Web runtime via llms.txt. ChatGPT writes the API calls.
Updated March 2026
ChatGPT doesn't support MCP yet. But it can still work with Sutrena -- give it the /llms-full.txt reference and your API key, and it generates working curl commands or Python scripts. Code Interpreter can even run them directly. Pages, forms, dashboards through conversation.
1. Get a Sutrena API key
No signup needed.
curl -X POST https://sutrena.com/api/trial
# Response: { "data": { "key": "st_trial_xxx", ... } }2. Give ChatGPT the API reference
Paste the /llms-full.txt URL or content into your conversation. It's a plain-text file built for LLMs -- every endpoint, parameter, and example.
# Option 1: Paste the URL and ask ChatGPT to read it
"Read https://sutrena.com/llms-full.txt and use it
as a reference for the Sutrena API"
# Option 2: Copy the content and paste it directly
curl https://sutrena.com/llms-full.txt | pbcopy3. Create forms through conversation
Ask ChatGPT what you want. It generates the API calls -- curl, Python, JavaScript, whatever you prefer.
# Example prompt:
"Using my Sutrena API key st_trial_xxx, create a
waitlist form with name and email fields, then create
a dashboard with a submission count and a line chart
showing signups per day."
# ChatGPT will generate the curl commands or Python
# code and you can run them directly.4. Use Code Interpreter for automation
With ChatGPT Plus, Code Interpreter can run Python that hits the Sutrena API directly. It creates the form and verifies it worked, all in one go.
# ChatGPT Code Interpreter example (Python):
import requests
API_KEY = "st_trial_xxx"
BASE = "https://sutrena.com"
headers = {
"Authorization": f"Bearer {API_KEY}",
"Content-Type": "application/json"
}
# Create a form
form = requests.post(f"{BASE}/api/forms", headers=headers, json={
"name": "Waitlist",
"fields": [
{"name": "name", "label": "Name", "type": "text", "required": True},
{"name": "email", "label": "Email", "type": "email", "required": True}
]
}).json()
print(f"Form created: {form['formId']}")
print(f"Hosted URL: https://sutrena.com/f/{form['formId']}")5. Set up a custom GPT (optional)
You can also build a custom GPT with the Sutrena schema as an action. Then anyone using your GPT can interact with Sutrena without touching curl.
# In the GPT Builder:
# 1. Go to "Configure" > "Actions" > "Create new action"
# 2. Import schema from URL:
# https://sutrena.com/api/schema
# 3. Set authentication to "API Key" with header
# "Authorization" and prefix "Bearer "
# 4. Users of your GPT can now say
# "Create a feedback form" and it worksNot yet. That's why we go through the REST API instead. If you want native tool-calling, use Claude or Cursor.
A plain-text file describing the entire API -- endpoints, schemas, auth, error codes, examples. It follows the llms.txt convention. Think of it as API docs formatted for an LLM to read.
Yes. It can generate the dashboard DSL and call POST /api/dashboards. Give it the /llms-full.txt reference and it knows the widget types -- metric_card, pie_chart, bar_chart, line_chart, data_table, text_block, action_table.
Use a trial key for experiments. Unclaimed trial accounts auto-delete after 24 hours -- visit the claimUrl to keep your data. For production, make a dedicated key and rotate it if needed. Don't put your main key in a shared GPT.
Sutrena is the web runtime for AI agents. Three primitives — pages, forms, and dashboards — accessible through one API. Your agent creates web artifacts, humans interact with them, and your agent gets the data back. Framework-agnostic. Works from any MCP client or HTTP client.
1. Get a trial key (no auth, no signup)
curl -X POST https://sutrena.com/api/trial2. Create a form + dashboard from a template
curl -X POST https://sutrena.com/api/forms \
-H "Authorization: Bearer st_trial_xxx" \
-H "Content-Type: application/json" \
-d '{"templateId": "waitlist", "createDashboard": true}'Get a trial API key instantly with no signup, or create an account for the full experience.