xhs-publisher_skill

Overview

This skill documents the Xiaohongshu (Little Red Book) creator platform publishing flows and step-by-step operations for both short image-text posts and long-form article posts. It provides the correct publishing entry URLs, required field formats, DOM targets, and safety guidance for automation scripts. The goal is to reliably prepare content and stop before the final publish action so a human can review and confirm.

How this skill works

The skill describes two main flows: 'image-text' for posts under 140 characters that use text-to-image generation, and 'article' for long-form posts (≥140 characters) using the rich text editor. It lists the exact entry URLs, element identifiers (buttons, textboxes, and template selectors), and recommended sequencing: fill all content first, then generate images or apply layout templates, then populate the final publish page fields. Automation should fill fields and pause for human confirmation instead of clicking the final publish button.

When to use it

• When preparing short image-text posts (< 140 characters) that require generating images from title and body.
• When composing long-form article posts (≥ 140 characters) needing full rich-text content and one-click layout.
• When automating repetitive content entry while requiring a human to confirm publish.
• When you need exact DOM targets and sequencing to avoid input failures on custom editors.
• When you must follow platform constraints like title length, tag formatting, and image limits.

Best practices

• Always use the correct entry URL: target=image for image-text, target=article for long-form.
• Fill all content first (title + body) before clicking generate or applying templates.
• Separate tags from body with two newline characters and use # for each tag, max 2–3 tags.
• Add 1–3 second waits between automated actions and randomize delays to reduce detection.
• Do not auto-click the final publish button; present content for manual review and confirmation.

Example use cases

• Batch-prepare product-review posts as image-text: input title and body, generate images, then stop for review.
• Automate long-form drafts: populate rich-text editor, apply one-click layout, and prefill publish fields for editor review.
• Integrate into a content pipeline to convert blog posts into Xiaohongshu-ready notes with tags and cover images.
• Use the DOM selectors to build UI tests that verify auto-save, template selection, and image-generation behavior.