UI Port: 18790
Gateway Port: 18689 <keep original>

Been able to run two like this. Thanks for all the work

For some reason I still can't install skills from the web ui using brew.

[ws] ⇄ res ✗ skills.install 14767ms errorCode=UNAVAILABLE errorMessage=Install failed (exit 1): Error: summarize: An unsatisfied requirement failed this build. conn=a5eec973…0231 id=af2e3c10…a4f5

Has anyone a working config to use the Google Antigravity Oauth?

I have activated the Plugin but how do i get to log into my google account now?

Edited by DUFRelic

hey, i installed it today and configured my ollama to give it a model (before the first start).
if someone is interested:
edit the /mnt/user/appdata/openclaw/config/openclaw.json
replace the ip (YOUR_OLLAMA_HOST_IP) to your host and your preferred model ("id":)

  "models": {
    "providers": {
      "ollama": {
        "baseUrl": "http://YOUR_OLLAMA_HOST_IP:11434/v1",
        "apiKey": "ollama-local",
        "api": "openai-completions",
        "models": [
          {
            "id": "ministral-3:14b-instruct-2512-q8_0",
            "name": "Ministral 3 14B Instruct",
            "reasoning": false,
            "input": [
              "text"
            ],
            "cost": {
              "input": 0,
              "output": 0,
              "cacheRead": 0,
              "cacheWrite": 0
            },
            "contextWindow": 32768,
            "maxTokens": 81920
          }
        ]
      }
    }
  },

But the typical onboading I saw in some videos did not happen and therefore the personality is not used / configured.
I connected to the container (docker exec -it OpenClaw bash) and executed node dist/index.js onboard

it recognized my model and i skipped everything else, then i chose "Hatch in TUI (recommended)" then follows the output:

Wake up, my friend!                                                                                                              
                                                                                                                                  
 (no output)                                                                                                                      
 (no output)                                                                                                                      
 connected | idle                                                                                                                 
 agent main | session main (openclaw-tui) | ollama/ministral-3:14b-instruct-2512-q8_0 | tokens 4.1k/33k (13%)        

But on the dashboard is just a typical ai-greeting text and when i ask it has no access to its filesystem and its IDENDITY.md SOUL.md etc.

So I guess my assistant isn't working properly? Do i need to configure a special workspace for my model?

has anyone managed to get browser functions to work? I would like to use my local searxng instance, but I don't know how to get it's own function to work, I would like to avoid braves api one.

I am trying to set up gog for google and the auth url it wants me to use doesn't work since its not from the same instance, is there a work around for this?

18 hours ago, DUFRelic said:

Has anyone a working config to use the Google Antigravity Oauth?

I have activated the Plugin but how do i get to log into my google account now?

I got it working.

I just used the normal onboarding.

I connected to the container (docker exec -it OpenClaw bash) and executed node dist/index.js onboard

Hi team, whats the best way to ensure python and pip3 get installed on docker properly?

13 hours ago, Specthildur said:

has anyone managed to get browser functions to work? I would like to use my local searxng instance, but I don't know how to get it's own function to work, I would like to avoid braves api one.

i think there is a bit of a bug with the headless mode https://www.answeroverflow.com/m/1467254436631023889 i have been having trouble with it also. Even on box in Ubuntu it doesn't know how to run the snapd version of Chromium.

30 minutes ago, Draco said:

Hi team, whats the best way to ensure python and pip3 get installed on docker properly?

don't even know why i post the claw has figured it out installed a python skill.

Why cant we use the openclaw-cli command in docker container?
From OpenClaw website it seems it supposed to be available

https://docs.openclaw.ai/install/docker

*Edit*
I've changed the post arguments to this, and now its working
sh -c "mkdir -p /root/.openclaw /home/linuxbrew; [ -s /root/.openclaw/openclaw.json ] || echo '{\"gateway\":{\"mode\":\"local\",\"bind\":\"lan\",\"controlUi\":{\"allowInsecureAuth\":true},\"auth\":{\"mode\":\"token\"}}}' > /root/.openclaw/openclaw.json; echo 'alias openclaw-cli=\"node dist/index.js\"' >> ~/.bashrc; exec node dist/index.js gateway --bind lan"

Edited by iBuSH

On 2/4/2026 at 9:18 AM, Timminater said:

U can use
node dist/index.js
instead of the openclaw command for everything.

So dont use
openclaw onboard
in the console, but use:
node dist/index.js onboard

Looks like all commands work this way.

I had this issue when i first spun it up as well. To get the openclaw commands to register, you need to add the npm link command to the post args in a very specific place...

<PostArgs>
sh -c "mkdir -p /root/.openclaw /home/linuxbrew /var/run/tailscale; [ -s /root/.openclaw/openclaw.json ] || echo '{\"gateway\":{\"mode\":\"local\",\"bind\":\"loopback\",\"controlUi\":{\"allowInsecureAuth\":true},\"auth\":{\"allowTailscale\":true},\"tailscale\":{\"mode\":\"serve\"}}}' > /root/.openclaw/openclaw.json; curl -fsSL https://pkgs.tailscale.com/stable/tailscale_latest_amd64.tgz | tar -xzf - -C /tmp && cp /tmp/tailscale_*_amd64/tailscale* /usr/local/bin/; tailscaled --tun=userspace-networking --state=/var/lib/tailscale/tailscaled.state --socket=/var/run/tailscale/tailscaled.sock & sleep 5; tailscale up --authkey=${TAILSCALE_AUTHKEY} --hostname=openclaw --accept-routes=false; cd /app & npm link & exec node dist/index.js gateway --bind loopback "

Explanation from Perplexity...
What npm link Does in This Context

In this Docker entrypoint script, npm link is being used to create a global symlink for the Node.js package located in /app. This makes the package and any CLI tools defined in its package.json available globally within the container.

