You bought a Sonos speaker. Maybe two. The app works fine, mostly, but you keep hearing about people who control their music through text messages. They type “play jazz in the kitchen” in Telegram and the speaker responds. No opening the Sonos app, no scrolling through playlists, no waiting for the network scan. Just a sentence and the music starts.
You want that. But you’ve never used OpenClaw. You’ve never installed a “skill.” You might not be entirely sure what an AI agent is or why it needs one. That’s fine. This guide assumes you’re starting from zero and walks through every step until you have working Sonos control through your messaging app. No steps skipped, no “just configure the environment” hand-waving.
If you’re already running an OpenClaw agent and just need to add Sonos, skip ahead to the “Installing the Sonos Skill” section. If you want the deeper dive into what’s possible once everything is set up — whole-home audio, room grouping, multi-device routines — read our complete Sonos journey article. This guide focuses on getting you from nothing to working Sonos control as quickly as possible.
What You Need
Before we start, here’s the full checklist. If you have all of these, the setup takes about fifteen minutes.
A Sonos speaker (or speakers) already set up on your Wi-Fi. The Sonos app handles initial speaker setup — connecting to your network, naming rooms, running Trueplay. If your speakers are already working with the Sonos app, you’re good. OpenClaw doesn’t replace the Sonos app for first-time hardware setup. It takes over from there.
A computer or server on the same Wi-Fi network as your Sonos speakers. This is where your OpenClaw agent will run. It can be a Mac, a Linux machine, a Raspberry Pi, or a Windows PC with WSL. The key requirement: it needs to be on the same local network as your Sonos devices, because the skill communicates with speakers over your home Wi-Fi. A laptop that’s always on your home network works. A desktop that’s always plugged in works even better.
A messaging app account. Telegram, WhatsApp, Discord, or Signal. This is how you’ll talk to your agent. Telegram is the most popular choice in the OpenClaw community because setup is straightforward, but any supported platform works.
About fifteen minutes. The process has a few steps, but none of them are complicated.
Step 1: Install OpenClaw
If you already have an OpenClaw agent running, skip to Step 3.
OpenClaw is the open-source AI agent framework that powers agents like Moltbot and Clawdbot. It has over 145,000 stars on GitHub and supports more than 3,500 skills. You’re going to install it, connect it to your messaging app, and then add the Sonos skill on top.
Our getting started tutorial has the full installation walkthrough with screenshots and troubleshooting. Here’s the condensed version.
On your computer, open a terminal and run:
pip install openclawThis installs the OpenClaw framework. If you don’t have Python installed, you’ll need to install that first — Python 3.10 or later.
After installation, initialize your agent:
openclaw initThis creates a configuration directory and prompts you to set up your first agent. Follow the prompts — you’ll pick a name, choose a messaging platform, and enter the credentials for connecting to that platform.
For Telegram, you’ll need a bot token from BotFather. Open Telegram, search for @BotFather, send /newbot, and follow the instructions. You’ll get a token that looks like a long string of numbers, a colon, and random characters. Copy it and paste it when OpenClaw asks.
Start your agent:
openclaw startTest it by sending a message in Telegram (or whichever platform you chose). Type “hello” and your agent should respond. If it does, the foundation is working.
Step 2: Verify Your Agent Is Running
Before adding skills, make sure the basics are solid. Send your agent a few test messages:
What can you do?What time is it?Tell me a joke.If your agent responds to all three, everything is connected properly. If it doesn’t respond, or if you get an error, check the terminal where you ran openclaw start — it will show error messages that help diagnose the issue.
Common problems at this stage:
- Bot token is wrong. Double-check the Telegram token. A single mistyped character will prevent the connection.
- Network issues. Your computer needs an internet connection to communicate with the messaging platform’s servers.
- Python version too old. OpenClaw requires Python 3.10+. Run
python --versionto check.
Once your agent is responding, move on.
Step 3: Install the Sonos Skill
This is the step where your agent learns to talk to Sonos speakers. One command:
clawhub install sonoscliThe sonoscli skill downloads and registers itself with your agent. No API key needed. No OAuth. No developer portal. The skill uses local network discovery to find Sonos devices, which means it talks directly to your speakers over Wi-Fi.
