E2LLM: The missing Puzzle in the LLM Workflow Mosaic.
We’ve been working with LLMs for a long time. Our field is integrations of open-source solutions.
That means a lot of scripts, connectors, and… forms.
For us, developers, a form is just a link, a simple UI. Who cares if it looks ugly? It works.
For clients, though, it matters. A form has to look nice, clear, welcoming.
The dialogue became routine:
Client: “Why does this form look like [censored]?”
Us: “It’s just the connector between products, functionally it’s fine.”
Client: “We want it nice and welcome.”
Where AI failed
We tried AI.
The answer? “The form is fine.”
Yes, fine as code, fine as function — but not fine as a user experience.
AI couldn’t help because it had no eyes.
We realized: to judge a UI, the AI must see the DOM. Not source code. The runtime snapshot.
Challenge accepted
MCP? Too messy, too many tokens.
Browser dev tools? Messier.
We needed something lean.
So we built it: Element to LLM (E2LLM).
An add-on that delivers the runtime DOM snapshot to an LLM in one click.
Free of charge.
What we use it for
Dev tasks.
QA.
Design reviews (the root cause of E2LLM existing).
Parsing complicated forms.
E2LLM is tiny, but it fills a huge missing piece in the LLM workflow mosaic. Curious if anyone else has fought the ‘ugly form vs happy client’ battle — would you use E2LLM in your workflow?
