How to

NOISEY posts to social apps from real phones. The phones are plugged into your own computer. It does not use an API, and it does not use a browser pretending to be a phone. You drop a video into a folder, and a rule you wrote posts it through the app on the phone you picked.

NOISEY has two parts. The desktop app on your computer controls the phones. This website holds your folders, automations and media, and tells the desktop app what to do.

The desktop app does the work on the phones. It mirrors their screens, taps and types on them, and runs your routines. It is the only thing that ever touches a phone. NOISEY's servers never do.

The website (my.noisey.app) is the control panel. It holds your folders, the webhook that feeds them, your automations, your activity and your account. When an automation fires, the server queues a job and sends it to your desktop app. The desktop app does the posting.

Go to my.noisey.app/signup. You need a name, an email and a password. Or press Continue with Google, which skips both the password and the confirmation email. Nothing else is collected, and we never ask for a card.

Confirm your address before you sign in for the first time. Look in your inbox for a message called Confirm your email address and click the link. Until you do, signing in says “confirm your email address first”.

No email? Check your spam folder. Then sign up again with the same address to get a fresh link. Google sign-ins skip this step.

The download button on noisey.ai signs you in first. It then takes you to the Get the desktop app card on your dashboard, which has the Mac and Windows installers. You sign in first because every copy of NOISEY belongs to an account, and your plan's limits are counted per account.

First time on a Mac: right-click NOISEY in Applications, choose Open, then confirm. You only do this once. Your Mac warns you because NOISEY isn't notarised with Apple yet.

First time on Windows: when you see “Windows protected your PC”, click More info, then Run anyway. You only do this once. The installer isn't code-signed yet. It installs for you alone and never asks for administrator rights.

• A Mac on macOS 13 or newer (Apple silicon or Intel), or a Windows PC on Windows 10 or 11 (64-bit).
• One or more Android phones, and a USB cable for each one. Use a powered hub once you have more than a few.
• An account, which is free to start.

iPhones are not supported.

1. Create an account and download NOISEY.
2. In the desktop app, choose Connect this computer and approve the code at my.noisey.app/link.
3. Plug in a phone with USB debugging on and accept the trust prompt on its screen.
4. Teach the desktop app a routine, such as how you post a video to TikTok. Give it everything it needs so it can run on its own.
5. On the website: make a folder under Media, create your webhook, then under Automations connect folder → routine → phone.
6. POST a video to the webhook with ?folder=your folder. Then watch Activity.

1. In the desktop app, press Connect this computer. The button sits in the top bar, and fills the first screen, whenever the app is signed out. A short code like ABCD-2345 appears and the app waits.
2. Your browser opens my.noisey.app/link. Sign in if you aren't already. The code carries through.
3. The page names the computer that is asking. If it's yours, press Connect it. The app signs itself in within a few seconds.

The app never asks for your password. So changing your password later, or adding Google sign-in, changes nothing in the desktop app.

Codes expire on purpose after a few minutes. Go back to the desktop app and ask it to connect again. It makes a fresh code straight away. Nothing about your account has changed.

Press I didn't ask. A code only reaches you because someone started the connect flow and sent it to you. If that someone wasn't you, approving it signs their computer into your account. They would get your phones, your media and your automations.

Declining kills the code for good. Nothing is connected, and the code cannot be used again.

Settings → Your computers lists every computer connected to the account, and when each was last seen. Sign it out works straight away: that computer stops syncing and stops getting jobs. It keeps controlling the phones plugged into it. Signing out cuts the account link, not the cable.

Yes, on purpose. Controlling phones plugged into your own computer never needs a server. Mirroring and manual control keep working with no connection at all.

The cloud half pauses: sync, media sent to your webhook, and jobs queued by the server. Those wait on the server and are sent to you when the app is back online.

1. Enable USB debugging on the phone (next question).
2. Plug it into the Mac running NOISEY.
3. On the phone's screen, accept Allow USB debugging? Tick Always allow from this computer so it lasts after you unplug.

The phone appears in the desktop app's grid and registers itself to your account. A moment later it also shows under Phones on the website.

1. On the phone, open Settings → About phone and tap Build number seven times. It tells you that you are now a developer.
2. Back in Settings, open Developer options and switch USB debugging on.

Menu names vary a little between makers, but tapping Build number seven times works on every Android phone.

The phone hasn't trusted your computer yet. Unlock it and look for the Allow USB debugging? prompt. It is easy to miss, and unplugging and replugging the cable brings it back. Tick Always allow from this computer and accept. If the prompt never shows, open Developer options, revoke USB debugging authorisations, and replug.

• online - connected to your computer and ready.
• busy - a job is running on it right now.
• offline - registered to the account, but not plugged into a computer running NOISEY right now.
• unauthorized - plugged in, but the phone hasn't accepted the USB debugging prompt.
• error - the desktop app can see it but cannot control it. Replugging usually fixes it.

Until a phone has a profile, it is named after what it is: its model, or its hardware id if the model is unknown. Give the phone a profile in the desktop app and it is named after who it posts as.

The free plan allows 2 phones. The paid plan allows 10. NOISEY counts each phone once, not each cable or address, so replugging, hubs and address changes never count a phone twice. The limit applies when a new phone registers, and the desktop app tells you when you reach it.

Under Phones, click a phone's name to rename it. Or press Name all phones: every unnamed phone takes the next name from a fixed list - Onyx, Ember, Indigo, Basalt and so on. The order is always the same, the names are easy to remember, and you can change any of them later.

There is one name everywhere. The desktop app shows the same word on the same tile, and renaming it there updates it here. “Post to Onyx” works. “Post to the ninth SM-G986U” does not.

A named set of phones that an automation treats as one target, such as “Morning rack”. When the automation fires, every phone in the group gets the job, spaced out by the gap you set.

You make a group where you need it. In the automation form, choose Several phones…, tick the phones, and name the group. It is saved, and later automations can use it.

A queue with a name. Files land in a folder through your webhook or a watched Drive folder, and automations watch folders. That link between folder and automation is the whole routing model: where a file lands decides what happens to it.

Press Media → New folder. You can also make one while creating an automation. A folder is just a name. There is nothing else to set.

Two ways today:

• Your webhook - anything that can POST a file sends it to your webhook URL and names the folder. n8n, Zapier, Make, a script or a render pipeline all work.
• A watched Google Drive folder - our servers check it every few minutes and pull in new videos.

There is no upload button yet, on purpose. NOISEY connects where your videos are made to the phone that posts them.

One URL, unique to your account, that accepts video uploads:

POST https://api.noisey.app/webhooks/media/<your-token>?folder=Morning posts

Create it under Media → Your webhook. The long token in the URL is the password. Anyone with the URL can send files to your account, so keep it secret.

Name it in the request. Add ?folder= with the folder's name; upper and lower case don't matter. For a folder inside another folder, use the path, like ?folder=Clients/Acme. You can use the folder's id instead, or send an x-folder header. One webhook serves every folder, so an n8n or Zapier flow feeding three folders needs one secret and one field, not three URLs.

An unknown folder name is refused with a clear error. NOISEY will not quietly make a folder that no automation watches, so make the folder in the dashboard first. Leave ?folder= off and the file lands with no folder.

Press Media → Your webhook → Copy URL. You can copy the working URL there at any time. It is stored encrypted with the server's own key, so a stolen database alone gives nothing away. Older webhooks, made before copying existed, say so on their row. Recreate those once and you can copy them from then on.

Press Media → Your webhook → Revoke. It works straight away: the old URL answers 404 on the next request. Then create a new webhook and update the tools that use it. Files you already sent are safe. A webhook token can only add files. It can never read or delete anything.

That is on purpose. NOISEY identifies a file by its contents. If it has seen those exact contents before, it recognises the file instead of taking it again: the webhook answers with "deduped": true and no automation fires. This stops a Drive re-sync or a retrying pipeline from posting yesterday's video to twenty phones twice. A file that is genuinely new, even by one frame, counts as new and triggers as normal.

NOISEY moves videos to your phones. It is not storage. On the free plan a file lives 24 hours from upload, so post it before then. Each file shows its own countdown under Media. The paid plan ($29/month) keeps files 90 days. Your plan also limits how many files you can hold at once: 30 free, 500 paid. Keep your originals where you made them.

Yes. When you create a folder, choose which folder it sits inside. Folders inside folders show as paths, like “Clients / Acme”. The webhook reaches one the same way: ?folder=Clients/Acme. Each folder is still its own queue, and an automation watches exactly one.

Press the bin icon on the folder's card. Folders inside it go too. Files are not deleted. They move to “No folder” and keep their own expiry clock. NOISEY refuses to delete a folder that an automation watches, directly or through a folder inside it, and it names the automation. A tidy-up can never quietly break the thing that posts at 9am.

Yes. Every file under Media has a download link. It gives you the exact file that was uploaded.

A standing rule: when a file lands in this folder, run this routine on this phone or group, using the new file. As soon as your webhook or a Drive sync puts a new file in the folder, the automation queues one job for each target phone. You don't have to click anything, or be at your desk.

1. Automations → New automation.
2. Pick the folder that triggers it, or create one right there.
3. Pick the routine to run. Only routines that can run unattended are offered.
4. Pick one phone, an existing group, or Several phones… to make a group on the spot.
5. For groups, set the seconds between phones, rename the group if you want, and press create. It starts turned on.

Three things: a folder, at least one registered phone, and a routine that can run on its own. Folders take seconds to make on the website. You teach routines in the desktop app, and they sync up on their own once your computer is connected. The Automations page tells you which of the three is missing and where to get it.

It goes through the automation's folder now, newest files first, up to a set limit. It queues jobs as if those files had just arrived. It tells you what it queued. If the limit left older files behind, it says how many.

Three explanations, in order of likelihood:

• This automation has already queued or posted every file in the folder on those phones. The same video is never posted twice (see below).
• The folder is empty.
• The automation is paused. Resume it first.

Not by the same automation on the same phone. NOISEY tracks each automation, file and phone together. If that combination already has a job that is queued, running or successful, it is skipped. This is true for webhook triggers and for Run now. There is one exception: a failed job can be retried, because the post never happened.

When an automation fires on a group, each phone's job starts that many seconds after the one before it, instead of all at once. Twenty phones uploading at the same time through one computer causes a traffic jam. The gap protects your own bandwidth and your own computer. That is all it is for.

Both buttons are on the automation's card. Pause keeps the rule but ignores new files until you resume. Files that arrive while it is paused are not queued later. Delete removes the rule. Jobs that already ran, and their history, are left alone.

One on the free plan, which is enough to test the whole loop. Unlimited on the paid plan. The limit counts standing rules, not how often they fire.

