downloading process | Friends of the Crustacean 🦞🤝 | Page 1

tired tundra Apr 15, 2026, 10:03 AM

#

the download is so confusing, i need help setting it up and everything

novel wigeonBOT Apr 15, 2026, 10:04 AM

#

yeah, the setup can feel like a lot at first.

The recommended easiest path is:

macOS / Linux

curl -fsSL https://openclaw.ai/install.sh | bash
openclaw onboard --install-daemon

Windows (PowerShell)

iwr -useb https://openclaw.ai/install.ps1 | iex
openclaw onboard --install-daemon

Then check:

#

openclaw gateway status
openclaw dashboard

Docs:

Reply with these and I’ll guide you step by step:

your OS
openclaw --version
if that works, node --version
the exact step or error you’re stuck on

If openclaw --version says “command not found”, just say that, and tell me your OS.

tired tundra Apr 15, 2026, 10:12 AM

#

cant move using arrow keys during setup

novel wigeonBOT Apr 15, 2026, 10:13 AM

#

ah, that usually means the interactive prompt isn’t getting proper TTY keyboard input.

Try this first:

Cancel setup with Ctrl+C
Re-run it in a normal terminal
- Windows: Windows Terminal, ideally WSL2
- macOS: Terminal or iTerm
- Linux: your regular terminal
Run:

openclaw onboard --install-daemon

A few things to avoid:

GitHub Codespaces/browser shells
weird embedded terminals

#

terminals opened through apps that don’t pass arrow keys correctly

If you’re on Windows, docs strongly recommend WSL2 for the smoothest setup:
https://docs.openclaw.ai/start/getting-started
https://docs.openclaw.ai/start/wizard

Reply with:

your OS
what terminal you’re using (PowerShell, cmd, Git Bash, WSL, VS Code terminal, etc.)
openclaw --version

and I’ll tell you the exact fix.

tired tundra Apr 15, 2026, 10:18 AM

#

after i click skip for now in selecting channels it says this :"o Select channel (QuickStart)
| Skip for now
TypeError: Cannot read properties of undefined (reading 'trim')
openclaw : File C:\Users\Eosperus\AppData\Roaming\npm\openclaw.ps1 cannot be loaded because running scripts is disabled on this system. For more information, see about_Execution_Policies at https:/go.microsoft.com/fwlink/?LinkID=135170.
At line:2 char:1

openclaw onboard --install-daemon

  + CategoryInfo          : SecurityError: (:) [], ParentContainsErrorRecordException
  + FullyQualifiedErrorId : UnauthorizedAccess

PS C:\Users\Eosperus>"

novel wigeonBOT Apr 15, 2026, 10:19 AM

#

yep, that’s two separate Windows problems:

PowerShell is blocking openclaw.ps1
That trim crash after “Skip for now” looks like a real onboarding bug, not you

First, bypass the PowerShell script block

In PowerShell, run the .cmd shim instead:

openclaw.cmd --version
openclaw.cmd onboard --install-daemon

That usually works even when openclaw.ps1 is blocked.

If you want to fix PowerShell permanently, run this once and reopen PowerShell:

#

Set-ExecutionPolicy -Scope CurrentUser RemoteSigned

For the `trim` crash

That part looks like a setup bug on native Windows. I’d try updating first:

npm i -g openclaw@latest
openclaw.cmd --version
openclaw.cmd onboard --install-daemon

Docs:

Windows: https://docs.openclaw.ai/platforms/windows
Updating: https://docs.openclaw.ai/install/updating

#

My honest recommendation

Since you’re on native Windows and already hitting Windows-specific issues, WSL2 is the smoother path for OpenClaw.

If you want, I can give you the exact WSL2 setup steps next, one by one.
Also paste the output of:

openclaw.cmd --version

tired tundra Apr 15, 2026, 10:25 AM

#

it still says :"o Select channel (QuickStart)
| Skip for now
TypeError: Cannot read properties of undefined (reading 'trim')
PS C:\Users\Eosperus>
"

novel wigeonBOT Apr 15, 2026, 10:26 AM

#

okay, don’t keep fighting QuickStart. That crash is happening on the QuickStart channel step.

Try the manual flow instead:

openclaw.cmd onboard --flow manual

If that works, you can skip channels and just use the browser dashboard first, because channels are optional:

openclaw.cmd dashboard

Also paste this so I can tell if you’re on a buggy build:

#

openclaw.cmd --version

If manual flow also crashes, I’d stop wasting your time on native Windows and switch to WSL2, which is the recommended Windows path:
https://docs.openclaw.ai/platforms/windows

If you want, send the output of openclaw.cmd --version and I’ll give you the next exact step.

tired tundra Apr 15, 2026, 10:26 AM

#

i cant move using arrow keys during choosing y/n lock down and quickstart/manual

novel wigeonBOT Apr 15, 2026, 10:28 AM

#

try these keys, Clack supports vim-style keys too:

j = down
k = up
h = left
l = right
Enter = confirm

So:

QuickStart / Manual → use j / k, then Enter
y/n lockdown → use h / l, then Enter

If those also don’t work, then your Windows terminal is mangling interactive key input, and I’d switch to Windows Terminal or, better, WSL2:
https://docs.openclaw.ai/platforms/windows

If you want, try j/k/h/l once and tell me what happens.

tired tundra Apr 15, 2026, 10:28 AM

#

now how do i resume downloading

novel wigeonBOT Apr 15, 2026, 10:29 AM

#

you probably don’t need to re-download.

If this works:

openclaw.cmd --version

then OpenClaw is already installed, and you can just continue setup with:

openclaw.cmd onboard --flow manual

Then use:

j/k to move up/down
h/l for left/right

#

Enter to confirm

If openclaw.cmd --version says command not found, then rerun the installer:

iwr -useb https://openclaw.ai/install.ps1 | iex

If you want, paste what openclaw.cmd --version prints and I’ll tell you the exact next step.

tired tundra Apr 15, 2026, 10:30 AM

#

btw i cant only move during the first 2 choices, the ai model and channel choices i can use arrow keys

novel wigeonBOT Apr 15, 2026, 10:31 AM

#

ah, that makes sense.

