The video tool now has nicer looking log out put.

Inspector:

1. Post an article to website from the inspector panel
2. Set the cover image for the article,
3. Pick the language,
4. Pick publishing date with a calendar
5. Set Author name,
6. Pick category,
7. Pick Formats,
8. Additional tags
9. Delete button is snow more accessible, it is the first item in the Inspector
10. Save settings button
11. Publish button
12. File upload in the chat
13. Added MedGemma 4b for image recognition
14. Token counter used and max available
15. Reset session button

Quick update. The video tool is now looking nicer. It will not fill the console log but instead count things. We now have great improvements with the inspector panel. We can now set cover images for articles just by drag and dropping or clicking on it. We can pick the language. We can set a post date with a calendar. That's nice. We can set the author name. We can set the category. Video blog. Short stories. Kurzgeschichte. Cuentas Nuevos. That will determine in what language our post is published. We can set the format for the downloads. EPUB. Sie. One-sheet. We can include the post image in the article or not. And we can choose to have table of contents or not. In case we haven't tracked and dropped any images we can add location of the image right here. Image all text. We can add additional tags. Save the settings. And I can publish directly to Eleventy host website. Move the delete button to the top. Because I got tired to always have to scroll down if I actually wanted the things. Which I will not do. Now here's a good example. We have the cover image. The language set. The date. And the kicker. Which is really cool. I can now just publish again. And he will just overwrite. So if I have an article in my webpage. And I want to change the image. I can just do it. From the inspector. And then publish. Be changed. Awesome. Now this is so fat. We now have file upload. In the chat window. So I can click here. Attach a file. I can just drag it up into the text field. I have added Medjammer. If I ever have the need to look at an x-ray. And get a doctor's opinion. Which claimed to be 80% right at times. And this is a very small model. And the idea is to have it even running on device. So the patient's information doesn't lead to the evildoers. This is also cool. We now have down here. In the right corner. A token counter. We start off with 1k. 1000 tokens. And he will automatically show us the MEX tokens available. So if I talk to OpenAI. I say hi. Hi. What can I help you with today? He will always update the tokens that are available. And the tokens that I have used. Per model. So if I switch to Mr. Large as an example. I will update the tokens accordingly. That I can use. In that session. I have the little icon here. Where I reset the context. And now I'm back. At my beginning context. Another thing that's new. I have taught him to either read a file in our sandbox. The transcript folder. Or open the file. Which he will then add to the binder as a new text document. I can also now ask him to read all the tabs. In my current browser session. Because I had 46 open tabs. Constantly. Because I haven't built Mudd yet. Didn't know if I'm going to have a look at them again. So I can ask him. Show me a list of open browser tabs. With a name and URL. And he gets to work. And shows me a list of my current open tabs. And I can even be more specific. And say. Show me a list of my open browser tabs. But only the YouTube videos. With title and URL. And bam. He's listing. All the YouTube videos I have open. I can probably imagine. To write the next plugin. To give him. The right to close browser tabs. And then I can just pick and choose. What I want him to close. That's it for the nightly update.