Hi everyone!
Sharing the results of the first day of building a panel for running a Telegram channel.
1. Built a website where I can now add my Telegram accounts via the API and load the channels I'm going to run. I also connected the ad account via the API, so now, through its MCP, I can place Telegram Ads with an agent and this site.
To get there I had to struggle through KYC on Fragment and become a small holder of the GRAM token.
2. Added neural network support for proofreading, clean formatting and image generation. Now I can drop thoughts and screenshots from my work here, edit several posts in parallel, and publish them once they're ready.
3. I'll write articles about every project. I thought for a long time about how to show you the project as openly and informatively as possible - the prompts, the models' answers, the intermediate results and so on - and decided the articles will be based on a retrospective analysis of my prompts, which I'll track in apiToken.
Now I'll be able to map them out for anyone interested and show how much time and money the development cost.

October 12, 2026 7