dpaperwork
All guides

Connect Claude with MCP

Beyond the in-app Chat, dpaperwork exposes an MCP (Model Context Protocol) endpoint so you can talk to your Workspace from Claude Desktop, Claude Code, or any other MCP-compatible client — using the same tools Chat itself uses to query records, create and update things, move deals, draft documents, and trigger Playbooks.

Set it up

  1. Go to Settings → Personal → MCP.
  2. Click New token and give it a name you'll recognise later (e.g. "Claude Desktop", "laptop").
  3. Copy the token when it's shown — it's only shown once. The dialog also gives you a ready-made claude mcp add command and a Claude Desktop config snippet, each with its own copy button.
  4. Paste the command into your terminal, or the config snippet into your MCP client's config file, and you're connected.

The endpoint URL is your Workspace's own address (.../api/workspaces/<workspaceId>/mcp) — copy it from the same page if you're wiring up a client by hand instead of using the generated snippets.

What it can do

A connected client gets the same tools as the in-app AI assistant: querying records, creating and updating Companies/Contacts/Deals/Tasks, logging Interactions, moving deal stages, drafting documents and Playbooks, pulling a pipeline snapshot or a metric, and triggering a Playbook — plus a chat tool that runs a full conversational turn for you.

A token acts as you. Every call is scoped to your own role permissions — if your role can't close a Deal in the app, it can't through MCP either. A Workspace that's read-only (payment paused) blocks writes the same way it does everywhere else.

Keeping tokens safe

Treat an MCP token like a password — anyone with it can act as you in this Workspace. Give each device or client its own token so you can tell them apart, and Revoke any token you no longer use or suspect has leaked; revoking is immediate and can't be undone.