social-push

Social-push is a specialized automation skill for AI coding assistants that leverages agent-browser to streamline the process of publishing content to various social media platforms. By automating browser interactions, it eliminates the need for manual navigation and clicking, allowing users to draft articles, tweets, and posts with a single command. It acts as an bridge between your AI development environment and the web-based interfaces of popular social networks, ensuring consistent, efficient, and reliable content distribution workflows.

• Automated browser orchestration using agent-browser with remote debugging capabilities.

• Multi-platform support including Xiaohongshu (photo/long-form), X (Twitter), Weibo, WeChat Official Accounts, Juejin, and Linux.do.

• Self-evolution mechanism allows the agent to verify and fix interaction flows when webpage structures change.

• Safety-first design: The skill defaults to drafting content and restricts automatic final publishing, requiring the user to confirm the content before it goes live.

• Markdown-based configuration files enable quick and simple addition of new platforms without complex scripting.

• The skill requires a browser with remote debugging enabled (e.g., Google Chrome or Microsoft Edge with port 9222).

• Users should handle the initial login session; the skill supports persisting state to maintain long-term access.

• Typical inputs include the target platform, content type, text, and media paths; output is an interactive browser session holding the drafted content.

• It is highly recommended to use the --auto-connect flag to seamlessly link with existing browser profiles.

• The agent relies on snapshot analysis to identify UI elements, which ensures robustness against minor frontend updates on target websites.