Those first two are likely horizontal prompts, so they use:

Left / Right
or h / l
then Enter

The later ones, like AI model and channel, are vertical lists, so they use:

Up / Down
or j / k
then Enter

So basically:

lockdown y/n → left/right

#

quickstart/manual → left/right
model/channel lists → up/down

Glad we caught that. If you want, keep going and paste the next thing it gets stuck on.

tired tundra Apr 15, 2026, 10:31 AM

#

oh now i found out that i can move using both arrows and jkhl for a split second and then it just doenst allow me to move anymore while choosing the first 2 choices

novel wigeonBOT Apr 15, 2026, 10:32 AM

#

yeah, that sounds like a real terminal/input bug on those first two prompts.

At this point, I’d stop using the interactive wizard and bypass it completely with non-interactive onboarding.

Pick the one that matches your AI provider:

OpenAI

openclaw.cmd onboard --non-interactive --mode local --auth-choice openai-api-key --openai-api-key "YOUR_OPENAI_API_KEY" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Gemini

openclaw.cmd onboard --non-interactive --mode local --auth-choice gemini-api-key --gemini-api-key "YOUR_GEMINI_API_KEY" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Anthropic

#

openclaw.cmd onboard --non-interactive --mode local --auth-choice apiKey --anthropic-api-key "YOUR_ANTHROPIC_API_KEY" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Then check:

openclaw.cmd gateway status
openclaw.cmd dashboard

If you’re using something else like Ollama, tell me which one and I’ll give you the exact command.

tired tundra Apr 15, 2026, 10:33 AM

#

nevermind it works now, but after i choose local gateway in workspace directory it just freezes

novel wigeonBOT Apr 15, 2026, 10:34 AM

#

that next workspace directory step is a text input, not a menu.

So it can look “frozen”, but it’s usually waiting for you to:

press Enter to accept the default workspace path, or
type a folder path and then press Enter

So after choosing Local gateway, when it shows the workspace directory, just hit Enter if the default looks fine.

If pressing Enter there really does nothing for 10 to 15 seconds, paste the exact last line it shows and I’ll help from there.

tired tundra Apr 15, 2026, 10:35 AM

#

Workspace directory
| C:\Users\Eosperus.openclaw\workspace█
—
and then it freezes

novel wigeonBOT Apr 15, 2026, 10:35 AM

#

that actually doesn’t look frozen. The █ means the cursor is sitting in the text field.

Just press Enter to accept:

C:\Users\Eosperus\.openclaw\workspace

If you want a different folder, edit it first, then press Enter.

If you press Enter and it still doesn’t move, tell me exactly what happens after that.

tired tundra Apr 15, 2026, 10:36 AM

#

why cant i cntrl + V in powershell when it ask for my api key

novel wigeonBOT Apr 15, 2026, 10:36 AM

#

because that prompt is running in raw keyboard input mode, so normal Ctrl+V paste often doesn’t work there.

Try one of these instead:

Right-click in the terminal window
Shift+Insert
Ctrl+Shift+V if you’re in Windows Terminal

So the flow is:

copy your API key
click the PowerShell window
use right-click or Shift+Insert
press Enter

You’re not doing anything wrong, that’s just a weird terminal/prompt thing on Windows.

tired tundra Apr 15, 2026, 10:40 AM

#

novel wigeonBOT Apr 15, 2026, 10:40 AM

#

choose Loopback (127.0.0.1).

That’s the right default for most people, especially if you’re just using OpenClaw on your own PC.

Quick guide:

Loopback: only this computer can access it, safest default
LAN: other devices on your home network can access it
Tailnet: only if you already use Tailscale
Auto: fallback behavior, not needed for a normal setup
Custom IP: advanced/manual setup

So yep, pick Loopback and continue.

tired tundra Apr 15, 2026, 10:41 AM

#

gateway auth token or password? (im completely new)

novel wigeonBOT Apr 15, 2026, 10:41 AM

#

choose Token.

That’s the better default for a new setup, and OpenClaw can usually generate one for you automatically.

Simple rule:

Token = recommended, safer default, no need to invent a password
Password = only if you specifically want to type your own password manually

So yep, pick Token and keep going.

tired tundra Apr 15, 2026, 10:41 AM

#

what about tailscale exposure?

novel wigeonBOT Apr 15, 2026, 10:42 AM

#

choose Off.

Only use Tailscale exposure if you already use Tailscale and want to access OpenClaw from another device securely.

For a first setup:

Off = correct default
turn it on later only if you know you need remote access

So: Off.

tired tundra Apr 15, 2026, 10:42 AM

#

How do you want to provide the gateway token?
| > Generate/store plaintext token (Default)
| Use SecretRef
—

novel wigeonBOT Apr 15, 2026, 10:42 AM

#

choose Generate/store plaintext token (Default).

That’s the right choice for a first setup.

Generate/store plaintext token = easiest, normal default
Use SecretRef = more advanced, for people managing secrets from env/files/providers

So yep, keep the default and press Enter.

tired tundra Apr 15, 2026, 10:43 AM

#

Gateway token (blank to generate)
| Needed for multi-machine or non-loopback access
—
?

novel wigeonBOT Apr 15, 2026, 10:43 AM

#

leave it blank and press Enter.

OpenClaw will generate the token for you automatically, which is perfect for a first setup.

That note is just saying the token matters if you ever access it from another machine later. For now, blank is fine.

tired tundra Apr 15, 2026, 10:44 AM

#

configure chat channel now?

novel wigeonBOT Apr 15, 2026, 10:45 AM

#

if you’re completely new, I’d choose Skip for now.

You do not need a chat channel to get started, you can use the dashboard in your browser first.

