How to Automate Bulk Xiaohongshu Posting with AI Phone Control

Why Bulk Xiaohongshu Posting Slows You Down

Publishing a single Xiaohongshu note takes a handful of taps: open the app, tap the compose entry, pick images from the gallery, write the caption, add topic tags, choose a cover, set visibility, and publish. On its own, that is a few minutes of work.

The problem is not difficulty. It is repetition, and the repetition is nearly identical every time. The compose entry sits in the same place, the image picker behaves the same way, and the caption and topic fields are the same two fields. Publishing a batch — a week of notes in one sitting, or the same story across several phones — means most of your time goes into tapping the same spots instead of deciding what to say.

Stage of posting	What you do by hand	Good candidate for a template?
Reach the compose entry	Find and tap it	Yes — the entry point rarely moves
Pick images	Choose from the gallery in order	Yes, if the image order is predictable
Write caption and topics	Type free text	Yes — leave the text as a parameter
Set cover and visibility	Tap through a couple of settings	Yes, if the settings stay stable
Decide what to post	Write and choose the angle	No — keep this with a human

Before you build anything, line up four things: your account context, so the AI picks the right account; your captions and topics, which become parameters; your image source in the phone gallery, in the order you want them picked; and your target phone, with a clear alias. Then start from the install selection page and confirm at least one phone shows as online in Devices.

Step 1: Describe the Posting Loop in a Conversation

The first real step is not configuration — it is describing the flow clearly. Open Conversation, pick an online phone, and describe the publishing steps in plain English. There are two modes:

• Conversation mode — you know the steps, so you send the instruction and the AI plans and operates the phone right away.
• AI refine mode — you know the goal but not every step, so you talk it through with the AI first, filling in what to tap and what to type before you run it.

For a first attempt, use AI refine mode. A prompt you can start from:

Open Xiaohongshu, go to the compose page, select the latest image in the gallery as the cover, fill the caption with the text I give you, add my topic tags, set the cover and visibility, then stop before publishing.

Note the last line. On a first run, hold irreversible actions like the final publish for human confirmation; remove the limit once the flow behaves reliably. A description you like can be saved in Prompts and reused for the next batch.

Step 2: Freeze It Into a Reusable Workflow Template

A conversation is good for trying a flow; a workflow is good for running it again and again. In Workflows, create a workflow in the library and open the canvas. A typical sequence reads:

Start automation environment → wait for the interface → open Xiaohongshu → enter compose → select gallery images → fill caption → add topics → set cover → set visibility → publish.

Three canvas features are worth using here:

Canvas feature	What it does for a posting flow
Validate	Checks the links and parameters before you run
Auto-layout	Tidies a long sequence so the flow stays readable
AI workflow collaboration	Generates or edits steps so you do not build from scratch
Test run	Runs the flow on one online phone before you rely on it

The key design decision is to leave the changing parts as parameters. Caption, topics, and image index differ on every note, so pass them at run time. One template then covers the whole batch instead of one workflow per note. The short video publishing workflow guide follows the same pattern.

Step 3: Test-Run on a Device and Add Image Templates

Before trusting a template, run it once: pick one online phone, use the test run option in the canvas, then check the result in Execution History.

The AI reads the screen through text recognition and interface understanding, which handles most screens well. Two cases are harder: icon-only buttons with no text, and custom-drawn interfaces. The Xiaohongshu compose entry is a common example.

The fix is Image Templates. Crop the button into a small image, store it in the library, and describe it in three parts:

• Name — short, for example “Compose entry”.
• Description — visual traits, for example “orange round plus button at the bottom of the home screen”.
• Usage — the trigger, for example “use after reaching the Xiaohongshu home screen”.

In Conversation you can attach templates to the current session, and the AI consults them when it needs to locate a button. Adding templates for every ambiguous icon is the most effective way to make bulk Xiaohongshu posting stable.

Step 4: Set Your Posting Cadence With Scheduled Tasks

Once the template is reliable, put it on a rhythm. The Schedule page supports two kinds of schedules:

• Workflow schedule — runs a fixed template on a chosen phone at the set time.
• Conversation schedule — replays a conversation’s content at the set time.

Posting has fixed steps, so a workflow schedule is usually the better fit. Two prerequisites apply: your computer must keep the iEasyRun workstation running, because scheduling is local, and the target phone must be online or able to connect at that moment.

After each run, review Execution History for the task source, status, target device, and error message. A fixed order saves time: an “automation environment” error sends you to the install page, “device offline” to Devices, “unauthorized or expired” to the workflow library, and an image or OCR failure to the image template library.

The Bottom Line

Automating bulk Xiaohongshu posting is not about letting AI decide what you post. It is about removing the taps you would repeat anyway. Get a phone online, describe the flow in a conversation, freeze it into a workflow template with parameters, add image templates for awkward buttons, and run it on a schedule while checking execution history.

The judgment — the angle, the copy, the timing — stays with you; the repetitive tapping moves to a real phone driven by AI. iEasyRun is completely free, has no paywall on its features, and runs on your own computer. Try it with one phone and one note before scaling, and see the installation guide to start.