In the automation form, change the trigger from When a file lands to On a schedule. Pick the times, the weekdays, and the folder to take files from. Each run posts the newest file that phone hasn't posted yet. That is one file per phone per run, so a daily schedule posts once a day and never repeats.

Times use your timezone. The card shows the next run, so you can check the schedule before it matters.

It is the latest a run may start. “9:00, window 30” runs at 9:00 when everything is ready. If your computer is catching up, it may start as late as 9:30. After that the run is marked missed, rather than posting something stale at 14:37. It is not a randomiser. Start times are never shuffled inside the window.

The start window closed before your computer was connected and free. Nothing broke and nothing posted. The file is still there, and the next scheduled run picks it up. If runs keep getting missed, your Mac isn't open at those times, or the window is too short. Make the window longer on the automation.

Check each step. All of them are visible:

1. Media: did the file land, and in the folder the automation watches? A typo in ?folder= is refused by the webhook, and the tool that sent it logged the error.
2. Was it genuinely new? Identical bytes are deduped and fire nothing.
3. Automations: is the rule enabled, and watching that exact folder?
4. Activity: is there a job? If it says queued, the server has done its part and is waiting for your desktop app to be open and connected. If it says failed, the error is printed under it.

Yes. Open the routine in the desktop app and find What it should tell you. Name each thing you want read off the phone, such as replyCount or postUrl, and describe it in your own words. When the routine runs, those values come back and show under the run.

A run can only fill in the fields you named. It cannot send back something you did not ask for. That is deliberate, because the screen it is reading might be showing a client’s messages. Anything that looks like a one-time code is thrown away on your own computer before it is recorded.

Treat these as what the phone appeared to say, not as a measurement. A model read the value, and nothing can tell a number that was read from one that was guessed. That is why every value shows where on screen it came from. Check that before you build something that depends on it.

Make a flow on the website and set its trigger to something happens on a phone. Enter the id of the routine you care about to narrow it down. A flow with nothing filled in fires on every phone event, which is rarely what you want.

Add an HTTP step pointing at your own address. Anything the routine reported is available by name, so a flow can use {{replyCount}} directly instead of digging through a list.

It fires once, when the routine has finished, on the real result. It does not fire on each retry. A routine that ran while your computer was offline reports when your Mac reconnects, so a flow can be late but is never lost.

Open a phone in the desktop app and press the red Record button next to Claw. Do the task on the screen in front of you, press Record again, and give it a name. It is saved as an ordinary routine.

If you repeat the same gesture three or more times, it becomes one step with a count you can change afterwards. Record ten swipes, then change the number to thirty. That is how it is meant to work, and no model runs when it replays.

Anything you type while recording is never saved. The routine asks you for it when it runs, so a recording can include a WiFi login without storing one. A recorded routine is tied to the phone you recorded it on, because it replays exact positions. Widen it in Routines once you have watched it work.

A task you teach the desktop app in plain words: “open TikTok, start a new post, pick the newest video, write the caption, post it”. NOISEY works out the taps and typing on the real phone. Automations and schedules then run that routine without you.

In the desktop app, with a phone connected, describe what you want done and run it. Fix the instructions until it works, then save it with a name. It syncs to your account on its own. A moment later the website offers it when you create an automation or a schedule.

Yes. In the automation form, press Write one next to the routine picker. Give it a name and write the steps in plain words, one per line. Your Mac picks it up the next time it connects. Until then the routine is labelled waiting for your computer, and anything scheduled against it waits too.

Write real values, not placeholders. A scheduled run has nobody to ask. The first run on the phone works everything out as it goes. Let it learn in the desktop app and later replays are free.

A learned routine has been turned into exact steps. Replaying it is fast and costs nothing. That is “(free)”.

An unlearned routine asks the AI to work the task out again on every run. That is flexible, but it is slower and it costs model usage each time. Run the routine successfully in the desktop app and let it learn. It then becomes the free kind.

Because it asks a question while it runs. A routine with an input set to ask each run needs a person to answer, and nobody is there at 9am. So it cannot run on its own and is left out of the list. The page says how many were hidden and why. In the desktop app, give that input a fixed value, or take the value from the incoming video. The routine can then be scheduled.

Routines live on the Mac that taught them. The same routine name on two Macs is really two files, and they can differ. The computer name, as in “Post to TikTok - Studio Mac”, tells you which computer does the work. That matters once you run two.

In the desktop app, where you made it. Edit the instructions or the inputs and save. The website picks up the new version through the same sync that delivered it. There is no routine editor on the website, on purpose: you can only test a routine on a real phone, and the phones are plugged into your computer.

1. Add an HTTP Request node at the point where your flow has the finished video as binary data.
2. Set the method to POST. Set the URL to your webhook URL plus ?folder=Morning posts. The folder is the only routing choice, so use an expression if your flow feeds more than one folder.
3. Set the body content type to n8n binary file, pointing at your binary property.
4. Add a header x-filename with the file's name. This is the name shown in Media.

The node's response tells you what happened. jobsQueued above zero means phones are about to post.

Same shape as n8n. Use a raw HTTP action: in Zapier, Webhooks by Zapier → Custom Request; in Make, HTTP → Make a request. POST the file as the body to your webhook URL, with ?folder= naming the folder and an x-filename header. Send the file itself as the raw body. Do not wrap it in a form or in JSON.

curl -X POST \
  -H "content-type: video/mp4" \
  -H "x-filename: sunrise-take-1.mp4" \
  --data-binary @sunrise-take-1.mp4 \
  "https://api.noisey.app/webhooks/media/<your-token>?folder=Morning posts"

Use --data-binary. Plain -d damages binary data. The Media page gives you this exact command, filled in, when you create your webhook.

Request:

• POST with the file itself as the raw body. It is streamed, so large videos are fine.
• content-type - the file's type, e.g. video/mp4.
• x-filename - what to call it. Optional; without it NOISEY makes a name from the date and time.
• ?folder= (or x-folder) - the folder name or id. Optional; without it the file has no folder.

Response, on success:

{
  "media": { "id": "…", "filename": "…", "size_bytes": 1234567, … },
  "automationsTriggered": 1,
  "jobsQueued": 2
}

A file NOISEY has seen before answers "deduped": true and triggers nothing. An unknown ?folder= gives a 404 with a message naming the folder. A revoked or mistyped token gives a 404 that says so.

1. Media → Watch a Drive folder (at my.noisey.app/media).
2. Connect the Google account the videos live in, and allow Drive when asked.
3. Browse to the folder and pick it. You choose it by looking at it, not by pasting an id. Name the NOISEY folder it feeds, then confirm.

New videos in that Drive folder are pulled into the NOISEY folder you chose. There they trigger automations, just like files sent to your webhook.

They are different permissions, and you should only grant what you use. Signing in shares your name and email. NOISEY asks for Drive access the moment you connect a folder, and not before. It reads the folders you browse and downloads the videos it takes in. It never writes to your Drive, and it never deletes anything.

Google sometimes cancels a connection, usually after a password change or a security review on their side. The source keeps its settings and shows the error clearly instead of failing quietly. Open Sources, reconnect Google when asked, and syncing carries on where it stopped.

Our servers check every few minutes, so files come in and jobs are queued with your computer closed. The posting still needs the desktop app open with the phones plugged in. Queued jobs wait, and are sent the moment it connects.

The person a phone posts as. A profile holds a name, a handle, a picture, the platform, and an optional brief about the account holder. Give a phone a profile in the desktop app and it stops being called SM-G986U1 everywhere. Profiles live at my.noisey.app/profiles.

Schedules belong to profiles. Open Profiles, pick the profile, and add times and weekdays. Each slot names the routine to run and the folder to take files from. It runs on whichever phone carries the profile, in your timezone.

Like automations, a schedule needs a routine that can run on its own, and the profile has to be on a phone. The page shows the next run, so you can see it is set.

A short description of the real person who holds the account. You write it; NOISEY never generates one. Routines use it for context, so a fitness coach's captions don't read like a barber's. It is optional, and it only ever describes a real account holder.

Go to Settings → You. Your browser crops the picture to a square and shrinks it to 128px before saving, so nothing large leaves your computer. It is only ever shown small. The name is whatever you want to be called.

Go to Settings → Email, type the new address, and press Change. A confirmation link goes to the new inbox. Nothing changes until you click it, so a typo cannot lock you out. After that, sign in with the new address.

Go to Settings → Password. You need your current password. The sign out my other sessions box is ticked by default, because changing a leaked password should also kick out whoever has it. Your current browser stays signed in either way.

Yes. Settings → Password spots a Google-only account and offers to email you a link to set a password. Once it is set, both ways in work. That helps on the day Google sign-in is unavailable.

my.noisey.app/forgot-password emails you a reset link. If no email arrives, check your spam folder, and check you typed the address the account was made with. The page never says whether an address exists, on purpose.

Go to Settings → Sessions → Sign out everywhere else. Every browser session except the one you are using ends straight away. Desktop apps are separate. They are listed under Your computers on the same page, each with its own Sign it out.

You cannot delete it yourself yet. Email support@noisey.app from the account's own address. We will remove the account, along with its media, phones and automations.

2 phones, 30 stored uploads, 1 automation, and 7 days of activity history. Files live 24 hours from upload. That is enough to run the whole loop on a small setup, as long as you post what you take in. Every account starts free, and we never ask for a card at signup.

$29/month raises the limits to 10 phones, 500 stored uploads, unlimited automations and 90 days of history. Files are kept 90 days instead of 24 hours. You cannot upgrade inside the app yet, so every account currently runs on the free plan.

The action is refused the moment you try it, with a clear message such as “automations limit reached (1/1)”. Nothing you already have is changed. Limits are checked on the server, so they read the same everywhere.

• An unconfirmed account is asked to confirm first. Check the message above the form, and your inbox.
• Very strict browser privacy settings block all cookies, including the one that signs you in. Allow cookies for noisey.app.
• If the page says it could not reach the server, that is our problem, not yours. Try again shortly.

Google sign-in is switched off on the server for now. The email and password form below it still works, and an account made with email can add Google later. If you only ever signed in with Google, use Forgot your password to set one and get in that way.

Queued means the server is holding jobs for a desktop app that is not there. Open NOISEY on the Mac the routine lives on, signed in to the same account, with the target phones plugged in and trusted. Jobs go out within seconds of it connecting. Check that Settings → Your computers shows that computer as recently seen.

Activity prints the error under the job, in red, with the number of attempts. Failures are never hidden. A failed post can be queued again, because the double-post guard skips failures on purpose. Fix the cause and run the automation again to retry it.

• Swap the cable. Charging-only cables are common.
• Check USB debugging is on and the trust prompt was accepted. Unlock the phone, then replug it to bring the prompt back.
• Try another port, or plug in directly instead of via a hub.
• If the desktop app sees it but the website doesn't, check that the app is signed in. The phone registers to your account through the app.