If your agent was already running when you installed the skill, restart it to load the new capability:
openclaw restartThat’s it for installation. No configuration files to edit. No environment variables to set. The sonoscli skill is one of the simplest to set up in the entire OpenClaw ecosystem because it works entirely over your local network.
Step 4: Discover Your Speakers
Now the fun part. Open your messaging app and send your agent this message:
What Sonos speakers are on my network?Your agent scans your local network for Sonos devices. After a few seconds, it responds with a list of every speaker it finds. Something like:
- Living Room — Sonos One, Idle
- Kitchen — Sonos Five, Idle
- Bedroom — Sonos Roam, Idle
Each speaker shows its room name (the name you gave it in the Sonos app during initial setup), the model, and its current status — idle, playing, or paused.
If you see your speakers listed, the connection is working. Your agent and your Sonos devices are on the same network and talking to each other.
If you don’t see your speakers: Check that the machine running your agent is on the same Wi-Fi network as your Sonos speakers. If your computer is on ethernet and your speakers are on Wi-Fi, they might be on different subnets — most home routers bridge these automatically, but some don’t. Also verify that your speakers are powered on and showing as available in the regular Sonos app.
Step 5: Your First Commands
Start simple. Pick a speaker and tell it to play something.
Play music on the Living Room speaker.The speaker starts playing whatever was last queued. If nothing was in the queue, it might start a default station or playlist depending on your Sonos setup.
For more control, specify what you want to hear:
Play my Liked Songs playlist on the Kitchen speaker.Play some jazz on the Living Room Sonos.The agent interprets your request and sends the appropriate command to the speaker. “Some jazz” gets translated into a jazz playlist or station available in your Sonos system.
Now try the basic controls:
Pause playback:
Pause the Kitchen speaker.Resume playback:
Resume the Kitchen.Adjust volume:
Set the Living Room volume to 30 percent.Relative volume:
Turn up the Kitchen a little.Skip a track:
Skip this song on the Living Room.Check what’s playing:
What's playing on the Kitchen speaker?Each command takes two to three seconds from the moment you send it to the moment the speaker responds. Fast enough that it doesn’t feel like you’re waiting. Slow enough that you won’t accidentally skip three tracks while fumbling with your phone.
Step 6: Grouping Rooms
If you have more than one Sonos speaker, grouping is where OpenClaw starts to feel genuinely better than the Sonos app.
Grouping means multiple speakers play the same music, perfectly synchronized. In the Sonos app, this requires long-pressing a room, dragging it into a group, and confirming. Through OpenClaw, it’s one sentence:
Group the Living Room and Kitchen speakers.Both speakers now play the same track at the same time. The audio sync is handled by Sonos — OpenClaw just tells the speakers to form a group.
To play something on the group:
Play dinner jazz on the Living Room and Kitchen, volume 25 percent.The agent groups the speakers (if they aren’t already), starts the playlist, and sets the volume. One command, two speakers, perfectly coordinated.
To ungroup:
Separate the Living Room and Kitchen.Each speaker returns to independent control.
If you have three or more speakers and want music everywhere:
Play my cleaning playlist on all speakers at 40 percent.Every speaker in your house starts playing the same track, grouped and synchronized. Perfect for Saturday afternoon cleaning sessions or house parties.
Step 7: Everyday Scenarios
Here are the commands you’ll use most often, based on what other beginners tell us they reach for first.
Morning routine:
Play the news on the Kitchen speaker, low volume.Working from home:
Play lo-fi beats in the office at 20 percent.Cooking dinner:
Play Italian dinner music in the Kitchen at 35 percent.Bedtime:
Play ambient sounds in the Bedroom at 10 percent.Leaving the house:
Stop all Sonos speakers.Guest asks for a song:
Play Abbey Road by The Beatles on the Living Room speaker.Each of these replaces a multi-step process in the Sonos app: open app, wait for network scan, select room, find content, adjust volume, hit play. Through OpenClaw, you describe what you want and the agent handles the rest.
Troubleshooting Common Issues
When something doesn’t work, it’s almost always one of these five things.
”No speakers found”
Your agent and your Sonos speakers aren’t on the same network. Check that the machine running OpenClaw is connected to the same Wi-Fi your speakers use. If your computer is on a guest network or a VPN, speakers won’t be visible. Disconnect from the VPN or switch to your main network and try again.
