Half a year ago a content operator asked me whether her team could batch publish on Xiaohongshu at all, given that nobody there wrote code.
I asked what publishing one post involved. Open the app, pick photos, write the copy, add hashtags, tap publish. Two or three minutes each, a dozen or more a day.
I said none of that needs code. Just say the steps out loud. She was sceptical, tried it that afternoon on one phone, and three days later told me the flow was already saved as a template and now runs with a single click before she leaves.
That said, let me pour some cold water first: not everything can be done this way. Drawing the boundary clearly is more useful than only listing what is possible.
What works, what does not
Tasks you can hand off without code share one trait: the actions are on screen, and the judgement comes from looking.
| Task type | Without code? | Notes |
|---|---|---|
| Repeating a fixed flow | Yes | Open an app, tap, type, upload images, submit |
| Deciding the next step from the screen | Yes | Close a dialog if it appears, treat a completion label as success |
| Capturing evidence | Yes | Screenshot each step for later human review |
| Running one flow across many phones | Yes | Dispatch to a whole group |
| Pulling API data into a form | No | Needs code to call the API |
| Branching on a return value | No | Complex conditions need code |
| Integrating with CI | No | That is engineering integration |
| Writing results into a database | No | Needs data handling logic |
That table should tell you which side you are on. If your work is repeating the same actions daily on a few platforms, you are in the yes column. If it involves processing data, making decisions and connecting systems, you need code, and forcing a no-code route will only create friction.
The two conversation modes, and why picking wrong is tiring
The conversation page offers two modes. The difference is not capability, it is whether the requirement gets clarified first.
Conversation mode suits a case where you already know what to do. Type something like open Settings and wait three seconds, and it plans and acts immediately. For simple requests this is the least effort.
Refinement mode suits a case where you only have a vague idea. You describe roughly what you want, and it asks back: what does the screen look like, how long after the tap, what if a dialog appears. After a few rounds it turns your request into clearly ordered steps, and you confirm before it runs.
For a first attempt at anything, use refinement mode. The reason is practical: you do not know the screen well yet, so a single description tends to miss details, the run stalls on the third step, and you cannot tell whether the problem was the description or the action. Talking it through first lands more reliably.
Refined descriptions can be saved on the prompts page and pulled up next time instead of being rewritten.
Turning a working flow into a template
Conversation is for trying. Workflows are for running long term.
When a flow has been clean for several days, save it as a workflow. Afterwards you only pick a phone and parameters, with no re-describing.
A few canvas features are worth using:
| Feature | What it is for |
|---|---|
| Validate | Checks step connections and parameters so errors surface before the run |
| Arrange | Tidies the layout once steps multiply |
| AI workflow collaboration | Generates or edits steps instead of building from scratch |
| Trial run | Exercises the flow on one phone before committing to it |
The key design decision is leaving the changing parts as parameters. In a posting flow, the fixed parts are open the app, pick photos, tap publish, and the changing parts are the copy and image numbers. Turning the changing parts into parameters means one template covers every post, rather than one flow per post.
Always trial run first. Results are recorded in execution history, which is more reliable than watching the screen.
When an icon button cannot be recognised
One category of interface causes trouble: pure icon buttons with no readable text, and heavily custom-drawn screens. Text recognition cannot find those, and the AI gets stuck.
The fix is image templates. The method is plain: capture a small image of the button, save it, and describe three things.
Name: short, such as publish entry. Description: the visual characteristics, such as orange circular plus button at the bottom of the home screen. Usage: when to apply it, such as when the publish entry is needed after reaching the home screen.
On the conversation page you can attach these entries to the current session. Patching every easily confused icon with a template is what makes a no-code setup stable.
How to run your first task
If you decide to try, follow this order and you will see a result within half an hour.
Install and confirm the phone is online. Follow the installation selection guide, then scan and register the device on the devices page and confirm it shows as online.
Pick something you were going to do today anyway. Not something complex. Writing one line in Notes is enough. The point is proving the path works, not testing the ceiling.
Describe it in plain language. Start a new conversation, select the phone, and say what you want. Use refinement mode the first time.
Send it and watch the phone. If nothing moves, read the error in execution history. Most failures come down to two causes: an offline device, or an automation environment that never started.
Save what works as a workflow. Add a schedule if it needs to run daily. One caveat: the workstation on the computer has to be on, and the target phone online at that moment.
Once that first task runs, you will see that the value of a no-code route is not that it does everything. It is that it takes away the daily repetition that is not worth hiring someone to script. The operator from the beginning of this piece later turned posting, comment replies and evidence capture into three templates, saving about two hours a day.
The tasks that need judgement and data still belong with the colleagues who write code. Keeping that division clear is what makes the whole thing work. Install it and try one small task to find out which side your work falls on.
Frequently asked questions
- Can I really do phone automation with no coding at all?
- For tasks built on interface actions and visual judgement, yes. You describe what to open, where to tap, how long to wait and what counts as done, and it runs on a real phone. Publishing, liking, replying, filling forms and capturing evidence all fall in that range.
- What can I not do without code?
- Anything needing precise values or complex branching. Pulling data from an API into a form, deciding which path to take from a response, integrating with CI, or writing results into a database all still need code.
- Conversation mode or refinement mode, which should I use?
- It depends on how clear the requirement is. If the steps are already decided, conversation mode sends one instruction and it runs. If you only have a rough idea, refinement mode lets it ask questions first, filling in ordering and exception handling. For a first attempt, use refinement mode.
- Do I have to describe the same task every time?
- No. Once a flow runs cleanly, save it as a workflow template, and afterwards you just pick phones and fill parameters. Values that change each time, such as titles, prices and image numbers, become parameters, so one template covers a whole batch.
- What if a button is a pure icon with no text?
- Capture an image template for it: save a small screenshot of the button and describe what it is and when to use it. Once that entry exists, the AI selects it when it needs to find that button. This is the most effective fix for icon recognition.
- A run stalls halfway. How do I find where?
- Check execution history. Every task records its source, whether it is conversation, schedule or a workflow trial, plus execution status, target device and error message. Device offline points to the device page, automation environment points to the install page, and unauthorised points to workflow status.
- Can it run on a schedule?
- Yes. Scheduled tasks come in workflow schedules and conversation schedules, and fixed-step work suits workflow schedules. The workstation on your computer has to stay on, the target phone has to be online, and USB-connected devices need the connection intact at that moment.
- Do several phones running at once interfere with each other?
- No, because each device runs its own session. That said, on a first batch run it is worth staggering the start times slightly, not because of software conflicts but because it makes it obvious which device is which when something goes wrong.
Try it without code
Describe the task once and let the phone do it
Software is completely free and runs on your own computer. No coding needed.