The error body says which of the two it is:

• unknown or revoked token - the URL is mistyped, no longer valid after a revoke, or cut short by the tool sending it. Create a fresh webhook and paste it carefully.
• no folder called “…” - the ?folder= value matches no folder on your account. Create the folder first, or fix the spelling. Upper and lower case don't matter, but typos do.

The desktop app has a bug button at the foot of its sidebar. Reports sent there arrive with the app version and platform attached, which saves a round of questions. For anything on the website, email support@noisey.app with what you did, what you expected, and what happened instead.

No, and it is built that way. Phones are plugged into your computer, and only the desktop app can control them. The server stores rules and media, queues jobs, and hands them to your app. A hacked server still could not tap a screen. There is no route from the server to a phone that does not go through software on your own computer.

The token is made with a secure random generator. NOISEY checks requests against a stored hash, the same way it handles passwords. The copyable value is kept encrypted, so Copy URL works without weakening that. A leaked token can only add files to your account. It can never read, delete or post. Revoking it works from the next request.

The desktop app never handles your password. It shows a short-lived code that you approve while signed in on the website. The approval page starts with the name of the computer that is asking, and makes I didn't ask as easy to see as the approve button. The only real attack is talking you into approving someone else's code, so the page puts that question in front of you.

Your account details: your name, your email, and either a hashed password or your Google identity. Your folders, automations, profiles and schedules. The videos you upload, for as long as they live. Job history, for as long as your plan keeps it. For each phone, an id, a model and a status, and never the contents of the phone. If you grant Drive access, we use it to read the folder you picked and nothing else. The privacy policy lists everything, how long we keep it, and your rights over it.

At noisey.ai/terms, and the privacy policy at noisey.ai/privacy. You agree to them when you create an account. In short: NOISEY is for research, testing, and automating phones and accounts you own or are authorised to manage. What your automations do is your responsibility, as with any automation tool.

You are. NOISEY runs the routines you write, on your phones, on your accounts. We do not review, approve or control them, and we cannot see why you run them. Your relationship with the platforms you post to is between you and them. Their rules apply to you directly, and any action they take, including restricting your account, is yours to deal with. The terms of service spell this out.

A routine is what a phone does: tap this, type that, check it worked. You draw it on a canvas at Routines.

An automation decides when a routine runs: when a file lands in a folder, or at nine every weekday.

A flow is everything around that. Something happens, a model reads it, and one or more things follow. Running a routine is one thing it can do. Calling a webhook or writing to your timeline are others.

They stay separate so the simple case stays simple. “Post what lands in this folder” should not require drawing a diagram.

Routines → Draw a routine. Pick steps from the palette on the left, then drag from the dot under one box to another to say what happens next. The box marked Start is where it begins.

Saving checks the wiring. It refuses a line pointing at nothing, a box nothing reaches, or a step this version cannot run. Better to find out now than at three in the morning on a phone.

Use Pick it off the phone in the inspector. With a phone connected, it shows that phone’s screen as it is right now and outlines everything NOISEY can identify. Click the thing you mean. It then works out the most reliable way to find it again: its id first, then its accessibility label, then its text. If the best name available is a generic one, it tells you. A name like button2 matches whichever button happens to sit second, so it will press something else on the next phone.

Yes. Press Run it once at the bottom of the editor. It saves, then runs on the phone you choose. The boxes colour in as it goes: green where it worked, red where it stopped. The step that stopped carries a picture of the screen at that moment, which is usually the whole answer.

Add an Ask a model step. It uses your own AI account. NOISEY does not resell model access, so you pay your provider directly and nothing else.

Your computer keeps its key in the system keychain: the macOS Keychain, or Windows Credential Manager. Flows run on our servers so they keep working overnight, and our servers cannot reach that keychain. So you can also store a key with us under Settings. It is encrypted, never shown again, and one click deletes it. Without one, a flow stops when it reaches an AI step.

Each account has a monthly limit on AI steps. NOISEY checks it before each call, so you never find out from a bill. Settings shows how many you have used.

Yes. Use the Call a URL step. It calls public addresses only. A flow cannot be pointed at localhost, a private network range or a cloud metadata address. The request comes from our servers, not your browser, so allowing those would let one account reach things it should not.

It spots engagement. Without it, a flow can act on a schedule or a file. With it, a flow can react when somebody comments on your client’s post. It is an Android app you install on your own phones over adb.

It reads notifications from six apps: LinkedIn, TikTok, Instagram, X, Facebook and YouTube. That list is built into the app. No setting can add to it, and neither can we or anyone else. Only a new version that you choose to install can change it.

It will never read:

• messages and SMS, including WhatsApp, Telegram and Signal;
• your dialer, or mail;
• banking, payment and wallet apps;
• authenticators and password managers;
• Android’s own system notifications.

The blocked list is checked before the allowed list, so a mistake in one cannot open the other. Anything that looks like a verification code is thrown away on the phone, before it is written down or sent anywhere. The app can only read. It cannot reply to a notification, open it, or dismiss it.

Install the APK over adb, open it, and press the button that opens Android’s own Notification Access screen. You grant the permission there, by hand. NOISEY will not grant it for you. An app that quietly gives itself notification access is not one you should trust.

Take the permission away the same way, in Android Settings, at any time. The app shows whether access is granted, and how many notifications are waiting for your computer to collect.

The app has no internet access at all. It writes to a file on the phone. Your computer collects that file over the same cable it already uses to control the phone, then passes it to NOISEY so your flows can react. Contents are deleted after 72 hours.

You get a copy on your own account. It is yours from then on. If the author publishes a newer version, what is already running on your phones does not change until you install the update. Anything the author filled in for their own account arrives blank, so set your own values before you run it. An installed flow arrives switched off.

The steps, and nothing you filled in. Before you publish, NOISEY shows you exactly what it will remove:

• any text a step types, unless it is a {{placeholder}};
• the values you set for the routine’s inputs;
• the phones, groups and folders it was pointed at;
• headers on any webhook step, which often carry keys.

NOISEY refuses a webhook URL that looks like it holds a token, rather than quietly emptying it. Publishing it would publish the token. Emptying it would ship a flow that looks complete and does nothing.

The same line the product itself draws: nothing built to help automation avoid being detected or enforced against. Staff remove anything that breaks that rule, and repeat offenders lose the right to publish. Report anything that crosses the line, or email support@noisey.app.

A forum inside the app. It has threads and replies under a few topics: announcements, automations & routines, phones & hardware, feature requests, and show & tell. It is where people who run NOISEY compare notes. You must be signed in to read or post. There is no public, anonymous side.

Your display name and the picture from your settings. Never your email address. If your account has no name yet, the composer asks you to pick one before your first post. Change it any time in Settings, and your posts show the current name.

Short list, firmly held:

• Keep it lawful, and keep it civil. No harassment, no abuse.
• No spam - one thread per idea, no self-promotion floods.
• No detection-evasion techniques. We remove anything meant to help automation avoid being caught by a platform: anti-ban tricks, fingerprint spoofing, “undetectable” methods. That is the same line the product itself draws. Reliability work such as retries, waits and spacing is fine to discuss.
• No other people’s personal information, and nothing confidential.
• Everyone here is 18 or over. It is in the terms.

Staff remove what breaks these rules and ban accounts that keep breaking them. The full version lives in the terms of service.

Every thread and reply has a Report button. Pick a reason, add a note if it helps, and send. It lands in a staff queue that shows the content in context, and one decision there settles every report on the same post. To write to us instead, or to appeal a moderation decision, email support@noisey.app and a person will read it. We remove illegal content when we find it, and may report it to the authorities where the law requires.

Yes, both, on your own threads and replies. Edits show a small “edited” marker. Deleting a reply leaves a “removed by its author” note, so conversations keep their shape. Deleting a thread takes its replies with it. Content that staff removed for breaking a rule is kept privately as the moderation record, but it is never shown again.

Announcements is where NOISEY posts release notes and news. Staff start those threads, and anyone can reply. Every other topic is open. Pick the one that fits. Staff sometimes move or pin threads so the useful ones stay easy to find.

Staff locked it, usually because it has been answered or has stopped being useful. You can still read a locked thread. It just takes no new replies. If you think a lock is wrong, email support@noisey.app.

One new thread every two minutes, and one reply every ten seconds. It stops a stuck script, or a busy afternoon of pasting, from burying everyone else’s conversations. Wait for the timer it shows you, then post again.

Not yet. Posts are text, and links work. Text keeps the forum fast, and keeps moderation manageable at this size. If you need to show something, host the image elsewhere and link to it. Your Media library is a separate thing: private storage for the videos your automations post.

The left rail has three sections: Phones, Media and Routines. Under them sit your groups, and the panels for whatever you have selected. Phones shows how many phones are ready. Routines shows how many routines you have saved.

Activity and Settings are the two icons at the far right of the top bar, not rail items. They are places you visit, not things you do. Both toggle, so pressing again takes you back to Phones. The foot of the rail always shows the adb status and the bug button. Tick some phones and a batch bar appears under the grid.

The ones you have ticked, not the one you happen to be looking at. The Phones list in the panel, the batch bar, the Toolbox and Claw all use the same selection. The Phones header shows it as N/M, with All and None underneath. Clicking a tile's header ticks it too.

Know this before you press anything: a phone is ticked automatically when it connects. Plug a phone in mid-session and it joins the target of your next batch action.

Almost always one of four things is missing:

• A phone selected - most of the app.
• Exactly one phone - recording.
• This computer connected to your account - sync, profiles, and jobs sent by the server.
• A live mirror running on that phone - needed for typing as keystrokes, and for any text outside ASCII.

Some controls also depend on what the phone can do. The Shell section and the app library disappear when an iPhone is selected, and Put selected devices onto TCP goes flat.

Press Activity, the icon at the top right. Its badge counts lines, and shows 99+ past ninety-nine. Every action reports there: the result of a bulk setting, shell output, a screen read, a rack key install, and the name of any phone that drops off. This is your Mac's log of what it did to the phones in front of it. Activity on the website is a different list: it records the jobs the server sent out.

The sun/moon icon at the far right of the top bar. The choice is remembered on this Mac.

Not to control phones. Plug a phone in, mirror it, tap on it, run a routine you wrote here. None of that asks a server for permission, and none of it stops when your connection does.

You need an account for the cloud half: routines syncing to the dashboard, profiles per phone, automations the server runs, and Google Drive. The first screen and the bell both tell you this.

Four places, all doing the same thing: the button in the top bar, shown on every screen while you are signed out; the account icon at the foot of the rail; Settings → Account → Connect this computer; and the first-run screen.

The app gets a code, shows it in large type, and opens your browser at my.noisey.app/link so you can approve it. It checks every 3 seconds, and changes to Computer connected with your address once you approve. You never type a password into the desktop app. It never sees one.

