More and more, your best thinking comes out of an LLM as a rich HTML page, and getting it in front of people is the awkward part. Email the file and the JavaScript breaks, so it renders wrong; publish it and it is suddenly public on someone else's host. TAG renders your page properly and shares it through a private link to the exact people you choose, then captures their feedback by voice, pinned to the section, as a clean file you can hand straight back to the LLM. It works just for you too, a fast way to view your own HTML output and talk your notes back into the loop.

Privately share your LLM-produced HTML docs with collaborators. They dictate their feedback, you get every word transcribed in an MD, ready for your next loop. Made for the AI era, where HTML makes more sense than decks and docs and voice is the ultimate modality.

Sending is invite-only during beta. If someone gave you an invite link, open it first, then sign in with that same email. You do not need an account to receive and Tag a document someone shares with you, that link just works.