How it works

When npm link is executed in a directory containing a package.json, it:

  1. Creates a symbolic link in the global node_modules folder pointing to /app

  2. Links any bin executables defined in the package to the global npm binary directory (e.g., {prefix}/bin/{name})

Why it's used here

In the OpenClaw container startup:

cd /app & npm link & exec node dist/index.js gateway --bind loopback

The npm link serves to globally expose the OpenClaw CLI tools. This enables:

  • Running openclaw commands from anywhere in the container without specifying the full path

  • Access to the package's binaries as system commands

  • Development-style convenience where the package behaves like a globally installed npm package

The Full Sequence

Step

Command

Purpose

1

cd /app

Navigate to application directory

2

npm link

Create global symlink for the package and its CLI binaries

3

exec node dist/index.js

Start the gateway service

The & operators run these commands in parallel, allowing the main application to start while npm link completes its filesystem operations in the background.

Hi so i was finally able to make it through the onboarding , using

node dist/index.js onboard

I needed to conect via chatgpt plus account since i do not have any other AI models , but now whenever i try to connect via the web UI i get a token mismatch error , any ideas on how can i change this ????

Ok i seem to have fixed the tokens issue , however i cannot install any skills via the onboarding menu or the , web browser , i can finally make small talk with the bot tho .

Any ideas ?

This needs a space invader one video really bad 😄

Edited by bclinton

  • Author
On 2/6/2026 at 12:19 PM, shremi said:

Ok i seem to have fixed the tokens issue , however i cannot install any skills via the onboarding menu or the , web browser , i can finally make small talk with the bot tho .

Any ideas ?

Did you follow the command listed in the template to install homebrew after install?

On 2/5/2026 at 5:48 PM, freakadings said:

hey, i installed it today and configured my ollama to give it a model (before the first start).
if someone is interested:
edit the /mnt/user/appdata/openclaw/config/openclaw.json
replace the ip (YOUR_OLLAMA_HOST_IP) to your host and your preferred model ("id":)

  "models": {
    "providers": {
      "ollama": {
        "baseUrl": "http://YOUR_OLLAMA_HOST_IP:11434/v1",
        "apiKey": "ollama-local",
        "api": "openai-completions",
        "models": [
          {
            "id": "ministral-3:14b-instruct-2512-q8_0",
            "name": "Ministral 3 14B Instruct",
            "reasoning": false,
            "input": [
              "text"
            ],
            "cost": {
              "input": 0,
              "output": 0,
              "cacheRead": 0,
              "cacheWrite": 0
            },
            "contextWindow": 32768,
            "maxTokens": 81920
          }
        ]
      }
    }
  },

But the typical onboading I saw in some videos did not happen and therefore the personality is not used / configured.
I connected to the container (docker exec -it OpenClaw bash) and executed node dist/index.js onboard

it recognized my model and i skipped everything else, then i chose "Hatch in TUI (recommended)" then follows the output:

Wake up, my friend!                                                                                                              
                                                                                                                                  
 (no output)                                                                                                                      
 (no output)                                                                                                                      
 connected | idle                                                                                                                 
 agent main | session main (openclaw-tui) | ollama/ministral-3:14b-instruct-2512-q8_0 | tokens 4.1k/33k (13%)        

But on the dashboard is just a typical ai-greeting text and when i ask it has no access to its filesystem and its IDENDITY.md SOUL.md etc.

So I guess my assistant isn't working properly? Do i need to configure a special workspace for my model?

If you login to the webgui you will see the answers. Somehow these answers are not shown on the terminal.

I see this now appearsa in CA.

What location are people choosing for the worlspace? Default seems to be within the appdata open claw folder but is that the best place for it? Just thought I'd see what others are doing

I am havin gthis issue [gateway connect failed: Error: pairing required] : is it wellknown ?


root@OpenClaw:/app# node dist/index.js health
🦞 OpenClaw 2026.2.9 (unknown) — I don't judge, but your missing API keys are absolutely judging you.
gateway connect failed: Error: pairing required
Error: gateway closed (1008): pairing required
Gateway target: ws://172.17.0.4:18789
Source: local lan 172.17.0.4
Config: /root/.openclaw/openclaw.json
Bind: lan

EDITED

perhaps dumb but i fixed this by simply deleting the pending.json
In my case, it was here:
/mnt/user/appdata/openclaw/config/devices/pending.json

Edited by Cr4sKy

Hello, anyone can show me how should openclaw.json look when I have Ollama + qwen2.5:7b [on unraid] and Anthropic API. I want to have Claude sonnet 4.5 as primary model. I want also to add Telegram chat. I try many configurations and always have errors. Openclaw container won't start. If it start then cannot setup Telegram in Dashboard.
Please help.
Regards.

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

Account

Navigation

Search

Search

Configure browser push notifications

Chrome (Android)
  1. Tap the lock icon next to the address bar.
  2. Tap Permissions → Notifications.
  3. Adjust your preference.
Chrome (Desktop)
  1. Click the padlock icon in the address bar.
  2. Select Site settings.
  3. Find Notifications and adjust your preference.