The dialog tells you what to do instead. Go to my.noisey.app/link on any device, including your phone, and type the code it shows. There is also an Open the page again button. If the code has expired, the dialog says That code expired and offers Try again, which makes a fresh one straight away. Codes last a few minutes on purpose.

The chip in the top bar. The address is blurred by default. Click the eye to show it. A shared screen or a screenshot should not give away your email. Clicking the address opens your dashboard. The dot is green when the cloud can be reached, and amber when this computer is signed in but cannot reach it right now. Settings → Account says the same thing in words.

Press the top-bar chip, or the rail's account icon, then Manage / sign out → Sign out. It asks first, and the dialog says what will happen: signing out stops sync and new jobs for this computer, and the phones in front of it keep working. Activity logs This computer is signed out, and the Connect button comes back.

Exactly what it says: Phones keep working; sync and new jobs wait until the connection is back. Nothing here is lost. Controlling a phone never waits on a server, so mirroring, tapping and routines you run by hand carry on. The app checks the connection every 60 seconds, so the banner shows what is true now, not what was true when you opened the app.

Two things must be true before a tile appears. USB debugging is on in the phone's developer options, and adb is working on this Mac. The foot of the rail shows a green dot and the adb version when it is. Then plug in, accept Allow USB debugging on the phone, and press Rescan.

It worked when a tile appears with a green dot and the top bar reads N ready. If the phone is listed but marked unauthorized, you still need to accept the trust prompt. An iPhone never appears this way. It has its own setup, below.

NOISEY reaches Android phones over adb. Until it finds adb, it cannot control any of them. The bell says exactly that and gives you an Open settings button. iPhones are not affected. Settings → Android Debug Bridge shows the status, the version and the full path the app is using. The dot at the foot of the rail shows the same thing on every screen. Point NOISEY at a working adb and the dot turns green without a restart.

A banner above the grid reads N phones need attention. It lists each serial with its state, so you can fix named phones instead of hunting through the grid.

On Android, unlock the phone and accept Allow USB debugging. Tick Always allow from this computer so it lasts after you unplug. On an iPhone, unlock it and tap Trust This Computer. Each Mac needs its own pairing, and there is no way around tapping it. It worked when the banner's count drops and the dot turns green.

Press Use rack key… on the N phones need attention banner. The rack vendor's “licence file” is an adb private key. Installing it makes that key your Mac's adb identity, so every board set up with it authorises without anyone tapping anything.

adb restarts, every board reconnects, and the app rescans right away and again 4 seconds later. Activity logs rack key installed, and names the backup it saved under ~/.android/ if you already had a key. Android only. The button is hidden when every blocked phone is an iPhone, which has nothing like this.

Press Rescan, the circular-arrow icon in the top bar on the Phones section. It is turned off while a batch operation runs. Most of the time you will not need it, because the list refreshes itself every 5 seconds. A phone that disconnects is named in Activity instead of just vanishing, so you know which one went.

Go to Toolbox → Network. Set the subnet, a from and to range, and the port, then press Scan. You don't need to select anything first, because this is a search. Settings → This machine shows your own address and subnet, and the subnet box is filled in from it. The port box takes one port, 5555; a list, 5555,5037; or a range, 5555-5560. A range can cover at most 64 ports, so a typo cannot turn into millions of probes.

It worked when Activity reads found N adb endpoint(s), connected N. If devices answer but none of them speak adb, the app tells you which of two things is happening. Either the boards are on but not listening, so USB authorisation is still outstanding. Or the rack is switched off, or on another subnet.

Select the phones, then press Toolbox → Network → Put selected devices onto TCP. It uses the first port in the port box. Android only. The button is turned off if any selected phone is an iPhone, and the hint explains why. adb over TCP is an Android feature. An iPhone stays on USB for pairing and launching, and its control channel already runs over your network.

Press the ●/○ eye button on the phone's row in the panel's Phones list. Only the tile goes away. A hidden phone can still be selected and used, which is what you want for a board you are not watching but still want in a batch. The app remembers your choice between launches.

Click the phone's name on its tile, type a new one, and press Rename. When you are signed in, the rename goes through the server, so the grid and the dashboard agree. When you are offline, it is kept as a nickname on this Mac.

The tile shows the first name it can find, in this order: a profile from the website, then the server's name for that phone, then your local nickname, then the model, then the serial. If the phone has a profile, the name is not clickable. That name is set on the website, and the tooltip says so.

The quick way is the density icon in the top bar, which cycles through 110, 150 and 220 px. For exact sizes, use Settings → Device display → Small screen size (90-320 px), and Main screen size (50-100%) for the focused view.

This changes the display size only. It changes nothing on the phone, and it does not reduce load. For that, use the encoder settings two questions down.

• Green or amber dot - ready, or not ready.
• A small avatar - a profile is assigned to this handset.
• NNfps - the live mirror's frame rate.
• NN% - battery.
• ⤢ - open this phone on its own.
• The foot shows the serial. Hover over it for the serial, resolution and OS version.

Two ways, and the difference matters. Double-clicking a tile selects only that phone and opens it. The ⤢ button opens it and leaves your selection exactly as it was.

In the focused view, the phone on screen always gets what you send. A key press reaches it whether or not other phones are ticked. Sync keyboard decides whether the rest of your selection gets the same press, and the hint under the key row tells you how it is set.

Two ways, and the second is usually the one you want. The + on the rail's Groups header makes an empty group that you then fill. Or tick some phones and press New group from N selected in the panel, which names it and fills it in one step. The group appears in the rail with a count.

Select them, then press Move to group… on the batch bar. The dropdown includes No group, to take them out again. It only appears once you have at least one group.

Click the group's name in the rail. All phones clears the filter. An empty group says so, and tells you what to do: Select phones in All phones, then use ‘Move to group’ on the batch bar.

The panel's Groups section has a tick for each group. It only shows when you have groups. Unticking one takes its tiles off the grid. They are hidden, not disconnected: those phones are still plugged in and still working.

No. Groups live on this computer only and are not synced. Besides tidying the grid, they set scope: a routine's Where and when → Run on can target A group. So “the morning shelf” becomes something you can point a schedule at.

Press the mirror icon in the top bar. It lights up when it is on. Settings → Behaviour → Mirror screens live instead of polling screenshots is the same switch. When it is on, each phone gets a real video stream. When it is off, the app takes a screenshot every 3 seconds. It worked when tiles show an fps badge and the top bar reads N live.

The placeholder tells you which of three things it is:

• Waiting for a frame… - normal, and not a fault. A phone only sends video when its screen changes, so a phone sitting still sends nothing. Opening it in the focused view, or making the tile bigger, forces a fresh frame.
• Mirror failed or Mirror ended, with a Retry button.
• The state word, such as unauthorized or offline, when the phone cannot be controlled at all.

An iPhone's video comes a completely different way, through WebDriverAgent instead of the Android video stream. The tile behaves the same.

Use Settings → Device display → Small screen resolution (240-1080 px) and Frame rate of small screen (1-60 fps). These control the phone's own encoder, not the size of the picture on your computer. That is why the tile-size slider does not help and lowering these does. Changing either one restarts every stream, and Activity says so, so you are not left wondering why the grid blinked.

Click or drag straight on a tile, or on the big screen in the focused view. The phone you touched always gets the gesture. Sync decides whether the rest of your selection gets it too.

On a phone with no live mirror, the app collects the press and the release, then replays them as a tap or a swipe when you let go. Nothing moves on that phone until you release the button.

Sync in the top bar, also called Sync mouse in the panel, makes one gesture control every selected phone, not just the one under the pointer. It handles everything adb cannot do in bulk. Put twenty phones on the same settings screen, then tap through once.

Check they all landed on the same screen before you start. Different firmware can land one level up, and a synced tap would then press something else on that phone.

Sync keyboard sits in the panel next to Sync mouse. It controls how far keys and typed text go while you have one phone open in the focused view. When it is on, they also reach the other selected phones. When it is off, only the phone on screen gets them. Off is what you want when you are typing into one phone out of twenty.

The phone on screen gets the press either way. In the grid, where no single phone is open, keys and text go to every selected phone. The hint under the key row tells you how the switch is set, and how many other phones are listening.

They sit on the batch bar in the grid, and in a row under the phone in the focused view. In the grid, they go to every selected phone. In the focused view, the phone on screen always gets them, plus the rest of the selection when Sync keyboard is on.

On Android, Back also wakes the screen while a live mirror is running, which is what you want for a shelf of sleeping boards. On an iPhone, only keys that have a real equivalent work. iOS has no Back, so pressing it returns an error instead of quietly doing nothing.

Open one phone, then press Read screen in the bar under it. The result goes to Activity, up to 40 lines. After that it says screen read - see Activity.

On Android, that is uiautomator's accessibility tree. If the screen is still animating, it tries up to 4 times, then says it could not read the screen instead of handing you half of one. On an iPhone, the tree comes from WebDriverAgent through the same reader Claw uses, so you read exactly what the model reads.

Use Type on devices in the panel. Type your text, then press Type or Enter. It goes into whatever field is already selected on the phone. NOISEY does not pick the field, so tap into the box first, or let a routine do it. You need phones selected, and it follows Sync keyboard when you have a phone open in the focused view.

Tick Send as keystrokes under Type on devices. Left off, the text is pasted through the clipboard. Ticked, it is typed key by key. Password fields need that, because most of them refuse a paste.

Two rules the box does not mention. Keystroke mode handles ASCII only, and the app refuses anything else rather than sending broken text. It also needs a live mirror running on that phone. Without one it answers needs a live mirror session.

The clipboard route on its own. Emoji and any alphabet work, and it goes to every selected phone. The phone's clipboard sets the size limit. Go over it and the app refuses and prints both byte counts, rather than quietly cutting your text short. With no live mirror it falls back to adb's own text input, which on Android handles ASCII only. An iPhone has no such limit, because WebDriverAgent takes real UTF-8.

The phone had no live mirror, so the text went through adb's text input. On Android that cannot carry anything outside ASCII. The app says so instead of sending something close enough: needs a live mirror session - adb cannot type non-ASCII. Turn live mirroring on for that phone and send it again. iPhones are not affected.

In the panel, press Bulk settings → Wi-Fi or Data → On or Off. It applies to the phones you have selected, and the section states the count: Applies to the N selected phone(s). Activity reports the result: wifi turned on on N phone(s). iPhones in the selection are skipped, and the panel says so. These controls write to Android's settings over adb.

Press Bulk settings → Lock all to portrait. Unlike the other controls in that section, this one falls back to every ready phone when nothing is selected.

The app checks the change instead of assuming it worked. settings put reports success whether or not the value stuck, so the app reads the value back. Once, seven of nineteen boards were still unlocked after a console reported success on all of them. Activity reads N phone(s) locked to portrait, and the grid rescans a second and a half later so the tile shapes update.

