Skip to content
Faldaro

MCP server

Connect Claude to your forms

Faldaro hosts a Model Context Protocol server. Point Claude — or any MCP client — at it with one of your API keys, and your model can build forms, file submissions, run reports and work your data as tool calls, with exactly the permissions that key holds and not one more.

Try it free

  • The whole platform as a toolbox

    Nearly sixty tools, the same catalog the in-app agent uses: folders and forms, drafts and publishing, logic and workflow editing, submissions, transitions, reports, datasets, document and notification templates, test cases. If your key may do it, your model may do it.

  • Your key is the whole permission model

    The server adds no account of its own and holds no credential of its own. Each tool call is authorised as the API key that made it — the same privilege checks and the same tenant isolation as the raw API, enforced in the database.

  • Honest tool annotations

    Every tool declares what it is: read-only tools say so, and tools that can overwrite a draft carry a destructive hint, so a well-behaved client knows when to stop and ask you. Reading is safe by construction — evaluation and previews persist nothing.

  • Build and test forms from your editor

    Ask Claude Code to draft a form, wire its logic, add regression test cases and run them — then open the builder to see the result. The validation that guards the app guards these writes too: a broken logic document is refused before it is saved, not stored and ignored.

Two minutes to connected

One endpoint, one header

Create an API key under Settings → API keys, then hand it to your client. That is the whole setup.

Claude Code

claude mcp add --transport http faldaro https://faldaro.app/api/mcp \
  --header "Authorization: Bearer ffk_your_key_here"

Then ask for the work, not the clicks: “find last week's unassigned submissions and assign them to me”, “add a test case pinning the premium for a $2M building”. The MCP guide has the full tool list, other clients, and what to know before connecting a model to real data.

Questions

Fair questions, straight answers

What is MCP?

The Model Context Protocol is an open standard that lets AI applications call tools on other systems. Faldaro hosts an MCP server, so Claude — and any other MCP client — can operate your Faldaro account: list folders, build and publish forms, file and search submissions, run reports, all as tool calls you can watch.

Is this different from the in-app AI agent?

Same tools, different driver. The in-app agent runs inside Faldaro on our models and is metered in AI credits. Over MCP, the model is the one you already use — your Claude, in your editor or terminal — calling Faldaro as one of its tools. Faldaro meters those calls like ordinary API traffic, not AI credits.

What can a connected model do to my account?

Exactly what the API key you connect with can do, and nothing more. A key carries an explicit grant of privileges chosen when it was made; every tool call is checked against that grant and walled into your organisation by the database itself. There are no delete or archive tools at all — destructive actions stay in the app, with a person.

Which plans include it?

Team and Business. API keys and the raw API stay on every plan, Free included — what Team adds is the model-ready toolbox. And on any plan, MCP tool calls never consume AI credits: the model doing the thinking is yours.

Which clients work?

Any MCP client that speaks the Streamable HTTP transport and can send an Authorization header — Claude Code and the Claude API among them, and stdio-only clients through a local bridge. Connectors that require OAuth sign-in (such as claude.ai on the web) are not supported yet.

Put your process on Faldaro today

The Free plan is free forever — unlimited forms, ten thousand submissions a month, and the same builder, logic engine and workflows as every paid plan.

Start free See pricing

No card required.