Best beginner path:

Skip for now
finish setup
test it in the dashboard
add a chat channel later if you want phone/chat app access

If you want the easiest channel later, Telegram is usually the simplest.

tired tundra Apr 15, 2026, 10:46 AM

#

which search provider should i use?

novel wigeonBOT Apr 15, 2026, 10:46 AM

#

for a first setup, I’d choose DuckDuckGo.

Why:

no API key needed
easiest to get through setup right now
you can change it later

Quick rule:

DuckDuckGo = easiest beginner choice
Brave = better structured results, but needs a Brave API key
Skip / configure later = also fine if you just want to finish setup

So my recommendation: DuckDuckGo.
You can always change it later with openclaw configure --section web.

tired tundra Apr 15, 2026, 10:47 AM

#

|

Search provider
| Brave Search
| > Exa Search (Neural + keyword search with date filters and content extraction)
| Firecrawl Search
| Gemini (Google Search)
| Grok (xAI)
| Kimi (Moonshot)
| Ollama Web Search
| Perplexity Search
| SearXNG Search
| Tavily Search
| Skip for now

what do i do

novel wigeonBOT Apr 15, 2026, 10:47 AM

#

with that list, I’d choose Skip for now.

Reason:

most of those need extra setup / another API key
you can finish OpenClaw first
add search later once everything else is working

Only pick one now if you already use it already, for example:

Gemini if you already have a Gemini key and want Google-backed search
Ollama Web Search if you already run Ollama locally
SearXNG if you already have your own SearXNG instance

Otherwise: Skip for now is the best move.

tired tundra Apr 15, 2026, 10:48 AM

#

Configure skills now? (recommended)
| > Yes / No
—

novel wigeonBOT Apr 15, 2026, 10:48 AM

#

I’d pick No for now.

You’ve already had enough setup friction, and skipping skills will help you get to a working install faster.

Best beginner path:

finish core setup first
make sure the dashboard opens
add skills later once the basics work

So: No.

tired tundra Apr 15, 2026, 10:49 AM

#

Configure plugins (select to set up now, or skip)
| [ ] Skip for now
| [ ] @openclaw/amazon-bedrock-provider
| [ ] @openclaw/github-copilot-provider
| [ ] @openclaw/google-plugin
| [ ] @openclaw/huggingface-provider
| [ ] @openclaw/minimax-provider
| [ ] @openclaw/moonshot-provider
| [ ] @openclaw/ollama-provider
| [ ] @openclaw/xai-plugin
| [ ] ACPX Runtime
| [•] Device Pairing (1 field)

what do i do

novel wigeonBOT Apr 15, 2026, 10:49 AM

#

I’d skip plugins for now.

Do this:

move to Skip for now
press Space so it gets checked
if Device Pairing is still checked, press Space on that too to uncheck it
press Enter

That keeps the setup simple, which is what you want right now. You can add plugins later once the basic install is working.

tired tundra Apr 15, 2026, 10:50 AM

#

|

Enable hooks?
| [•] Skip for now
| [ ] 🚀 boot-md
| [ ] 📎 bootstrap-extra-files
| [ ] 📝 command-logger
| [ ] 💾 session-memory
—
?

novel wigeonBOT Apr 15, 2026, 10:50 AM

#

leave Skip for now checked and press Enter.

That’s the right move for a first setup. Hooks are extra behavior you can enable later once everything is working.

tired tundra Apr 15, 2026, 10:50 AM

#

Install Gateway service (recommended)
| > Yes / No
—
?

novel wigeonBOT Apr 15, 2026, 10:51 AM

#

choose Yes.

That makes OpenClaw easier to use because the Gateway service gets installed for you instead of needing to start it manually every time.

So:

Yes = recommended, easier
No = only if you want to run it manually yourself

If it fails afterward, we can fall back to manual, but for now pick Yes.

tired tundra Apr 15, 2026, 10:52 AM

#

now a node window with the name of openclaw-gateway pops up, what do i do

novel wigeonBOT Apr 15, 2026, 10:52 AM

#

that’s normal, it’s the Gateway process starting.

What to do:

leave that window open
go back to the onboarding/setup window
let setup keep running

So basically: don’t close the openclaw-gateway window unless it shows an error.

If setup finishes and you’re unsure what next screen means, paste it here and I’ll walk you through it.

tired tundra Apr 15, 2026, 10:53 AM

#

how do i know if the setup is finished?

novel wigeonBOT Apr 15, 2026, 10:54 AM

#

you’ll usually know setup is finished when:

the onboarding window stops asking questions
it shows a final success/summary message
you get your normal PowerShell prompt back

Then verify with:

openclaw.cmd gateway status
openclaw.cmd dashboard

What you want:

gateway status shows the gateway is running
dashboard opens OpenClaw in your browser

#

If you want, paste the final output once it stops and I’ll confirm whether you’re done.

tired tundra Apr 15, 2026, 10:54 AM

#

in powershell it says:"|
— Onboarding complete. Use the dashboard link above to control OpenClaw.

PS C:\Users\Eosperus>"

#

hello?

novel wigeonBOT Apr 15, 2026, 10:57 AM

#

yep, I’m here 🦐

Did the dashboard open okay, or did you get stuck on the next step?

tired tundra Apr 15, 2026, 10:57 AM

#

in powershell it says:"|
— Onboarding complete. Use the dashboard link above to control OpenClaw.

PS C:\Users\Eosperus>"

novel wigeonBOT Apr 15, 2026, 10:57 AM

#

yep, that means it finished successfully.

You’re done with onboarding. Next:

openclaw.cmd gateway status
openclaw.cmd dashboard

Or just click the dashboard link it printed above.

If the dashboard opens, you’re in 👍

tired tundra Apr 15, 2026, 11:00 AM

#

i told it to reply with "yes" and it doesnt even answer, all i see is
Assistant 5:59 PM gemini-3.1-pro-preview

novel wigeonBOT Apr 15, 2026, 11:00 AM

#

got it, that means the chat UI is opening, but the actual reply is probably failing.