Not by command. You do it by hand. Joining a network needs root, and these boards cannot be rooted, so the app does not offer a button that would quietly fail.

Here is what works. Bulk settings → Open on every selected device → Wi-Fi puts every selected phone on the Wi-Fi settings screen. Turn Sync on and tap through once, and they all follow. Three screens have buttons: Language, Wi-Fi and Settings. Check they all landed on the same screen before you start tapping. Different firmware can land one level up, and a synced tap would then hit the wrong thing.

Press Toolbox → Prepare for automation → Prepare selected. Do this once per phone, before your first routine. It turns off Android's full-screen confirmation overlay, which has no button to tap and otherwise blocks automation on every full-screen app. It also locks the phone to portrait. It worked when Activity reads N phone(s) ready for automation. Android only.

Select the phones, press Install app… on the batch bar, and pick a .apk. To install more than one file at a time, use Toolbox → Apps → App library.

Installs run in groups of 4, not all at once. Twenty at a time down one shared USB hub is slower, not faster. It worked when Activity reads install → N phone(s) OK. Failures come as a count, with the first message attached. Android only. An iPhone needs a signed IPA and a provisioning profile for each app, so the whole library section is hidden when one is selected.

Press Toolbox → Apps → App library, then type a path or press Browse. Every .apk, .apkm and .xapk in the folder is listed with its package id, version and size, or marked unreadable. Split bundles are marked split. The app remembers the folder for next time.

Press Toolbox → Apps → Quick install → Check N devices, then Install N missing apps. Each app goes only to the phones that do not have it, and progress is reported for each phone. You need phones selected and a library with something in it. It worked when you get either All N app(s) are on every selected device, or a list reading missing on X of Y followed by N installs succeeded.

Because two files in the folder declare the same package. The library reads the package instead of trusting the file name: N package(s) appear more than once. The file name is not the app.

It is worth acting on. Three downloads named for Reddit, X and Bluesky once turned out to be identical copies of the same app store. Installing them would have put that one store on twenty phones three times, and none of the three apps anywhere.

Press Toolbox → Apps, type the app's id, then press Launch. On Android that is a package name like com.example.app. On an iPhone it is a bundle id like com.apple.Preferences. The placeholder changes to match your selection, and a hint says which one it wants.

No. The desktop app has no uninstall control, on either platform. Remove the app on the phone, or with adb yourself.

Press Toolbox → Shell, type the command, then press Run. It runs on every selected phone, and the output goes to Activity: the first 3 phones, cut to 120 characters each. Twenty full outputs is not a log anyone reads. The whole section disappears when an iPhone is selected. iOS has no shell, and the app does not pretend otherwise.

Select them, then press Restart on the batch bar. The button is styled as destructive, and it names the count before it goes. Mirrors are shut down first, so the streams come back cleanly instead of reconnecting to a phone that is halfway through booting. Android only. You cannot restart an iPhone from NOISEY.

There is no screenshot button. The app only uses screenshots for the still preview it takes when live mirroring is off, and it never saves any of them to your computer. If you want something you can act on, use Read screen in the focused view. It gives you the elements on screen instead of a picture, which is usually what you wanted.

Press Media → Choose folder… and pick a folder on this Mac. The videos in it become your library. Drop a clip in and it is there, ready to send to any phone. The app remembers the folder.

This is not the same Media as the website's. The website's is a cloud library that automations watch. This one is a folder on your own disk that you send to phones by hand, or with auto-send.

Press Refresh in the bar at the top of Media. Nothing watches the folder for you, so a file that lands while you are looking does not appear on its own. The same bar shows the folder path and the video count. The empty state names the folder, which is the fastest way to spot that your clip went somewhere else.

Select the phones under Phones first, then press Send to N devices on the file's row in Media. Media has no phone picker of its own. With nothing selected, the buttons are flat and the hint tells you to go and select some.

For a one-off file outside the library, use Toolbox → Send a video → Choose a video…. It takes mp4, mov, m4v and webm. It worked when Activity says the file is now visible to pickers on N phones. Android only.

That is the app refusing to call a half-finished job a success. Sending a video does two things: it copies the file to the phone, then makes Android index it. A file that is on the phone but not indexed is invisible to every upload picker, which is the only reason you sent it.

Send it again. If it keeps failing on the same phone, look at that phone's media scanner, not at NOISEY.

Press Media → Auto-send → New. You need a media folder chosen and phones selected, or the button stays flat. The rule then reads as a sentence: when a video is added to that folder, send it to N devices, so many seconds apart.

It starts off. Press Turn on when you are ready, and it checks every 15 seconds. Send now runs it immediately. The trash icon deletes it. Each rule remembers what it has already sent, so a file is never sent twice. The seconds apart protect your own computer and bandwidth. One computer pushing to twenty phones at once causes a traffic jam. That is all the gap is for.

Auto-send moves files onto phones. It does not post anything. The words are worth keeping straight, because this feature used to be called Automations, and that name covered two unrelated things:

• Routines - what a phone does.
• Automations - when a routine runs. Written on the website.
• Auto-send - how a video gets onto a phone. It lives under Media because that is what it is about.

Press Routines → New routine, either from the empty state or from the button at the foot of the list. Name it and press Create. There is only one kind of routine. Whether it has been learned, and whether it is scheduled, are states it is in, not separate things to make. It syncs to your account, so a moment later the website offers it when you build an automation.

Open Routines → the routine → Instructions. Put one thing per numbered row, then press Add an instruction for the next. They run in order, each on whatever screen the last one left. So write “open TikTok”, then “press the plus”, not one row describing the whole post.

Reorder rows with the up and down chevrons, and remove one with the × on the row. Each row has Carry on even if this one fails, for steps that are allowed to miss, like a cookie banner that is not always there.

Write {{something}} inside an instruction. A Needs entry appears the moment you type it, waiting for you to say where its value comes from. That is the whole step. There is no form to fill in first and no library to set up, and a placeholder and its definition cannot drift apart, because one creates the other.

Each need is either Some text or A video, and takes one of three sources. Ask me each run asks you when it runs, so the routine cannot run on its own. Always the same is a fixed value you type. Newest unused file in a folder takes the next file. It worked when the warning about an undefined placeholder disappears and Run comes back to life.

Set the need's kind to A video and its source to Newest unused file in a folder, then pick the folder. The file is sent to the phones before any instruction runs. What gets put into your wording is the filename. That is deliberate. The instruction ends up saying “pick clip-08-13.mp4 from the gallery” instead of “pick the newest video”, which breaks as soon as something else on the phone is newer.

Use Routines → Guidance for Claw. Write what you want in plain text, such as “captions are lowercase, one line, no hashtags”. It shapes what Claw decides, not what is sent to the phone.

A learned routine does not read it. Once the steps are compiled, those steps are what runs. Guidance only matters while a routine is still thinking, so changing it changes the next time it learns, not the next replay.

Open Routines → Where and when → Run on and pick All phones, A group, or Chosen phones. For chosen phones, press Use the N selected.

The list is worked out when the routine runs, not when you save it. Chosen phones are matched by hardware id, not by serial. Matching on serials would quietly drop a board whose address changed after a router reboot. It would post to eighteen of nineteen phones, with nothing in any log to say so.

Your choice, under Routines → Where and when → Afterwards. Pick Go to the home screen, which is the default, Leave it where it ends, Close the app, or Lock the screen. This matters most when the next routine expects to start from home. A phone left inside an app is the usual reason a second run behaves differently from the first.

Press Run on N devices at the top of the routine. It saves your edits first, so you run what you can see on screen.

The button is flat when the routine has no instructions, no phones in scope, or a need with no source. The tooltip says which. A learned routine reports N ok, or X/N failed with the first failure quoted in full. The count tells you something broke, and the message tells you what. A routine that has not been learned opens the Claw drawer instead, with one run per phone.

Press Save in the routine's bar. Beside it, a live indicator reads Saved or Unsaved changes. Switching to another routine saves the one you are leaving, so clicking away does not lose work. The button is there to tell you where you stand, not to be the only thing between you and losing an afternoon.

Press Routines → Where and when → Run it on a schedule, then pick Daily at a time, or Every N min. It worked when the row in the list shows a green dot and next HH:MM, and the line under it reads daily at 09:00 instead of not scheduled.

Two limits to know before you rely on it. The scheduler runs inside the app, so routines only fire while NOISEY is open. And scheduled routines run one at a time. Two firing at 09:00 across eighteen boards would fight over the same USB hub.

Run it once with Claw, then press Learn. That turns what Claw worked out into steps that replay with no model behind them. The chip changes from Not learned to Learned, and the cost chip from thinks every run to free.

It learns from one phone's run, not from all of them. The compiled steps find things by name rather than by position, so they work on other phones. Paying a model nineteen times for nineteen copies of the same route would defeat the point. The first phone to finish successfully is the one that teaches.

Read the note under the routine's bar. It says one of two things: Learned N step(s). Ends with a check, so a run that did not work says so. Or, in bold, It has no way to prove it worked.

The second one means the screen it finished on had nothing specific enough to check for. A success there only means that no step failed. That is not the same as “it posted”, and the app will not blur the two for you.

Press Forget, next to Learn. It is flat unless the routine is learned. It drops the compiled steps, so the next run works the task out again.

This also happens on its own when you edit the instructions, and Activity says so: the routine changed, so its learned steps were dropped, and it needs one run to learn again. Otherwise a replay would keep following the old route while the screen showed your new words. Only instruction edits do this. Renaming or rescheduling does not.

Look at Recent runs under the routine. It shows the last 8. Each row gives the time, N-failed/total ok, whether it replayed or ran through Claw, and scheduled if a timer started it. Failed runs are highlighted and show the first error, so the list answers “has this been working?” without a trip to Activity.

Press the trash icon in the routine's bar. It goes straight away. There is no confirmation and no undo, so check you are looking at the right routine before you press it.

The part of the desktop app that controls phones from plain English. Press Claw in the top bar, say what the phones should do, and press Run on N devices. You need phones selected and an engine set up in Settings → Claw.

Every selected phone gets its own run. If one goes off script, it is worked out on its own instead of blindly copying another. Each run shows as a card with a status chip and its steps.

Because no engine is set up yet. Claw controls phones with a language model, and NOISEY does not resell one. Bring either an OpenRouter key or your own Claude subscription, and add it in Settings → Claw. Keys and tokens are kept in the system keychain, either the macOS Keychain or Windows Credential Manager, never in a file. Nothing else in the app needs this. Mirroring, typing and learned routines all run without a model.

Press Settings → Claw → Engine → OpenRouter - pay per token. Paste your sk-or-… key and press Save key. It worked when the key row's dot turns green and reads In the keychain.

