Claude integration: ask Roadmark from Claude

Claude connects to outside systems through custom connectors, which Anthropic builds on the Model Context Protocol (MCP). Roadmark's connector is an MCP server, so adding Roadmark to Claude is a matter of pasting its address into Claude's connector settings and signing in. There's nothing to install and nothing for your developers to build.

What your team can ask Claude

Anything they could find in Roadmark, asked in their own words. A dispatcher asks which loads picking up tomorrow still have no truck. Someone in sales asks why the last five Northgate Home quotes were lost. Finance asks what's over 45 days unpaid, by customer. An owner compares margin across the group's three companies this quarter. See the AI assistants page for more questions by team.

For example, with Roadmark's example group, an owner asks Claude how Northgate Home did this quarter compared with last. Claude reads the loads and invoices in Roadmark and answers, in US dollars: 118 loads, up from 94, $271,480.00 in revenue at a 15.9% margin, down from 17.4%. Most of the drop is Hamilton to Columbus, where carrier cost rose $190.00 a load since July and the rate stayed the same. The owner asks for a renewal with Columbus up 6% and everything else flat. Claude drafts the rate agreement in Roadmark and says so in the chat: the draft is waiting for approval, and nothing has been sent to Northgate Home.

Two sets of permissions, both enforced

Roadmark's. The connector runs as the person asking. It carries exactly their access: their companies, their customers, nothing else. A Roadmark admin decides how far assistants may go, from read only, to read and draft, to acting within the rules your team has set for the agents. Work asked for in Claude goes through the same agents as work asked for in the app, so a dollar limit on a lane, a customer on credit hold or a carrier blocked by compliance applies to it the same way. See AI controls.

Claude's. Anthropic lets Team and Enterprise Owners decide, per connector, whether read-only tools and write tools are always allowed, need approval or are blocked, as described in Use connectors to extend Claude's capabilities. Anthropic also notes that even when Claude is allowed a write action, the person still needs the permission in the source system to make the change. In Roadmark's case, that's the access described above.

Every action taken through Claude is on the record in Roadmark, with the person, the assistant and the time. A quote asked for in Claude shows up in the quote's history as "asked Claude to price", then the Quoting agent's draft, then the person who approved and sent it.

Setting it up on a Team or Enterprise plan

Anthropic's setup guide splits the work in two. An Owner adds the connector once for the organization, from Organization settings > Connectors. After that, each member finds Roadmark under Customize > Connectors, clicks Connect and signs in with their own Roadmark account. That sign-in is what gives Claude exactly that person's access, so two people asking the same question can get different answers if they work in different companies.

On an individual Pro or Max plan, you add the connector yourself from Customize > Connectors. A Free account can add one custom connector.

Anthropic's guide also warns to connect only to servers you trust and to review the permissions a server asks for. Claude reaches connectors from Anthropic's cloud rather than from your computer.

In a chat

Connectors are turned on per conversation: click + in the chat box, open Connectors and switch Roadmark on. From there, ask in plain words. When a question needs a record, Claude calls Roadmark and shows that it did, such as "Read loads and invoices in Roadmark". When a request needs work done, the result lands in Roadmark as a draft, with the reasons, in the same queue your team approves everything else from.

The same connector works in ChatGPT and Gemini too, and the Developers page describes the MCP server behind it for teams building their own tools.