Extract Clean JSON from Any X/Twitter Post or Article

ADHX is a post-reading skill that converts an X/Twitter link into structured JSON through the ADHX API. It supports ordinary short posts and long-form X Articles, returning content in a form suitable for summarization, analysis, and key-point extraction. The workflow uses a direct API request instead of scraping x.com or opening a browser. The API is free and requires no authentication.

The result separates post identity, text, author information, creation time, engagement metrics, and article content. This distinction matters because an article post can have an empty short-form text field while its substantive material lives in the article object. The reader checks that object rather than treating an empty tweet-text value as proof that no content exists.

Use this skill when someone shares an X/Twitter post and wants a summary, an analysis, author information, engagement figures, or the complete content of a long-form article. It applies to post links on x.com, twitter.com, and adhx.com with the user/status/id path structure. The supplied examples include summarizing a particular post and answering a question about its number of likes.

Keep the workflow within that post-reading scope. Required inputs, permissions, safety boundaries, and success criteria must be clear, and the output does not replace environment-specific validation or expert review. If the API returns an error or an empty response, tell the user that the post may be unavailable instead of inferring its contents.

The input is the full post URL. First parse the path to extract username and statusId, regardless of which supported domain supplied the link. Those values populate the documented endpoint:

https://adhx.com/api/share/tweet/{username}/{statusId}

The workflow fetches JSON using curl and then uses the returned fields to answer the actual question. It does not require a separate browser scraping step. A request for a summary uses the available text or article content; a request about likes reads engagement.likes from the response rather than estimating it from the prose.

The response schema includes id, the original post URL, short-form text, and createdAt. The author object contains a display name, username, and avatar URL. Engagement contains reply, retweet, like, and view counts. For long-form posts, the article object includes a title, roughly 200 characters of preview text, a cover-image URL, and full markdown content with images. These fields provide different kinds of information, so article extraction and engagement reporting use their corresponding objects.

ADHX integrates with its share API through an HTTP request made by curl. Installation can use the Claude Code plugin marketplace or manually retrieve the skill file from its repository into the local skill directory. Those installation options supply the reading instructions; the actual data path remains URL parsing followed by the ADHX JSON endpoint. The result can then be consumed directly for the requested explanation or extraction.