First, please run these in PowerShell and paste the output:

openclaw.cmd --version
openclaw.cmd status
openclaw.cmd models status

If you want one more useful check too:

openclaw.cmd gateway status

Then I can tell you exactly what’s wrong.

tired tundra Apr 15, 2026, 11:01 AM

#

in the logs it says "17:59:55 [agent/embedded] embedded run agent end: runId=8a401b23-e8c8-4c80-8641-56727c0bd2bf isError=true model=gemini-3.1-pro-preview provider=google error=LLM error: {
"error": {
"code": 400,
"message": "API key not valid. Please pass a valid API key.",
"status": "INVALID_ARGUMENT",
"details": [
{
"@type": "type.googleapis.com/google.rpc.ErrorInfo",
"reason": "API_KEY_INVALID",
"domain": "googleapis.com",
"metadata": {
"service": "generativelanguage.googleapis.com"
}
},
… rawError={"error":{"message":"{\n "error": {\n "code": 400,\n "message": "API key not valid. Please pass a valid API key.",\n "status": "INVALID_ARGUMENT",\n "details": [\n {\n "@type": "type.…"

novel wigeonBOT Apr 15, 2026, 11:01 AM

#

perfect, we found it.

The problem is not OpenClaw now. Your Gemini API key is invalid.

Fix

Get a valid Gemini API key from Google AI Studio, then reconfigure the model:

openclaw.cmd configure --section model

Then:

choose Gemini API key
paste the new valid key
finish the model setup

After that, verify with:

#

openclaw.cmd models status

And try the dashboard again.

Docs:
https://docs.openclaw.ai/providers/google

Also, don’t paste your API key here in Discord. If you want, I can walk you through the configure --section model prompts too.

tired tundra Apr 15, 2026, 11:03 AM

#

Model/auth provider
| Anthropic
| Arcee AI
| BytePlus
| Chutes
| Cloudflare AI Gateway
| Copilot Proxy
| Custom Provider
| DeepSeek
| Fireworks
| Gemini CLI OAuth
| GitHub Copilot
| > Google (Gemini API key + OAuth)
| Hugging Face
| Kilo Gateway
| LiteLLM
| LM Studio
| Microsoft Foundry
| MiniMax
| Mistral AI
| Moonshot AI (Kimi K2.5)
| Ollama
| OpenAI
| OpenAI Codex
| OpenCode
| OpenRouter
| Qianfan
| Qwen Cloud
| SGLang
| StepFun
| Synthetic
| Together AI
| Venice AI
| Vercel AI Gateway
| vLLM
| Volcano Engine
| xAI (Grok)
| Xiaomi
| Z.AI
| Skip for now
— help me choose

novel wigeonBOT Apr 15, 2026, 11:03 AM

#

choose the provider you actually have a working key/account for.

For you, the best rule is:

if you want to use Gemini and you can get a new valid Gemini API key, keep Google (Gemini API key + OAuth)
if you already have an OpenAI key, choose OpenAI
if you already have an Anthropic key, choose Anthropic

My recommendation right now:

stay on Google (Gemini API key + OAuth)
then paste a fresh valid Gemini API key

I would not pick:

Skip for now, because then it still won’t be able to answer
Gemini CLI OAuth, unless you already use the gemini CLI and know you want that route

#

So: pick Google again, but use a new valid key this time.

tired tundra Apr 15, 2026, 11:04 AM

#

wait, i can move in the selection menu

novel wigeonBOT Apr 15, 2026, 11:04 AM

#

nice, that’s good 😄

Then go ahead and choose:

Google (Gemini API key + OAuth)

and enter a fresh valid Gemini API key.

If you want, send the next prompt after that and I’ll keep walking you through it.

tired tundra Apr 15, 2026, 11:05 AM

#

wait i can type but cant paste the api key

novel wigeonBOT Apr 15, 2026, 11:05 AM

#

yep, that raw prompt is being annoying again.

Try paste in this order:

Right-click inside the PowerShell window
Shift+Insert
Ctrl+Shift+V if you’re in Windows Terminal

Do not use Ctrl+V there, it usually won’t work in that prompt.

If none of those work:

copy the key
click the terminal once
try right-click paste
then press Enter

If it still refuses, tell me whether you’re in Windows Terminal, PowerShell, or cmd, and I’ll give you the workaround for that exact one.

tired tundra Apr 15, 2026, 11:06 AM

#

im in powershell and both dont work

novel wigeonBOT Apr 15, 2026, 11:07 AM

#

okay, then let’s bypass the paste prompt entirely.

Cancel that prompt with Ctrl+C, then run this in PowerShell, replacing the key locally on your PC:

openclaw.cmd onboard --non-interactive --mode local --auth-choice gemini-api-key --gemini-api-key "YOUR_NEW_VALID_GEMINI_KEY" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Important:

replace YOUR_NEW_VALID_GEMINI_KEY with your real key
do not paste your key here in Discord

That avoids the broken interactive paste step completely.

After it finishes, test with:

#

openclaw.cmd models status
openclaw.cmd dashboard

If you want, I can stay with you while you run that.

tired tundra Apr 15, 2026, 11:07 AM

#

do i put the key in " "

novel wigeonBOT Apr 15, 2026, 11:07 AM

#

yes, keep the quotes.

Put the key inside them, like this:

--gemini-api-key "AIza..."

So:

keep the "
replace only YOUR_NEW_VALID_GEMINI_KEY

Example shape:

openclaw.cmd onboard --non-interactive --mode local --auth-choice gemini-api-key --gemini-api-key "AIzaSyEXAMPLE..." --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

#

Just don’t paste the real key here in Discord.

tired tundra Apr 15, 2026, 11:10 AM

#

after i paste the command it says this :"
Non-interactive setup requires explicit risk acknowledgement.
Read: https://docs.openclaw.ai/security
Re-run with: openclaw onboard --non-interactive --accept-risk ...
PS C:\Users\Eosperus>"

novel wigeonBOT Apr 15, 2026, 11:10 AM

#

yep, that’s expected. Just add --accept-risk.

