Developers

Connect DM Handled to your own systems

Three ways in, one set of keys. Your system sends customers, consents and listings; DM Handled tells your system what happened in the chat; and an AI agent can run the whole account.

REST API

Your system → DM Handled. Keep customers and their consents in step, give the chat a same-browser pass, and push listings as knowledge the moment they change.

REST API →

Webhooks

DM Handled → your system. Signed events when a customer confirms their details, gives or withdraws a consent, or needs a person.

Webhooks →

MCP server

For AI agents such as Claude. The same operations as the API, plus the inbox, replies, bookings, knowledge and analytics.

MCP server →

Before you start

  1. Plan. The API, webhooks and the MCP server are part of the Scale plan.
  2. A key. In DM Handled, open Settings → API & MCP and create a key. It is shown once; keep it on your server.
  3. A webhook (optional). In Settings → Webhooks, add your https address and store the signing secret shown once. Or create it through the API: POST /api/v1/webhooks.

How the pieces fit

Your system is the master record of your customers and their consents. You tell DM Handled what you hold, the chat uses it, and whatever the chat learns — an address confirmed, a consent given or withdrawn — comes back to you as an event.

your website form ──PUT /api/v1/contacts/{ref}──▶ DM Handled
your web page     ◀─POST …/identity (a pass)──────  (your server asks, the page carries it)
DM Handled chat   ──signed webhook: consent.granted, contact.confirmed, …──▶ your endpoint
your system       ──GET /api/v1/events?after=…──▶ (catch up after downtime)