ai.backengine/backengine-mcp

Official

by BackEngine-ai

Surface customer & prospect context from Slack, email, transcripts and tickets in any MCP client.

Surfaces customer context from Slack, email, transcripts, and tickets.

<p> <img src="assets/logo.png" alt="BackEngine" width="128" height="128"> </p> <h1>BackEngine MCP Server</h1> <p> The customer-context layer for your revenue team — available to Claude and any MCP client. </p>

What is BackEngine?

BackEngine is a multi-tenant SaaS platform that ingests all of an organization's customer and prospect communications — Slack threads, emails, call and meeting transcripts, and support tickets — and distills them into structured, queryable context. Raw conversations are processed into signals (categorized, attributed moments), sources (the underlying transcripts, emails, and tickets), and rolling project overviews, all isolated per tenant and scoped by role and access controls. The MCP server exposes this layer over the Model Context Protocol, so an LLM client can query customer signals, reconstruct context, prep for meetings, surface at-risk accounts, and draft grounded outreach from real conversation history — without leaving the chat.

Connecting

The server is remote and speaks streamable HTTP at https://backengine.ai/mcp. You'll need a BackEngine account; the server authenticates your session on connect.

Clients with native remote MCP support

json

{
  "mcpServers": {
    "backengine": {
      "type": "streamable-http",
      "url": "https://backengine.ai/mcp"
    }
  }
}

Clients that require a stdio bridge

For clients that don't yet support remote servers directly, proxy through mcp-remote:

json

{
  "mcpServers": {
    "backengine": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://backengine.ai/mcp"]
    }
  }
}

Tools

The core tools for querying customer and prospect context are listed below. The exact tool set is resolved per connection — additional tools (creating and updating records, and integration-specific actions for Jira, Zendesk, HubSpot, Salesforce, and Slack) become available based on your permissions and which integrations your workspace has connected.

Tool	Description
`list_projects`	List customers and prospects; filter by name, type, etc.
`get_project`	Retrieve a customer or prospect, optionally with its overview.
`get_project_overview`	Pre-computed rolling 12-week summary for a customer or prospect.
`list_signals`	List signals (categorized moments); filter by project, role, and time.
`find_similar_signals`	Semantic search across signals from a natural-language query.
`list_sources`	List sources — transcripts, emails, Slack threads, and tickets.
`get_source`	Retrieve a source's content (summary, verbatim, or metadata only).
`list_roles`	List the perspective lenses (e.g. product, sales, CS) used to filter signals.
`find_contacts`	Find contacts by name, email, job title, or project.
`list_signal_speakers`	List who has been speaking across signals.

Install

claude_desktop_config.json

{
  "mcpServers": {
    "backengine-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://backengine.ai/mcp"
      ]
    }
  }
}

Desktop config is stdio-only; this bridges via mcp-remote. Native remote: Settings > Connectors.