Run it again like this:

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice gemini-api-key --gemini-api-key "YOUR_NEW_VALID_GEMINI_KEY" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Same deal:

keep the quotes
replace only YOUR_NEW_VALID_GEMINI_KEY
don’t paste the real key here

After it finishes, run:

#

openclaw.cmd models status

tired tundra Apr 15, 2026, 11:16 AM

#

"🦞 OpenClaw 2026.4.14 (323493f) — I'll do the boring stuff while you dramatically stare at the logs like it's cinema.

Windows detected - OpenClaw runs great on WSL2!
Native Windows might be trickier.
Quick setup: wsl --install (one command, one reboot)
Guide: https://docs.openclaw.ai/windows
Config overwrite: C:\Users\Eosperus.openclaw\openclaw.json (sha256 eb5528b556eb730f3a5877a65a4fdaf38cb5731e58864bde8710c8622907f4b9 -> efae768443c9c62210e3a4a99cf288db060174551d7a3ed69cfc962ad5d92dcd, backup=C:\Users\Eosperus.openclaw\openclaw.json.bak)
Updated ~.openclaw\openclaw.json
Workspace OK: ~.openclaw\workspace
Sessions OK: ~.openclaw\agents\main\sessions
Gateway service install failed: Error: schtasks unavailable: schtasks produced no output for 5000ms
Tip: native Windows now falls back to a per-user Startup-folder login item when Scheduled Task creation is denied; if install still fails, rerun from an elevated PowerShell or skip service install.
Gateway service install did not complete successfully.
Run openclaw gateway status --deep for more detail." is that good?

novel wigeonBOT Apr 15, 2026, 11:16 AM

#

almost. The important part worked, but the Gateway service install failed.

So the situation is:

✅ config updated
✅ workspace/sessions look fine
❌ Windows auto-start service did not install

That’s not fatal. You can still use OpenClaw right now by running the gateway manually.

Do this next

First check status:

openclaw.cmd gateway status --deep

#

If it is not running, start it manually:

openclaw.cmd gateway run

Leave that window open, then in another PowerShell window run:

openclaw.cmd dashboard

Later fix for auto-start

If you want the service installed properly later, rerun PowerShell as Administrator and try:

#

openclaw.cmd gateway install

So: not fully good, but still workable.
Paste the output of openclaw.cmd gateway status --deep if you want and I’ll tell you the exact next move.

tired tundra Apr 15, 2026, 11:20 AM

#

🦞 OpenClaw 2026.4.14 (323493f) — Pairing codes exist because even bots believe in consent—and good security hygiene.

|
o
Service: Scheduled Task (registered)
File logs: ~\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Command: C:\Program Files\nodejs\node.exe C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789
Service file: ~.openclaw\gateway.cmd
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~.openclaw\openclaw.json
Config (service): ~.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: running (pid 7148, Startup-folder login item installed; listener detected on port 18789.)
Warm-up: launch agents can take a few seconds. Try again shortly.
RPC probe: failed
RPC target: ws://127.0.0.1:18789
timeout

Port 18789 is already in use.

pid 7148: "C:\Program Files\nodejs\node.exe" C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789 (127.0.0.1:18789)
Gateway already running locally. Stop it (openclaw gateway stop) or use a different port.
Listening: 127.0.0.1:18789
Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting

novel wigeonBOT Apr 15, 2026, 11:21 AM

#

that’s close, but not fully healthy yet.

Good news:

the gateway is installed
it’s running
it’s listening on 127.0.0.1:18789

Bad news:

the RPC probe timed out, so OpenClaw isn’t responding properly yet

Do this next

Wait about 15 to 20 seconds, then run:

openclaw.cmd gateway status

#

If it still says RPC probe: failed, restart it:

openclaw.cmd gateway restart
openclaw.cmd gateway status

If it still fails after that, run:

openclaw.cmd logs --follow

and paste the first error lines here.

You’re pretty close now.

tired tundra Apr 15, 2026, 11:22 AM

#

|
o
Service: Scheduled Task (registered)
File logs: ~\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Command: C:\Program Files\nodejs\node.exe C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789
Service file: ~.openclaw\gateway.cmd
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~.openclaw\openclaw.json
Config (service): ~.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: running (pid 7148, Startup-folder login item installed; listener detected on port 18789.)
RPC probe: ok

Listening: 127.0.0.1:18789
Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting

novel wigeonBOT Apr 15, 2026, 11:22 AM

#

perfect, that’s good now 🎉

The important line is:

RPC probe: ok

That means the gateway is healthy and responding.

18:23:24 [agent/embedded] embedded run agent end: runId=f287cbc5-37ce-494c-b902-0cc01d0e0990 isError=true model=gemini-3.1-pro-preview provider=google error=⚠️ API rate limit reached. Please try again later. rawError={"error":{"message":"{\n "error": {\n "code": 429,\n "message": "You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https:…
18:23:27 [agent/embedded] embedded run agent end: runId=f287cbc5-37ce-494c-b902-0cc01d0e0990 isError=true model=gemini-3.1-pro-preview provider=google error=⚠️ API rate limit reached. Please try again later. rawError={"error":{"message":"{\n "error": {\n "code": 429,\n "message": "You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https:…
18:23:31 [agent/embedded] embedded run agent end: runId=f287cbc5-37ce-494c-b902-0cc01d0e0990 isError=true model=gemini-3.1-pro-preview provider=google error=⚠️ API rate limit reached. Please try again later. rawError={"error":{"message":"{\n "error": {\n "code": 429,\n "message": "You exceeded your current quota, please check your plan and billing details. For more information on this error, head to: https:…

novel wigeonBOT Apr 15, 2026, 11:24 AM

#

yep, that means your Gemini key is valid now, but your Google Gemini quota/billing is exhausted.

So the problem changed from:

before: invalid key
now: valid key, but no available quota

What this means

This is a Google-side billing/quota issue, not an OpenClaw install issue.

Your options

Option 1, keep Gemini

check your Gemini / Google AI Studio billing and quota
wait for quota reset, or add billing/credits

Option 2, easiest way to get unstuck now

#

switch to another provider you already have quota for, like OpenAI or Anthropic

To switch providers:

openclaw.cmd configure --section model

Then pick a provider with a working paid key/quota.

