Punk XHS Cards

Punk XHS Cards is a Codex Skill from the GitHub user Adrian Punk that turns your character and a long Markdown article with screenshots into a set of vertical knowledge cards ready to post on Xiaohongshu. The first time, it builds a reusable personal character from your photo (or registers a character image you already have) and settles your name, the pose you strike on the cards and a theme color; if you don't pick a color, it suggests one from your clothes or accessories for you to confirm. It then parses the article's title, structure and key screenshots, calls another Skill to draw a 16:9 theme illustration, composes a 1080×1440 cover, and lays out the body cards from 01 on, checking for missing images, cut-off text and off-scale figures. At the end you get a suggested title, three alternatives, the post text and 6 to 10 hashtags. There are three checkpoints along the way, for the character, the card style and the cover, and nothing is batch-generated until you approve.

Who it's for:writers who post long articles on X or WeChat and want to repurpose them as Xiaohongshu cards featuring their own character.

Before you start: you need Codex with image generation available, the same author's Punk IP Article Illustrations Skill installed first, and Python 3.10+ with Pillow; on Linux you'll also need Chinese fonts. Your photos, character profile and articles stay in a local working folder, but when images are generated your reference photos go to Codex's image generation service, and the author asks you to decide whether that's right for your platform and account. Using your own photo is fine; get permission before using anyone else's photo or character. It only writes the post copy and never logs in to or posts on Xiaohongshu for you. The license is free for personal, non-commercial use; companies and organizations (even non-profits), brand accounts and client work count as commercial and need the author's written, paid permission first. The old MIT license only covers earlier releases.

Based on the project's public documentation. Not tested by AI Inspiration World.