For the model, take the tested default, the stronger fallback model (most reliable in the benchmark, ~6× cost), or type any OpenRouter model id. If you choose a model yourself, that is the one that runs. A failed run only retries on the stronger model when you left the tested default selected.

Press Settings → Claw → Engine → Your Claude subscription. It opens a three-step checklist: Claude on this Mac, with an Install Claude button if it is missing; Your Claude subscription, which is the next question; and Model - Sonnet (fast, on every plan), Opus (most capable, needs Max) or Haiku (cheapest, simple tasks). Then press Test it. If you are already signed in to Claude Code on this Mac, the card says so and gives you nothing to press.

Two limits. Runs use your own Claude plan and its rate limits. And Approve every step is not available on this engine, because a run with no window cannot stop to ask. If ANTHROPIC_API_KEY is set on this Mac, the card warns you. Runs then bill your API account by usage instead of your subscription.

Because claude setup-token opens a browser, then waits for you to paste the authorisation code back into standard input. An app window has no standard input to paste into. So Get a Claude CLI token opens a terminal instead. Follow it through, copy the sk-ant-… token it prints, paste it into the field, and press Save.

Improve, under the Claw box, rewrites what you typed as numbered steps a phone can follow. “Open X, go to a profile, follow ten of their followers” is one sentence to a person and about fourteen actions to a phone.

It replaces the text in place, so you can read and edit it before anything happens. The hint says the same: nothing runs until you press Claw. It may not add steps you did not ask for. You can edit the prompt behind it in Settings → Claw → Improving a request, with Save prompt and Reset to default. Leave it empty and you get the built-in one, shown greyed out in the box as a placeholder.

Tick Approve every step before it happens in the Claw drawer before you run. Each step then waits for you. OpenRouter engine only. The Claude subscription engine runs with no window and cannot pause to ask, and the checkbox says so instead of quietly ignoring you.

Press Settings → Claw → Devices at once. Pick All of them, or a number from 2 to 20. Anything over the limit waits its turn and shows as queued. Every running routine shares the limit, so two at once still add up to this number.

The hint states the real cost, and it changes with the engine. On OpenRouter, nothing runs on your computer, so this is only about how hard you push the phones. On the Claude engine it is about memory: that engine runs one process per phone, at roughly 500 MB each.

The run card shows the question with a box under it. Type the answer and press Send. The chip on that card reads asking while it waits, so a stopped run is never a mystery. Only that phone's run is waiting. The others carry on.

Stop on the run card, shown while a run is running, asking or queued.

The run card lists the last 6 steps: the tool it used, and why it chose it. Failures are highlighted. Above them, a strip shows one dot per instruction - done, now, or failed - and a count like 3/5. It is a strip rather than a sentence because nineteen of these have to fit on one screen. Clear finished on the Runs header tidies the drawer when you are done.

The Claw drawer's header shows a running dollar total to four decimal places, priced against the default model's list price. It shows money, not tokens. It is your key and your account either way. On OpenRouter, they bill you per token. On the Claude engine, it comes out of your own subscription's limits.

On a finished run, press Save as routine and name it. It arrives already learned, because it just succeeded on that phone. It is also scoped to that phone rather than all of them. Widen it under Routines → Where and when once you have watched it do the right thing on another phone.

That is on purpose. Both are right-hand panels about 340 px wide. Having both open on a laptop leaves too little grid to be worth looking at, so opening one closes the other.

Two ways in, and the first is easier:

1. Open one phone in the focused view and press Record in the top bar. Do the task on the screen in front of you, then press it again. It now reads Stop · N, with the step count.
2. Or Recordings → Record on the selected device in the panel.

You need exactly one phone selected. Otherwise the rail says Select exactly one device to record. Name it, press Save routine, and it appears under Routines like any other. Cancelling throws it away.

For the things adb cannot do in bulk: changing the system language, joining a Wi-Fi network, or anything that only exists as a screen in Settings. Record it once on one phone and replay it on twenty. When the screen has names to aim at, written instructions move between phones better. The reason is two questions down.

No. Anything you type during a recording becomes an Ask me each run need with no stored value. The characters are dropped when the recording is compiled, and never written into the routine. So a recording can include a password without storing one.

That matters more than it looks, because a routine's contents sync to your account unencrypted. Recorded counts, meaning how many times you repeated something, are kept as Always the same and stay editable under Routines. That way a scheduled run does not stop to ask a question nobody is there to answer.

Because it replays exact positions, and a position has no name and no check. A tap at a position cannot fail. Every phone would report success while pressing whatever happened to sit at those pixels.

So it is scoped to the phone you recorded on, and the app says so when you save it. Widen it under Routines → Where and when once you have watched it do the right thing on a second phone. You have to make that change yourself. It is never the default.

Yes. Timings are recorded and replayed exactly as you did them, because screens need time to settle and replaying instantly races the UI. Nothing is randomised. The delays are the ones you actually took, and they are kept for reliability, not to make anything look like anything.

Failures only. It holds recordings that could not be turned into routines. They are kept as local replays instead, they stay on this Mac, and they never sync. Each has Play, which sends it to every selected phone, and × to delete it.

A recording that compiled cleanly is a routine and appears under Routines, not here. An empty Recordings list is the good outcome.

You need Xcode and an Apple Developer account, and there is no way around them. Apple only lets one app control another during UI testing. So NOISEY builds and signs WebDriverAgent on your Mac, for your specific phone. There is no download-and-run version, and there cannot be.

That makes iPhone support a tool for the person who owns the phone, not something you can hand to a customer. Android needs none of it: one switch in developer options, and a cable.

Press Settings → the iPhone card → Set up iPhone. One button does every step a computer is allowed to do, then tells you what is left for you. It is safe to press again and again, and that is how it is meant to be used. Progress lines appear in the card while it runs, because the downloads inside take minutes. It worked when the card reads Ready - an iPhone tile can be opened and a tile appears.

Seven checks, each with its own fix. Three run themselves. Four are physical, or need your Apple ID. That is why it is a checklist and not a progress bar.

• Xcode installed - yours to do. Install it, then open it once to accept the licence.
• iOS platform support - automatic. Xcode 26 comes without it and can download it on its own.
• Device tools - automatic.
• Apple Developer team - yours to do. Open Xcode → Settings → Accounts and sign in. NOISEY finds the team afterwards.
• WebDriverAgent source - automatic, downloaded when needed. It is never bundled, because it has to be signed with your own team.
• iPhone connected - yours to do. Plug in, unlock, and tap Trust. A charge-only cable looks exactly like a data cable, and your Mac simply never sees the phone.
• Developer Mode - yours to do. Open Settings → Privacy & Security → Developer Mode, then restart the phone and unlock it.

Check again on the same card runs all seven again without redoing work.

The bell gives the reason: No iPhone on this Mac, or iPhone not ready yet. The body is the fix text for whichever check is blocking, and there is a Check again button. It is almost always one of the three physical steps: the cable, Trust, or Developer Mode. The notice only appears for people who have been through the setup, so nobody running Android boards is nagged about a phone they do not own.

Quite a lot. Where it can, the app hides the control instead of failing on the phone:

• No shell. The Toolbox's Shell section disappears when an iPhone is selected.
• No installing from a file. That needs a signed IPA and a provisioning profile for each app, so the app library is hidden.
• No Android settings. Bulk settings skip iPhones. Change those in the Settings app, or point Claw at them.
• No adb over TCP, and no restarting, and no sending a video to the gallery.
• No Back key. iOS has none, so pressing Back returns an error instead of quietly doing nothing.
• Launch takes a bundle id, not a package name.

Most of what matters does work: mirroring, tapping, swiping, typing, reading the screen, routines and Claw. Typing takes real UTF-8, with none of Android's ASCII limit.

Four tiles across the top, each one a link: Phones connected, showing live out of registered; Automations on, showing enabled out of total; Media waiting, the files you are holding right now; and Posted this week.

Under them, in order: Needs a look, when something is wrong; the three-step setup checklist, until your account is set up; a seven-day chart of posts; and Get the desktop app. Nothing here controls a phone. The dashboard tells you what is happening and sends you to the page that can change it.

The first number counts the phones a connected computer can see right now. The second counts what is registered to the account. Zero means none of your computers is connected. Or one is, but it cannot see the phones, usually because of the cable or an unaccepted trust prompt. The second number never drops on its own. A phone stays registered while your computer is off, which is what lets an automation you build today reach it next week.

Three things, gathered in one place. The card only appears when there is something in it: a source whose last Google check failed, the number of failed jobs, and the number of scheduled runs marked missed, worded “was the Mac open?”. Each line links to the page that explains it. There is no green all-clear version. No card is the good outcome.

1. Connect your computer - the pairing code from the desktop app.
2. Plug in a phone - this opens the help centre in the same tab, so use your browser's back button to return to a half-finished setup.
3. Create an automation.

The checklist disappears once all three are done: a computer connected, at least one phone registered, and at least one automation. It does not come back.

The last seven days in your own timezone, oldest first. Green is posted, sitting on the baseline. Red is failed, stacked on top. The total sits above each bar, and hovering shows the exact split. It counts finished jobs only. Queued, running and missed jobs do not appear, so a day when your computer was shut shows an empty bar instead of a row of failures.

Everything except touching a phone.

• Works with the Mac shut: the whole dashboard, writing and editing routines, automations and flows, the webhook receiving files, Drive sources, flows running on our servers, the calendar, the store and the community.
• Waits for the Mac: every job dispatch, Try it on a phone, Pick it off the phone, and a flow’s Run a routine step.

Jobs queued in the meantime are not lost. They are sent the moment the desktop app connects, unless their start window has already closed.

The dashboard reads from api.noisey.app on a different domain, sending your session cookie with each request. A browser that blocks third-party cookies makes every one of those reads look signed out. Pages come up empty, sign-in loops back to sign-in, and my.noisey.app can sit on “Loading…” forever. The give-away is that nothing shows an error. It is all just blank. Allow cookies for noisey.app, or try another browser to check.

Yes. On narrow screens the left rail hides behind a menu button at the top left, and closes as soon as you go somewhere. The routine and flow editors drop from three columns to one below 1100px. The calendar steps from 7 columns to 4, then 2, then 1. Lists, activity and the community read well on a phone. Drawing a graph on one is possible, but not pleasant.

Press the book icon at the foot of the left rail, between Settings and Sign out. It opens this help centre in a new tab, so you keep whatever you were half-way through.

Three kinds share one list.

• Drawn - a graph you made on the canvas. Open takes you back to it.
• Written in sentences - plain instructions a Mac teaches itself the first time it runs them.
• On a Mac - taught in the desktop app and mirrored here so an automation can choose it. It says learned or not learned yet.

The kinds overlap. A routine taught on a Mac now brings its graph with it, so it can be both taught on a Mac and drawn.

