Developers · Public Tool API & MCP

ThatApp inside the AI client of your choice.

Mint a token, paste it into Claude, ChatGPT desktop, or any Model Context Protocol client, and that client can reach ThatApp's tools and your connectors directly.

# add ThatApp as a custom connector
endpoint  mcp.thatapp.io
transport  streamable-http
token     tap_•••••••••
→ tools available · flows · data lake · saved views
What you get

A library of tools at one endpoint

Platform tools at one surface and a meta-dispatch interface to your iPaaS connector library at another. Your AI client can query, build, and report.

Mint, list, and revoke tokens

Issue tokens from your dashboard and revoke them when done. A leaked token is recoverable, honored within about sixty seconds.

Scoped per surface

Each token is scoped to a specific surface, so a token for one cannot be reused for the other. Destructive cross-tenant tools are intentionally disabled.

Two transports

Stdio for local clients like Claude Desktop, and streamable HTTP for remote web clients. Use whichever your client supports.

Three steps to a connected client

1 · Mint a token

Issue a scoped token from your ThatApp dashboard in a couple of clicks.

2 · Add the connector

Paste the endpoint and token into Claude, ChatGPT desktop, or any MCP client.

3 · Start working

Your AI client can now query your data lake, build flows, and pull reports through AVA's tools.

Put your tools where you build.

ARTIFICIAL = FAKE
Humans lie. Data doesn't.
Welcome to the world's first Organizational Intelligence platform.