So, honestly: OpenClaw is installed correctly now, but Gemini is out of quota. If you want, I can help you switch the model provider next.

tired tundra Apr 15, 2026, 11:25 AM

#

best model for me? i dont have money

novel wigeonBOT Apr 15, 2026, 11:25 AM

#

if your budget is basically $0, the honest answer is:

free tiers can work for testing, but they’re usually limited and unreliable
local/Ollama is technically possible, but for OpenClaw it’s usually not great on normal hardware
for the current cheapest reliable option, ask in #models because prices and free tiers change fast

Docs:

Models: https://docs.openclaw.ai/concepts/models
Community notes: https://docs.openclaw.ai/concepts/models#model-discussions-community-notes

So my recommendation is:

if your Gemini quota resets later, use that for basic testing
otherwise check #models for the current low-cost provider people are actually using successfully

If you want, I can also tell you the most realistic free-ish path vs cheap-but-works path.

tired tundra Apr 15, 2026, 11:27 AM

#

how do i change api keys

novel wigeonBOT Apr 15, 2026, 11:27 AM

#

easiest way:

Change your model API key

openclaw.cmd configure --section model

Change your search/web API key

openclaw.cmd configure --section web

Then:

pick the provider you want
enter the new key
finish the wizard

#

If you want to be safe afterward, restart the gateway:

openclaw.cmd gateway restart

Docs:
https://docs.openclaw.ai/cli/configure

If you want, I can tell you exactly which section to use for Gemini, OpenAI, or something else.

tired tundra Apr 15, 2026, 11:28 AM

#

help me change and paste the key directly

novel wigeonBOT Apr 15, 2026, 11:28 AM

#

yep, you can do it directly in PowerShell.

Use the one that matches your provider, and put the key inside the quotes.

Gemini

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice gemini-api-key --gemini-api-key "YOUR_KEY_HERE" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

OpenAI

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice openai-api-key --openai-api-key "YOUR_KEY_HERE" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Anthropic

#

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice apiKey --anthropic-api-key "YOUR_KEY_HERE" --gateway-port 18789 --gateway-bind loopback --install-daemon --skip-skills

Then check it with:

openclaw.cmd models status

If you want, tell me which provider you’re switching to, and I’ll give you just the one exact command.

tired tundra Apr 15, 2026, 11:32 AM

#

after i run the gemini api key change command it says this :"Windows detected - OpenClaw runs great on WSL2!
Native Windows might be trickier.
Quick setup: wsl --install (one command, one reboot)
Guide: https://docs.openclaw.ai/windows
Config overwrite: C:\Users\Eosperus.openclaw\openclaw.json (sha256 efae768443c9c62210e3a4a99cf288db060174551d7a3ed69cfc962ad5d92dcd -> 5c73fe955cb8e3307a16b579f7e53ca516d8b14ac62a1a115c1b95749a1d8862, backup=C:\Users\Eosperus.openclaw\openclaw.json.bak)
Updated ~.openclaw\openclaw.json
Workspace OK: ~.openclaw\workspace
Sessions OK: ~.openclaw\agents\main\sessions
Gateway service install failed: Error: schtasks unavailable: schtasks produced no output for 5000ms
Tip: native Windows now falls back to a per-user Startup-folder login item when Scheduled Task creation is denied; if install still fails, rerun from an elevated PowerShell or skip service install.
Gateway service install did not complete successfully.
Run openclaw gateway status --deep for more detail."

novel wigeonBOT Apr 15, 2026, 11:32 AM

#

that’s the same Windows service-install problem again, not necessarily a key problem.

The important part is:

Updated ~\.openclaw\openclaw.json → your config was updated
the failure is just the --install-daemon part on native Windows

Do this now

Check whether the new key works:

openclaw.cmd models status

and then try:

#

openclaw.cmd dashboard

Important

You already had the gateway installed/running before, so next time you change keys, do not include --install-daemon again.

If you need the direct Gemini command again later, use this version instead:

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice gemini-api-key --gemini-api-key "YOUR_KEY_HERE" --gateway-port 18789 --gateway-bind loopback --skip-skills

So for now, run openclaw.cmd models status and paste the result.

tired tundra Apr 15, 2026, 11:33 AM

#

🦞 OpenClaw 2026.4.14 (323493f) — I don't have opinions about tabs vs spaces. I have opinions about everything else.

Config : ~.openclaw\openclaw.json
Agent dir : ~.openclaw\agents\main\agent
Default : google/gemini-3.1-pro-preview
Fallbacks (0) : -
Image model : -
Image fallbacks (0): -
Aliases (1) : gemini -> google/gemini-3.1-pro-preview
Configured models (1): google/gemini-3.1-pro-preview

Auth overview
Auth store : ~.openclaw\agents\main\agent\auth-profiles.json
Shell env : off
Providers w/ OAuth/tokens (0): -

google effective=profiles:~.openclaw\agents\main\agent\auth-profiles.json | profiles=1 (oauth=0, token=0, api_key=1) | google:default=AIzaSyAc...aNnInOOg

OAuth/token status

none

novel wigeonBOT Apr 15, 2026, 11:34 AM

#

that output is good on the OpenClaw side.

It confirms:

OpenClaw has a Google API key saved
Gemini is set as the default provider
the config is being read correctly

What it doesn’t confirm is whether Google will actually let that key run requests without quota/billing errors.

Next step

Try the dashboard again with a tiny prompt, like:

reply with yes

If it still fails with the same 429 / exceeded quota error, then:

#

OpenClaw is configured correctly
your Google Gemini account/key still has no usable quota

At that point, your choices are:

wait / fix Gemini billing-quota
switch to another provider

So right now: test one more message in the dashboard and tell me whether it replies or throws the same 429.

tired tundra Apr 15, 2026, 11:35 AM

#

how do i fix gemini nilling quota

novel wigeonBOT Apr 15, 2026, 11:35 AM

#

if you mean Gemini billing/quota, that part is on Google’s side, not OpenClaw’s.

