Guide · 7 min read
Form builder API and automation guide
Updated 2026-09-17
TL;DR
A form builder API typically lets you create and manage forms with an API key over REST, get notified of new responses through signed webhooks instead of polling, and increasingly exposes an MCP server so AI coding agents can create and read forms directly.
Once a form is more than a one-off, you usually want it wired into something else: a script that creates forms programmatically, a system that reacts the moment a response comes in, or an AI agent that builds and edits forms as part of a larger workflow. This guide covers the three pieces that make that possible: the API itself, webhooks for real-time notifications, and MCP for AI agent access.
Authenticating with an API key
A REST API secured by API keys is the standard way to script form creation, list responses, or pull data into another system, without a browser session. Keep the key server-side; anything shipped to a browser is effectively public.
An OpenAPI spec (a machine-readable description of every endpoint) lets you generate a client library in your language of choice instead of hand-writing HTTP calls, and it's also what most API-testing tools use to autocomplete requests.
What a v1 REST API usually covers
Expect endpoints to create and update forms, list and fetch individual responses, and manage the basics of a form's configuration. Version the API from day one (v1, v2) so future breaking changes don't silently break scripts built against the current version.
Webhooks for real-time updates
Polling an API on a timer to check for new responses wastes requests and adds latency. A webhook pushes a payload to your endpoint the moment a response comes in instead, which is both faster and cheaper than polling.
Signed webhooks (HMAC-SHA256 is the common standard) let your endpoint verify the payload actually came from the form platform and wasn't forged, and automatic retries on delivery failure mean a brief outage on your end doesn't silently drop a submission.
CSV export and Slack notifications
Not every integration needs custom code. CSV export covers ad hoc analysis in a spreadsheet, and a Slack notification on new responses covers the common case of just wanting a team channel to see submissions land, without building anything.
MCP: letting AI agents work with forms directly
MCP (Model Context Protocol) is a way for AI coding agents like Claude Code, Cursor, or ChatGPT to call tools directly, rather than a human copying API responses back and forth. A form platform that exposes an MCP server lets an agent create a form, read responses, or adjust settings as part of a larger task, using the same underlying API but through a standardized interface the agent already knows how to call.
This matters increasingly for teams that build internal tools with AI agents: instead of writing custom API integration code for a one-off form task, you point the agent at the MCP server and describe what you want.
YeetForm's REST API v1 (docs at yeetform.io/developers) is authenticated with API keys, ships an OpenAPI spec, sends signed HMAC-SHA256 webhooks with automatic retries on new responses, and exposes an MCP server so agents like Claude Code, Cursor, or ChatGPT can create forms and read responses directly.