Yes. Press Open on its row. A routine taught in the desktop app arrives with its graph, so it opens on the canvas like any other, with the same boxes, wires and inspector. Edit it, save, and the new version goes back to every Mac on your account.

A row with no Open is written as sentences rather than drawn. You still change those in the desktop app, on the computer that owns them.

Whichever side saved last wins, and the page says so above the canvas. Your edit goes out as a newer version, and each Mac replaces its own copy instead of merging the two. A half-merged routine that runs differently on one phone is worse than a lost edit. Edit in one place at a time.

Five groups in the Add a step palette:

• Navigation - press back; press home; open an app by package name; open a settings page.
• Interaction - tap; tap if it is there, for a dialog that only sometimes appears; type text; scroll until found, with a give-up count; and tap a position, marked last resort because positions differ between phones.
• Wait & check - wait for something, with a timeout; wait a fixed number of milliseconds; check it worked; stop if this appears; and check the phone is healthy, with a minimum battery and a maximum temperature.
• Decisions - if it is on screen; if a value is; if a count is under; stop, with a message.
• Values - remember a value, readable later as {{name}}; count up.

Check it worked is the proof. A routine without one cannot say it worked, only that it finished.

Stopping loudly instead of tapping through something the routine does not understand: a rate-limit notice, a consent gate, an account warning. You name what to look for and the message to record. The job then stops with that message on its row in Activity. It is a stop sign, not a way around anything. When a platform tells you to slow down, the right answer is to stop.

The box with the Start pill. The first step you drop on an empty canvas becomes the start. To move it, select another box and press Start here in the inspector. Saving is refused if the start is not one of the routine's own boxes.

A decision has two dots under it instead of one: green yes on the left, red no on the right. Drag from the one you mean to the top of the next box. Ordinary steps have one outgoing dot. Stop has none. To remove a wire, click it and press delete. Both boxes are updated, so nothing is left pointing at a connection that no longer exists.

Click the box. The inspector on the right shows exactly the fields that step has, and nothing else. Delete in the inspector removes the box and clears any wires pointing at it, so saving does not then fail on wiring you never drew.

Press Save routine, top right. The name comes from the field above the canvas. Leave it blank and it saves as “Untitled routine”. Success reads “Saved as version N. Your Mac picks it up the next time it connects.” The first save also puts an id in the address bar, so refreshing keeps the routine instead of giving you a blank canvas.

A box nothing connects to is nearly always a forgotten wire. Draw it, or delete the box. Saving checks the whole graph and names one problem at a time:

• at least one box, and not more than the cap;
• every box with an id, and no two the same;
• every step one this build can actually run;
• every wire pointing at a box that exists;
• the start being one of the boxes, and every box reachable from it.

It is all refused here, rather than at three in the morning on a phone.

Write {{name}} inside a type-text step. The value comes from a Remember a value step earlier in the graph, or from an input the automation supplies. Do not leave a routine asking a person for the value. A routine that does is hidden from the automation and schedule pickers, because nobody is there at 9am to answer.

Two reasons the button refuses:

• No phone is connected. A trial is a real run on a real phone, so the desktop app has to be open with the phone plugged in.
• You already have three trials going: “you already have trial runs going - wait for one to finish”.

A trial saves the routine first and runs the saved version, so what you watch is what is on your screen. The panel keeps the last 12 steps, and a trial expires after 10 minutes.

Press Delete on its row and confirm by name: “Delete <name>? This cannot be undone.” It only appears on routines you made here. Delete a routine that lives on a Mac in the desktop app instead. That computer holds the copy, so deleting it from the website would just wait for the next sync to put it back.

The canvas draws graphs, and that routine is a list of plain instructions. There is nothing to lay out. Change it in the desktop app, on the Mac that owns it. There is no editor for a written routine on the website yet. If you would rather have it drawn, draw it. Routines → Draw a routine gives you the same task as boxes, and that is the only shape the store will publish.

It names the missing answer instead of greying the button out. The messages, in the order you meet them:

• “Pick the folder that triggers it.” - or, on a schedule, “Pick the folder it posts from.”
• “Pick the routine to run, or write one.”
• “Pick the phone or phones it runs on.”
• “Pick at least two phones. To use just one, choose it in the ‘On’ dropdown instead.”

A schedule also needs at least one time. Nothing is created until every answer is in. An automation with no target would quietly never run.

Press Edit on the automation’s card. The same form you built it with opens with every field filled in, and Save changes writes them back.

You can change its name, the folder it watches, the routine it runs, the phone or group it runs on, and the stagger. On a scheduled automation you can also change the times, days, timezone and window.

Pausing is deliberately not part of the form. Pause and Resume have their own button on the card, so saving an edit never quietly restarts something you had stopped. Jobs that already ran are not affected either way.

For a schedule: the start window (“starts by +30m or it’s marked missed”), then either the next run as a date and time, or “no upcoming runs”, or “paused”. The server works out that next run, not your browser, so it is the same moment the job will use. For a group that fires on arrival: the seconds between phones, and the date it was made. via <machine> appears when the routine belongs to one Mac.

Because 09:00 has to keep meaning 09:00 after a clock change. An offset like UTC+1 is only right for about half the year. A named zone carries its own rules and moves with them. Your own zone is at the top of the list, marked yours. A schedule set in someone else's zone appears on the calendar at the hour it will actually run for you.

You have routines, but every one of them asks a question while it runs, and a scheduled run has nobody to answer. In the desktop app, give that input a fixed value, or take the value from the incoming file. The routine then becomes selectable. With no routines at all, the page says Start with a routine instead, and offers New automation and Connect a computer.

Three places, all producing the same kind of run:

• Automations → New automation → On a schedule - the general one. Pick a routine, a folder, times, days, a zone, and any phone or group.
• Phones → Schedule on a phone's row - this phone, at this time, whoever it posts as.
• Profiles → a profile → Posting schedule - follows the person, and runs on whichever phone carries them.

Choose by what you want to keep working after a change. A profile's schedule survives the client moving to a different phone. A phone's schedule survives the client changing.

1. Phones → Schedule on that handset's row, then Schedule something.
2. Run - the routine. Posting from - the folder it takes files from.
3. Times, day toggles, the timezone, and how late it may start.
4. Schedule it.

It appears in that phone's list with its next run, and on the calendar, named “<Routine> on <Phone>”. You need at least one routine and one folder to already exist. The form says which is missing.

A schedule needs somewhere to run, so put the profile on a phone first. Go to Phones → the Posts as dropdown on the row → pick the client. The section on the profile then works straight away. Until then it says “<name> isn’t on a phone yet, so a schedule would have nothing to run on.”

Every row in a posting schedule has Pause, Resume and Remove, and shows its next run, “no upcoming runs”, or “paused”. Pausing one time leaves the others running. You cannot edit a time once it is added. Remove it and add the one you meant.

Go to Flows → New flow. Name it, pick what starts it from the chips under Starts when, add steps from the palette, and drag from the dot under one box to the top of the next. Then press Save flow. The wiring rules match the routine canvas: every step reachable from the start, and every wire pointing at a step that exists.

Five, and exactly one per flow:

• By hand - only when you press Run now. Good while you are building it.
• On a schedule - times, days, a named timezone.
• When a file arrives - anything landing through your webhook, optionally narrowed to one folder.
• When something happens to a phone - a job finished, failed or was missed; a routine ran; a phone connected or dropped off; a computer connected or disconnected.
• When a phone gets a notification - and you can narrow it to one platform: LinkedIn, TikTok, Instagram, X, Facebook or YouTube. The chip warns you that a flow saved against this trigger does not fire yet.

The trigger is deliberately not a box on the canvas. Every flow has exactly one. A box that people could delete or leave unwired would be a flow that can never fire.

Two groups.

• Thinking - Ask a model, which takes a prompt, an optional system prompt, a name to store the answer under, and a token limit; If a value is; and Remember a value.
• Doing - Run a routine, which takes a routine and one phone, and is where a flow reaches the real world; Call a URL; Write to the timeline; and Stop, with a message and a tick for whether it counts as a failure.

Use {{name}}, where the name is whatever you typed in the output field on Ask a model or Call a URL, or the name on Remember a value. Write it into any later step's text: a prompt, a URL, a body, or a timeline note. Routines use the same {{name}}, so there is nothing new to learn when moving between the two canvases.

Press Run now in the header. It saves first, runs the saved version, then lists every step with its detail and how long it took. Boxes turn green where they worked and red where they stopped. Grey boxes were never reached, which is usually the answer to “why did it not do that bit?”. The flow took the other branch.

On NOISEY's servers. That is the point of them: the trigger fires, the model thinks, and the URL is called with your computer shut. Anything a flow does to a phone still waits for your computer. A Run a routine step queues a job like any other, and it is sent when the desktop app connects, because only your computer can reach your phones.

A Run a routine step only offers routines you made on the website. A routine taught on a Mac is missing, and nothing on the page says why. Draw the routine at Routines and it appears in the list. Automations do not have this limit. They can point at a routine on a Mac.

Three refusals, all decided on the server: “that is not a URL a flow can call”; “a flow can only call http or https”; and “a flow cannot call an address inside the network”. Loopback, private ranges and cloud metadata addresses are blocked outright. The request comes from our servers, not your browser, so allowing those would let any account reach things that are not theirs.

Four causes, each with its own message:

• No key is stored. Flows run on our servers and cannot reach the keychain on your computer. Add a key under Settings → AI for flows.
• The month's allowance is gone: “that is this month’s N AI actions used up”. The free plan’s allowance is 0, so an Ask a model step fails the first time it runs on a free account.
• “the AI provider rejected the key - check it under Settings”.
• “the stored AI key can no longer be read - save it again under Settings”.

Press Pause or Turn on on the row. Delete asks you to confirm. A paused flow keeps its trigger and its graph, and simply does not fire. The row says “· paused” after the name. The list also shows the trigger, the number of steps, how many times it has run, and when it was last changed.

Free: 3. Paid: unlimited. Past the limit, creating one is refused with “that is your 3 flows used - the paid plan lifts the limit”. The limit counts flows you have, not flows running, so a paused flow you have finished with is worth deleting rather than leaving switched off.

The filename, the size, when it landed, when it expires (“expires in 19h”, “expires in 3d”), and a download link. A folder holding anything that expires within 48 hours repeats that warning at the top. The clock starts when the file is uploaded, not when it is posted, so a video sitting unposted is using up its life.

You cannot. Neither app has a delete control, and there is no route behind one. Files leave by expiring: 24 hours from upload on the free plan, 90 days on paid. If something has to go sooner, email support@noisey.app.

Two things follow from that. The stored-file limit counts what you are holding, so posting files and letting them expire is what keeps you under it. And an identical file is recognised rather than taken again, so an accidental repeat costs you nothing. A corrected copy, though, is genuinely new. It lands beside the original instead of replacing it.