If you have no money

the honest answer is: you usually can’t force more quota. Your options are:

wait for Google’s free quota to reset
switch to another provider
try a local setup later if you have strong enough hardware

If you want to check/fix it on Google’s side

go to your Google AI Studio / Google Cloud setup and check:

is the Gemini API enabled?
is billing attached to the project?
does the project still have available quota?

Your error means one of these is true:

#

free quota is used up
billing is missing
billing exists but quota is exhausted

So the short version is:

OpenClaw is working
Gemini quota is the problem
if you don’t want to pay, you’ll probably need to wait for reset or switch providers

If you want, I can help you switch away from Gemini next.

tired tundra Apr 15, 2026, 11:35 AM

#

how do i know if "is billing attached to the project?
does the project still have available quota?"

novel wigeonBOT Apr 15, 2026, 11:36 AM

#

easiest way to check is in Google AI Studio / Gemini API project settings.

Check if billing is attached

Go here:
https://ai.google.dev/gemini-api/docs/billing

What to look for:

on the Projects page, check the Billing Tier / Plan / Status
if it says things like Paid or shows a billing plan, billing is attached
if it says Set up billing or similar, it is not attached

Check if quota is available

Go here:
https://ai.google.dev/gemini-api/docs/rate-limits

What to look for:

your project’s tier

#

whether you’re on a small/free tier
whether you’ve hit the current request/token limits

You can also check in Google Cloud Console:

APIs & Services
Quotas
filter for Gemini API or Generative Language API

If you’re seeing:

“You exceeded your current quota, please check your plan and billing details”

then usually one of these is true:

billing is not attached
free quota is used up
the project is on a too-small tier

If you want, I can help you interpret what you see on the billing page.

tired tundra Apr 15, 2026, 11:38 AM

#

i dont have money or account, what are other ways to setup

novel wigeonBOT Apr 15, 2026, 11:38 AM

#

the honest answer is: OpenClaw itself is free, but it still needs some AI backend.

If you have no money and no account, your realistic options are basically:

Run a local model with Ollama or LM Studio
Wait until you can use a provider free tier / account

Most realistic no-money path

Local Ollama

But I want to be straight with you:

it works best on a strong PC
on normal hardware, OpenClaw can feel slow or unreliable
it’s the main “no account, no money” route

Docs:

Ollama: https://docs.openclaw.ai/providers/ollama

#

Models: https://docs.openclaw.ai/concepts/models

What is not a real solution

there isn’t a magic fully-free built-in cloud model inside OpenClaw
most hosted options still need an account, even if they have a free tier

So if you want, I can help you do the local Ollama setup next, but I’ll want to know:

your CPU
your RAM
whether you have an NVIDIA GPU

tired tundra Apr 15, 2026, 11:39 AM

#

how to check cpu and ram

novel wigeonBOT Apr 15, 2026, 11:39 AM

#

on Windows, easiest way:

Option 1, Task Manager

Press:

Ctrl + Shift + Esc

Then:

click Performance
click CPU
click Memory

That will show your:

CPU model
total RAM

Option 2, PowerShell

#

Run these:

Get-CimInstance Win32_Processor | Select-Object Name
Get-CimInstance Win32_ComputerSystem | Select-Object TotalPhysicalMemory

If you want RAM in GB:

"{0:N1} GB" -f ((Get-CimInstance Win32_ComputerSystem).TotalPhysicalMemory / 1GB)

Paste me:

the CPU name
the RAM in GB

#

and I’ll tell you whether local Ollama is realistic.

tired tundra Apr 15, 2026, 11:40 AM

#

intel i5 7300U CPU
24gb ram

novel wigeonBOT Apr 15, 2026, 11:40 AM

#

honest take: 24 GB RAM is okay, but that i5-7300U is going to be rough for local OpenClaw.

So:

can it run local models at all? yes, technically
will it be a good OpenClaw experience? probably not
expect slow replies, weaker tool use, and a pretty frustrating setup

So I would not recommend local Ollama as your main path on that laptop unless you’re okay with “just experimenting”.

Your most realistic options are:

wait for a free cloud quota / account option
ask in #models for the current cheapest/free-ish setup people are using
try local anyway, but with low expectations

Docs:

#

If you want, I can still give you the best-effort local Ollama path for that machine.

tired tundra Apr 15, 2026, 11:41 AM

#

alright ill wait

novel wigeonBOT Apr 15, 2026, 11:41 AM

#

honestly, that’s a reasonable call.

The good news is:

OpenClaw itself is installed
your gateway is working
later, when you have a usable provider/account, you’ll mostly just need:

openclaw.cmd configure --section model

and then add the new key.

If you want fresh low-cost/free-ish advice later, check #models too.
When you come back, I can help you finish the last step fast.

tired tundra Apr 15, 2026, 11:46 AM

#

how do i change to my openai api keys, help me paste btw

novel wigeonBOT Apr 15, 2026, 11:46 AM

#

yep, let’s switch it to OpenAI directly, without the annoying paste prompt.