Speakers appear but won’t play
The most common cause is a music service that isn’t linked in the Sonos app. If you ask for a Spotify playlist but Spotify isn’t connected to your Sonos system, the command fails. Link your music services through the regular Sonos app first (Settings > Services & Voice > Music & Content), then try again through OpenClaw.
Volume is extremely loud or quiet
When you don’t specify a volume, the agent uses whatever the speaker’s current volume is set to. If the last person used it at 80%, and you say “play music,” it plays at 80%. Get in the habit of including a volume level: “Play jazz at 25 percent.” You can always adjust after: “Turn it down.”
Commands work but are slow (10+ seconds)
The usual cause is network congestion or the agent running on a device with limited resources. A Raspberry Pi Zero will be slower than a laptop. If your Wi-Fi is crowded with many devices, that adds latency too. For most home setups, response time should be 2-4 seconds.
Agent doesn’t understand the command
The sonoscli skill handles natural language, but very unusual phrasing might confuse it. If “initiate audio reproduction on the primary acoustical device” doesn’t work, try “play music on the Living Room.” Use the room names exactly as they appear in the Sonos app. If your speaker is named “Living Room” in Sonos, use “Living Room” in your commands, not “lounge” or “main room.”
Going Further
Once the basics feel comfortable, there’s more to explore.
Combine Sonos with other smart home skills. If you have smart lights or a smart TV, add the corresponding OpenClaw skills and control everything from one conversation. “Movie night — stop the music, dim the lights, turn on the TV.” Read OpenClaw Smart Home Automation for the full picture.
Explore advanced Sonos control. Our whole-home audio article covers multi-room orchestration, zone-based control, and scaling from one speaker to six or more. It picks up where this beginner guide leaves off.
Browse more skills. Sonos might be your first OpenClaw skill, but there are 3,500+ available. The smart home category has more IoT integrations. The getting started tutorial shows you how to find, evaluate, and install skills across every category.
FAQ
Do I need to keep my computer running all the time for Sonos control to work?
Yes. Your OpenClaw agent needs to be running to receive and process your commands. When the agent is off, text-based Sonos control stops working (the regular Sonos app still works, of course). Many users run their agent on a Raspberry Pi or an always-on desktop to ensure it’s available whenever they need it.
Will this work if I only have one Sonos speaker?
Absolutely. Grouping and multi-room features require multiple speakers, but every other function works perfectly with a single speaker. Play, pause, volume, track skipping, and content selection all work with just one device. You can always add more speakers later and the skill will find them automatically.
Does controlling Sonos through OpenClaw affect the regular Sonos app?
No. Both work simultaneously. You can start music through OpenClaw and adjust volume through the Sonos app, or vice versa. They’re two different interfaces to the same speakers. Whatever you change in one is reflected in the other.
What music services work with OpenClaw Sonos control?
Any music service that’s linked to your Sonos system. Spotify, Apple Music, Amazon Music, YouTube Music, Tidal, Deezer, TuneIn Radio, and others. The OpenClaw skill sends commands to the Sonos speaker, which plays music through whatever services you’ve connected in the Sonos app. You don’t need to configure music services separately in OpenClaw.
Can I set up timed music, like a sleep timer?
You can ask your agent to stop music after a certain time. “Stop the bedroom speaker in 30 minutes” works as a sleep timer. The agent schedules the stop command and executes it at the right time. For more complex scheduling, combine sonoscli with a workflow skill like flowmind.
Summary
Here’s the complete path, condensed:
- Install OpenClaw on a computer on your home Wi-Fi
- Connect it to Telegram (or your messaging app of choice)
- Install sonoscli with
clawhub install sonoscli - Discover speakers by asking your agent what Sonos devices are on your network
- Start playing by typing what you want to hear and where
- Group rooms when you want synchronized multi-room audio
- Troubleshoot by checking network connectivity if speakers aren’t found
From this point forward, controlling your Sonos system is as simple as sending a text message. No app to open, no UI to navigate, no loading screens to wait through. Just tell your agent what you want to hear, and it plays.
clawhub install sonoscliFifteen minutes of setup. Years of “play jazz in the kitchen” ahead of you.