Added Nov 3, 2026
social-auto-upload
A command-line tool, with Skills for AI agents, that publishes videos and image posts to Xiaohongshu and other platforms.
- Ecosystem
- Agent Skills
- Skill category
- Marketing
- Creator / organization
- dreammis(GitHub 用户)
What it is
social-auto-upload is a multi-platform publishing tool by GitHub user dreammis. It has one sau command line plus a Skill for each platform, so agents like Codex, OpenClaw and Claude Code can publish videos, articles and image posts for you. The repo has 18 Skills, and Xiaohongshu (Rednote) is one of the supported platforms.
On Xiaohongshu you scan a QR code to log in the first time. After that it reads the local video, cover, images and text you point it to, uploads them and clicks publish. It can also use the platform's own scheduled publishing, which needs at least 2 hours' lead time. All platforms share one set of commands, and each platform's Skill installs separately, so you only add the ones you need. The current Xiaohongshu commands cover login, login check, video upload and image-note upload. It doesn't like or comment.
Who it's for:Creators and social media managers who post video to several platforms and don't mind setting up a Python environment. Readers who don't write code won't get much from it.
Before you start: Clone the repo, install with uv pip install -e ., add patchright's Chromium browser and fill in conf.py; install the Skills with npx skills add. It needs Python 3.10 to 3.12 and runs on Windows, Linux and macOS. Your Xiaohongshu session (cookies and browser local storage) is saved as JSON in a local cookies folder; the older web tool used a cookiesFile folder. The review found no code that sends cookies to the author's own service, but it didn't audit every line. It injects a stealth.min.js script into the page, and on first run it downloads and updates the biliup program on its own. If you let an agent run publishing directly, it uploads with your account and clicks publish for you. Installation doesn't use curl piped to bash. If you use these account automation features, weigh the safety risks and Xiaohongshu's platform rules yourself. AI Inspiration World is not responsible for any consequences. MIT license.
Based on the project's public documentation. Not tested by AI Inspiration World.