Run this in PowerShell, and put your real OpenAI key inside the quotes:

openclaw.cmd onboard --non-interactive --accept-risk --mode local --auth-choice openai-api-key --openai-api-key "YOUR_OPENAI_API_KEY" --gateway-port 18789 --gateway-bind loopback --skip-skills

Important:

keep the " quotes
replace only YOUR_OPENAI_API_KEY
do not add --install-daemon again

Then check it with:

#

openclaw.cmd models status
openclaw.cmd dashboard

If you want, after you run it, paste the models status output and I’ll confirm it switched correctly.

tired tundra Apr 15, 2026, 11:48 AM

#

Windows detected - OpenClaw runs great on WSL2!
Native Windows might be trickier.
Quick setup: wsl --install (one command, one reboot)
Guide: https://docs.openclaw.ai/windows
Config overwrite: C:\Users\Eosperus.openclaw\openclaw.json (sha256 5c73fe955cb8e3307a16b579f7e53ca516d8b14ac62a1a115c1b95749a1d8862 -> d5170463111905cc728938122344da76ca3463ea7a8696f354b94119747825cd, backup=C:\Users\Eosperus.openclaw\openclaw.json.bak)
Updated ~.openclaw\openclaw.json
Workspace OK: ~.openclaw\workspace
Sessions OK: ~.openclaw\agents\main\sessions
Gateway did not become reachable at ws://127.0.0.1:18789.
Last probe: gateway closed (1008): unauthorized: gateway token mismatch (provide gateway auth token)
Non-interactive local setup only waits for an already-running gateway unless you pass --install-daemon.
Fix: start openclaw gateway run, re-run with --install-daemon, or use --skip-health.
Native Windows managed gateway install tries Scheduled Tasks first and falls back to a per-user Startup-folder login item when task creation is denied.

novel wigeonBOT Apr 15, 2026, 11:49 AM

#

that actually makes sense.

What happened is:

your config got updated
the running gateway still has the old token
so now the CLI is hitting a token mismatch

Fix

Restart the gateway so it reloads the new config/token:

openclaw.cmd gateway restart
openclaw.cmd gateway status

Then check the model config again:

#

openclaw.cmd models status

and try:

openclaw.cmd dashboard

So the short version is: your OpenAI change probably wrote fine, but the gateway needs a restart.

tired tundra Apr 15, 2026, 11:51 AM

#

Service: Scheduled Task (registered)
File logs: ~\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Command: C:\Program Files\nodejs\node.exe C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789
Service file: ~.openclaw\gateway.cmd
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~.openclaw\openclaw.json
Config (service): ~.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: stopped (Startup-folder login item installed; no listener detected on port 18789.)
RPC probe: failed
RPC target: ws://127.0.0.1:18789
gateway closed (1006):

Service is loaded but not running (likely exited immediately).
File logs: C:\Users\Eosperus\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Logs: schtasks /Query /TN "OpenClaw Gateway" /V /FO LIST

Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting

novel wigeonBOT Apr 15, 2026, 11:52 AM

#

okay, the service is the part failing now.

Fastest way to get unstuck: run the gateway manually instead of fighting the Windows service.

Do this

In PowerShell:

openclaw.cmd gateway run

Leave that window open.

Then open a second PowerShell window and run:

#

openclaw.cmd models status
openclaw.cmd dashboard

What this means

your service is not staying alive
but OpenClaw can still work if you run the gateway manually

Important

If openclaw.cmd gateway run exits immediately or prints an error, paste that error here, because that will show the real cause.

tired tundra Apr 15, 2026, 11:53 AM

#

18:53:18 Gateway failed to start: gateway already running (pid 13152); lock timeout after 5000ms
If the gateway is supervised, stop it with: openclaw gateway stop
18:53:18 Port 18789 is already in use.
18:53:18 - pid 13152: "C:\Program Files\nodejs\node.exe" C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789 (127.0.0.1:18789)
18:53:18 - Gateway already running locally. Stop it (openclaw gateway stop) or use a different port.
18:53:18 Gateway service appears registered. Stop it first.
18:53:18 Tip: openclaw gateway stop
18:53:18 Or: schtasks /End /TN "OpenClaw Gateway"

novel wigeonBOT Apr 15, 2026, 11:53 AM

#

okay, good, we found the mess: there’s a half-running gateway process already holding the port.

Do a clean reset:

openclaw.cmd gateway stop
openclaw.cmd gateway status

If it then shows stopped / no listener, start it manually:

openclaw.cmd gateway run

If gateway stop does not clear it, use the Windows fallback it suggested:

#

schtasks /End /TN "OpenClaw Gateway"
openclaw.cmd gateway stop

Then try:

openclaw.cmd gateway run

So the immediate next command is:

openclaw.cmd gateway stop

tired tundra Apr 15, 2026, 11:54 AM

#

|
o
Service: Scheduled Task (registered)
File logs: ~\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Command: C:\Program Files\nodejs\node.exe C:\Users\Eosperus\AppData\Roaming\npm\node_modules\openclaw\dist\index.js gateway --port 18789
Service file: ~.openclaw\gateway.cmd
Service env: OPENCLAW_GATEWAY_PORT=18789

Config (cli): ~.openclaw\openclaw.json
Config (service): ~.openclaw\openclaw.json

Gateway: bind=loopback (127.0.0.1), port=18789 (service args)
Probe target: ws://127.0.0.1:18789
Dashboard: http://127.0.0.1:18789/
Probe note: Loopback-only gateway; only local clients can connect.

Runtime: stopped (Startup-folder login item installed; no listener detected on port 18789.)
RPC probe: failed
RPC target: ws://127.0.0.1:18789
gateway closed (1006):

Service is loaded but not running (likely exited immediately).
File logs: C:\Users\Eosperus\AppData\Local\Temp\openclaw\openclaw-2026-04-15.log
Logs: schtasks /Query /TN "OpenClaw Gateway" /V /FO LIST

Troubles: run openclaw status
Troubleshooting: https://docs.openclaw.ai/troubleshooting

novel wigeonBOT Apr 15, 2026, 11:55 AM

#

good, that means the stuck process is gone now.

Next step, start it manually:

openclaw.cmd gateway run

Leave that window open.

Then in a second PowerShell window run:

openclaw.cmd models status
openclaw.cmd dashboard

#downloading process

First, bypass the PowerShell script block

For the trim crash

My honest recommendation

Fix

Do this next

Later fix for auto-start

Do this next

Next

What this means

Your options

Change your model API key

Change your search/web API key

Gemini

OpenAI

Anthropic

Do this now

Important

Next step

If you have no money

If you want to check/fix it on Google’s side

Check if billing is attached

Check if quota is available

Most realistic no-money path

What is not a real solution

Option 1, Task Manager

Option 2, PowerShell

Fix

Do this

What this means

Important

For the `trim` crash