These docs as context
Giving Claude, ChatGPT or your editor the Trama documentation so it answers about Trama correctly.
This is the third door, and the cheapest: don't connect your data at all — connect the documentation. Useful when what you want isn't "what's pending today" but "how does the qualification gate work", or when you're writing code against the API and want your editor to stop guessing.
Any page, as Markdown
Add .md to any URL and you get that page as plain text, ready to paste into a chat:
https://trama.so/en/docs/quickstart.mdIt works in the three languages and on every page, the API reference included — there, what comes back is the OpenAPI fragment for that endpoint rather than a transcript, which is more useful: an assistant that reads it can build the request, not just describe it.
The whole site, for a model
| File | What it is |
|---|---|
/docs/llms.txt | An index: every page with its title, description and URL, in the three languages. Small enough to hand over whole |
/docs/llms-full.txt | The entire documentation as one text file |
Start with llms.txt. It's what lets an assistant find the right page instead of holding the whole site in context.
They live under /docs/, not at the root. trama.so/llms.txt is the site's index — the landing, the pricing, the blog — and it points here. If you only want the product documentation, the one you want is /docs/llms.txt.
In your editor
Point your assistant's fetch or docs tool at https://trama.so/docs/llms.txt and let it follow the .md URLs from there. That's the whole setup: no key, no account, nothing private.
A caveat worth stating
An assistant with these docs loaded is quoting documentation, not your account. It doesn't know your plan, your settings, or whether your WhatsApp is connected. When the answer depends on how your organization is set up, the source is the dashboard or the product connector — not this.