Skip to main content
Glama
AutoRepair Cloud MCP server

AutoRepair Cloud MCP server

AutoRepair Cloud MCP server for ChatGPT and Claude

Cotlet.dev connects AutoRepair Cloud to ChatGPT, Claude and other AI agents through your own signed-in Cloud Browser. The AutoRepair Cloud integration page reports no public API documentation, official MCP server or official ChatGPT or Claude connector, checked 17 November 2026.

What can I ask my agent to do?

  • Turn a customer concern into a repair quote for your approval.

  • Find repairs waiting for parts and review workflow delays.

  • Review workshop bookings, technician hours and daily sales.

  • Find customer vehicles and their repair history.

Quotes, orders, payments and anything a customer sees wait for your approval.

Related MCP server: Actionbook

How do I connect AutoRepair Cloud to ChatGPT or Claude?

  1. Sign up for Cotlet.dev.

  2. Enable AutoRepair Cloud and sign in at autorepaircloud.com/login in your Cotlet.dev Cloud Browser.

  3. In ChatGPT, open Settings > Apps & Connectors > Create. In Claude, open Settings > Connectors > Add custom connector.

  4. Add https://cotlet.dev/mcp/v1/connect and sign in to Cotlet.dev with OAuth.

Refresh the connector after enabling AutoRepair Cloud.

Can I connect an editor or a script?

Use https://cotlet.dev/mcp/v1 with an API key from your Cotlet.dev console.

{"mcpServers":{"cotlet":{"url":"https://cotlet.dev/mcp/v1","headers":{"Authorization":"Bearer ${env:COTLET_API_KEY}"}}}}

Can I run a local MCP bridge?

The included Python bridge forwards browser operations to the hosted Cotlet.dev MCP server. It requires Python 3.10 or later and uses the standard library. Its tools cover guidance, connection status, opening and reading a page, clicking, typing and closing the browser. The hosted connector above is the recommended setup.

export COTLET_API_KEY="your-key"
python3 autorepair_cloud_agent.py serve

Run python3 autorepair_cloud_agent.py connect for configuration help. Without a key, the bridge lists its tools and explains how to obtain a key when a tool is called.

Is this official?

Cotlet.dev is independent and is not affiliated with, endorsed by or sponsored by AutoRepair Cloud. AutoRepair Cloud is a trademark of its owner.

AutoRepair Cloud integration details and tool reference.

Available Tools

7 tools
autorepair_cloud_clickC

Click a numbered element on the current AutoRepair Cloud page.

Inputs: element (required).

autorepair_cloud_closeA

Close the AutoRepair Cloud browser session when the work is done, so credits stop running.

Inputs: no named parameters.

autorepair_cloud_guideA

How to work in AutoRepair Cloud (repair quotes, workshop bookings, parts and shop reports): its screens and the sign-in link. Call this first.

Inputs: task.

autorepair_cloud_openB

Open a AutoRepair Cloud page in your signed-in Cloud Browser (default https://autorepaircloud.com/login) to work with repair quotes, workshop bookings, parts and shop reports.

Inputs: url.

autorepair_cloud_readA

Read the current AutoRepair Cloud page: its text and numbered buttons, links and fields.

Inputs: element.

autorepair_cloud_statusA

Check that your Cotlet.dev account and AutoRepair Cloud are ready: account, credits and AutoRepair Cloud connection.

Inputs: no named parameters.

autorepair_cloud_typeC

Type text into a numbered field on the current AutoRepair Cloud page.

Inputs: text (required), element (required).

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.7/5.0

    Scored across 7 tools

    Disambiguation5/5

    Each tool has a clearly distinct role: guide, status, open, read, click, type, and close map to separate phases or browser actions. There is no meaningful overlap between reading, clicking, typing, opening, and closing.

    Naming Consistency5/5

    All tools consistently use the autorepair_cloud_ prefix followed by a short action-oriented name. The snake_case pattern is predictable throughout.

    Tool Count5/5

    Seven tools is well-scoped for a browser-based workflow server. Each tool covers an essential step: orientation, readiness, session lifecycle, and page interaction.

    Completeness4/5

    The surface covers the core browser workflow: open, read, click, type, close, plus guidance and status checks. Minor gaps exist around explicit navigation, waiting, scrolling, and select/dropdown handling, but most tasks can likely be worked around with read/click/type.

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      Not graded
      quality
      B
      maintenance
      Connects an AI assistant to a dealer account so it can build and link a shop website that shows live vehicle stock and takes enquiries and test-drive bookings with no database to run. It also lets the assistant read and change listings, prices, test drives, reviews and leads in plain language, with read-only access by default and every change written to an audit log.
      35 npm
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Connects MCP clients to an AI browser so agents can open and read pages, click, type, fill forms, take screenshots, and search within the user's logged-in browser session. Sensitive actions such as checkouts or payments require explicit approval.
      MIT
    • A
      license
      A
      quality
      B
      maintenance
      Connects a signed-in Phorest account to AI agents, enabling natural-language appointment booking and rebooking, schedule and gap checks, client history queries, and client-card notes without a Phorest API key.
      7
      Elastic 2.0