Fetch my logged-in X feed using View webpage and summarize my feed.
Important: You must NOT write code to parse HTML or markdown. The view_webpage tool returns cleaned markdown content that is ready for you to parse.
Important: You must complete the ENTIRE workflow end-to-end, your work is NOT done until the database has been udpated.
Workflow
-
Fetch the feed: Use View webpage with url x.com/home
- The user must be signed into their X account in Zo's browser
- If the user does not appear to be signed in, instruct them to sign in first
- Use the markdown text returned directly by the tool output - do NOT read the saved markdown file
- Note: The page may show a "javascript disabled" warning at the top, but actual tweet content will appear below it. Do not assume failure based on the warning alone
-
Extract tweet data: Parse the returned markdown text to extract tweet objects with these exact keys. If multiple tweets from the same author appear and appear to be part of a thread, join them into a single tweet.
content: tweet text (string, may contain newlines, quotes, special characters)author_url: absolute URL to author's profile (string)author_name: display name (string)url: absolute URL to the tweet (string, used as unique identifier)type: a string column, with value always set to 'feed'- IMPORTANT: continue to write and execute the script.
-
Summarize the tweets: Write the tweets in your response in a compact bullet list format.