Then nothing has been POSTed to it. The request routes a file, never the folder page. Add ?folder=<the name> to your webhook URL, or point a Drive source at it. The empty folder prints the exact query string to use. Files sent with no folder are not lost. They are listed under No folder. With no folders at all, the page offers Make your first folder.

Tokens made before the account-wide URL existed. Each one was tied to a single folder. They are listed under that folder and keep working until you revoke them. Some are older still and show “made before copying existed - recreate it once”, with no copy button. The account-wide URL replaced them, because a pipeline feeding three folders meant juggling three secrets.

Google Drive lives in Media now, at my.noisey.app/media. There used to be a separate Sources page. Each connected Drive folder is shown under the NOISEY folder it fills, as “N files seen · checked <time>”. Add one with Watch a Drive folder at the top of the page.

Yes. The top level of the picker lists your drives, and shared ones have a people icon instead of a folder icon. Browse into it and pick the folder as usual. You choose it by looking at it rather than by pasting an id, which stops a source watching the wrong folder under a name that looks right.

Press Check now on the source. Left alone, our servers check it every few minutes, so a Drive folder takes in files with your computer off. The posting still waits for your computer. Check now is for when you have just dropped a file in and want to watch it land.

Press the × on the source card. It works at once and asks for no confirmation. Files already taken in stay in their NOISEY folder until they expire. Nothing changes in Drive. NOISEY reads the folder you picked and never writes to your Drive.

Use the Posts as dropdown on the phone's row at Phones. Pick the profile, and it takes effect immediately. Nobody yet removes it. One profile per phone, on purpose. A phone posting as two clients makes “which account did this job post to?” impossible to answer, and that answer has to reach Activity and the calendar. The desktop app can set the same thing. Both write the same field.

It opens the profile that phone posts as. If it has none, it takes you to Profiles, where you can create one. That is quicker than hunting for the client by name and coming back.

You cannot, from either app. There is no delete control and no route behind one. A phone stays registered after you unplug it, which is what lets an automation you build today reach it next week. It also means a phone you have retired keeps counting against your plan limit, which is 2 free and 10 paid. Email support@noisey.app to have one taken off.

Because the page is read-only on purpose, and our servers never touch a phone. Everything you do to a phone happens in the desktop app: mirroring, tapping, installing, screenshots and batch operations. The desktop app is the only thing physically attached to the phone. A reboot button here would file a request and hope.

There is no group manager on the website. You make a group while building an automation: choose Several phones…, tick the phones, and name it. From then on it is offered anywhere you can pick a phone. The desktop app has a proper Groups list. The website only has this way in.

Go to Profiles → New profile. A name is the only thing required. A handle and a main platform are optional and can wait. You land on the profile itself, where you edit the brief, the accounts and the posting schedule. Creating a profile does not put it on a phone. Use the Posts as dropdown on the Phones page for that.

Go to Phones → the Posts as dropdown on the phone's row → pick the client. That is the whole step, and it takes effect immediately. The phone stops being called SM-G986U1 everywhere it appears, and Activity can answer “what happened on Jane’s phone today?”. A profile with no phone can still be written and briefed. It just cannot be scheduled.

Open the profile → Accounts → pick the platform, type the @handle, and paste a link to the profile if you have one → Add account. One client usually has several. The same person on LinkedIn and TikTok is one profile with two accounts, not two profiles. This is how a notification finds its way home: a reply to @jane on LinkedIn becomes this client, then their phone, then a routine.

No, and there is deliberately no button that would. You write this. NOISEY never invents a voice for an account. There is also no control for varying a voice across phones, and there will not be. A brief describes a real person who holds a real account. A control for making accounts sound unrelated to each other only makes sense if they are not.

Open the profile and use Picture → Add a picture. A square image looks best. A wider or taller image is cropped from the centre.

Your computer shrinks it to 128 pixels before uploading, so a large photo is never sent and then refused. The picture then appears wherever that profile does: the profiles list, the phone it posts from, and the calendar.

Your own account picture is a separate thing, set under Settings → You.

A read-only strip of platform icons, with the profile's main platform highlighted. It shows what the profile is set to. It is not where you add accounts. That is Accounts, further up the page.

Everything booked to run in a week: scheduled automations in green, and scheduled flows in amber, faded when paused. Automations that fire when a file arrives never appear here. They have no time, so there is nothing to draw, and an account that works perfectly can have an empty calendar. The server works out the times, so the calendar and the job agree on what “09:00 in Europe/London” means.

← Previous, This week and Next → above the grid, with the date range and “shown in <your timezone>” beside them. This week is greyed out when you are already on it.

Yours. A schedule set in another zone appears at the hour it will actually run where you are. 09:00 New York sits at 14:00 in London. The entry's own zone is in the hover tooltip. This is why the schedule form asks for a named zone rather than an offset: the conversion has to survive a clock change.

The time, the automation or flow name, and the client. The client is the profile name, or the phone name when the phone posts as nobody. The tooltip tells you which phone does the work: name · as <profile> · on <phone> · timezone. Include paused, on by default, keeps paused items visible and faded.

Then nothing you have is on a clock. Automations that fire when a file arrives never appear here, so check Automations before assuming something is broken. The empty state offers the two ways to fill it: Schedule a phone and New automation.

Go to Activity, which has two tabs. Timeline mixes jobs, connections and failures together, per phone and per client. That matches the question people actually ask: “what happened on Jane’s phone today?”. Jobs is the narrow view: just the work, with its errors. Timeline loads 150 events. Jobs loads 100.

The All phones dropdown on the Timeline tab narrows it to one phone. Only trouble keeps warnings and errors and hides everything else. When that toggle finds nothing, it says so plainly: “Nothing has gone wrong. No warnings or errors in this window. That is the good outcome.”

• queued - the server has the job and is waiting for one of your computers to be connected.
• dispatched / running - it is with your computer, or on the phone now.
• success / failed - finished. A failure prints its reason on the row, in red.
• missed (amber) - the start window closed before your computer was there. Nothing broke and nothing posted.
• cancelled - stopped before it ran.

A row also shows the number of attempts when there was more than one.

Free: 7 days. Paid: 90. Events drop out of the window rather than hiding behind a filter, and there is no export. So read anything you need for a monthly report while the window still holds it.

Nothing has run yet: “Connect NOISEY on your computer and the timeline fills as your phones work.” If you know jobs have run, the page is more likely signed out than empty. See the blank-page answer in the dashboard section.

In the editor's run panel, for the run you just started, and nowhere else. There is no flow run history page yet. To leave a permanent record, give the flow a Write to the timeline step. That lands in Activity with everything else. Add one wherever a flow makes a decision you will later want to explain.

Each row shows the kind, the author, how many installs it has, its rating and how many people rated it (or “unrated”), which version of NOISEY it was built against, its age, and the author's description. The Everything, Routines and Flows chips filter the list. That “built against” line is the closest thing to a freshness date, so read it before an old routine surprises you.

Press one of the five icons on the row, which are labelled one to five stars. Rating again replaces your previous rating, and the row then shows the new average. It does not show what you gave it, so write your own verdict down if you want to remember it.

Free: 3 live listings. Paid: unlimited. Only a drawn routine can be published. A routine written in sentences is a set of instructions for your own phones and does not move to someone else's, so the picker says “No routines to publish yet” when nothing qualifies. Flows can be published as they are.

Filter the store by Mine to see only what you published, then press Unpublish on the listing and confirm.

It disappears from the store immediately. Copies people already installed are theirs, and are untouched. Taking a listing down stops new installs. It does not reach into anyone’s account and remove a routine they are using.

Six: Announcements, where staff start the threads and anyone replies; General; Automations & routines; Phones & hardware; Feature requests, one request per thread so that likes mean something; and Show & tell. The chips along the top filter to one topic. All returns to everything.

1. Community → Start a thread.
2. Pick the topic.
3. A Title, up to 120 characters, and the post itself, up to 10,000.
4. Post thread.

You land on the new thread. If your account has no display name yet, the composer asks for one there and then. That is the moment you need it, so you don't have to go to Settings first.

Use the Reply box at the foot of a thread, up to 5,000 characters, then press Post reply. The heart on any thread or reply likes it and fills in. Press it again to unlike. Both have a cooldown: one new thread every 2 minutes, and one reply every 10 seconds. They exist to stop a script, not to slow you down.

Staff pinned the thread to the top of the list, or closed it to new replies. A locked thread replaces its reply box with “This thread is locked. Staff closed it to new replies.” Only staff can do either. You can still read a locked thread, and still like it.

Press Load older threads at the foot of the list. It works from a marker rather than page numbers, so threads posted while you are reading do not push what you have already seen onto another page.

Go to Settings → AI for flows. Pick the provider, OpenRouter or Anthropic. Paste the key, name a model if you want to, and press Save key. Success reads “AI key saved. Flows can now use AI while your computer is asleep.” The card then shows which provider is stored, the model in use, and how many AI actions you have spent this month.

No, not even to you. The card can say that a key exists, replace it, or delete it. That is the whole list. A field that can show a secret is a field that can leak one, so keep your own copy with your provider.

Press Delete key and confirm: “Delete the stored key? Flows that use AI will stop until you add another.” Your computer's own key is untouched. That one lives in the system keychain, either the macOS Keychain or Windows Credential Manager, and is a separate thing. Deleting this one only affects flows, which run on our servers and cannot reach a keychain on a closed laptop.

The server is missing the encryption key it would protect yours with, so it refuses rather than storing a secret unencrypted: “this server cannot store a key securely - use the key on your Mac instead.” Routines run by the desktop app still work. Only a flow's Ask a model step is affected. Tell us at support@noisey.app. That one is ours to fix.

No. Settings keeps two separate lists, and they do not overlap. Sessions → Sign out everywhere else ends other browser sessions. Your computers → Sign it out disconnects a Mac. For a computer you no longer have, do both. Signing a computer out stops it syncing and getting jobs. It keeps controlling the phones in front of it, which is the point of the desktop app working offline.

Nowhere in the dashboard yet. There is no plan indicator, no usage meter, no upgrade button, and paying is not set up. Today the only place your plan shows up is the message when you hit a limit, such as “automations limit reached (1/1)”. Assume you are on the free plan unless you have arranged otherwise with us. The numbers for each plan are in this help centre.

There are none. NOISEY only sends email you asked for: confirming an address, resetting a password, and confirming a change of address. So there is nothing to opt out of, no digest, and nothing to set up. The newsletter on noisey.ai is separate, and you opt in to it.

Press the sign-out icon at the foot of the left rail, third of three after Settings and the docs book. It signs out this browser only. Other browsers, and any connected computers, keep their own sessions until you end them from Settings.