Integrations

Last updated: 6 days ago
← Back to Overview
oper8r Documentation

Integrations extend oper8r's capabilities by connecting external services, managing API access, and organizing reusable prompts.

When logged in, click Data in the navigation bar to access these settings.

MCP Integration

Connect your AI tools directly to oper8r using the Model Context Protocol (MCP). Once connected, you can run semantic search on your knowledge base and retrieve documents from within your preferred AI assistant.

Prerequisites

  • An active oper8r account with an organization
  • Content indexed in your knowledge library

Server URL

Use the following URL when configuring your MCP client:

https://oper8r.io/api/mcp

Available Tools

Once connected, your AI assistant will have access to these tools:

oper8r_search_knowledge

Search your organization's knowledge base using hybrid semantic + keyword search.

Parameters:

  • query (required) - your search query
  • document_types (optional) - filter by document type
  • source_types (optional) - filter by source type
  • limit (optional) - max results, 1–50, default 10

Returns relevant passages with document IDs, titles, URLs, and content snippets.

oper8r_get_documents

Fetch the complete original content of documents by their IDs (from search results).

Parameters:

  • document_ids (required) - array of document IDs from search results (max 10)

Returns the full document text with metadata.

Cursor

  1. Open Cursor Settings and navigate to the MCP section, or manually create/edit the config file at:
    • Project-scoped: .cursor/mcp.json in your project root
    • Global: ~/.cursor/mcp.json in your home directory
  2. Add the following configuration:
{
  "mcpServers": {
    "oper8r": {
      "url": "https://oper8r.io/api/mcp"
    }
  }
}
  1. Save the file - Cursor will detect that authentication is required and open your browser
  2. Sign in with your oper8r account and select Allow
  3. Open an Agent Mode chat and ask it to search your knowledge base

Requires Cursor 1.0 or later. If you don't see MCP options, update Cursor to the latest version.

VS Code

Requires VS Code 1.101+ with the GitHub Copilot extension installed.

Option A: Command Palette

  1. Open the Command Palette (Cmd+Shift+P / Ctrl+Shift+P)
  2. Run "MCP: Add Server"
  3. Select "HTTP (Streamable HTTP or Server-Sent Events)"
  4. Enter the server URL: https://oper8r.io/api/mcp
  5. Name it oper8r
  6. Choose where to save (workspace or user profile)
  7. A browser window will open - sign in with your oper8r account and select Allow

Option B: Manual Configuration

  1. Create or open .vscode/mcp.json in your project root
  2. Add the following:
{
  "servers": {
    "oper8r": {
      "type": "http",
      "url": "https://oper8r.io/api/mcp"
    }
  }
}
  1. Save the file - VS Code will detect that authentication is required and open your browser
  2. Sign in with your oper8r account and select Allow
  3. The server should show a Running status in the MCP panel

Claude Code

  1. Run the following command in your terminal:
claude mcp add --transport http oper8r https://oper8r.io/api/mcp
  1. Launch Claude Code with claude
  2. Run the /mcp command, select the oper8r server, and choose Authenticate
  3. A browser window will open - sign in with your oper8r account and select Allow

Use --scope user to make oper8r available across all projects, or --scope project to share the config with your team via .mcp.json.

Claude Desktop

  1. Open Claude Desktop and go to Settings (profile icon → Settings)
  2. Navigate to Connectors in the sidebar
  3. Scroll to the bottom and click "Add custom connector"
  4. Enter the server URL: https://oper8r.io/api/mcp
  5. Click Add and complete the OAuth authentication flow in the browser that opens

The Connectors feature is available on all Claude Desktop plans.

Windsurf

  1. Open the MCP config file at ~/.codeium/windsurf/mcp_config.json (you can access this via Settings → Cascade → MCP Servers, then click the gear icon)
  2. Add the following configuration:
{
  "mcpServers": {
    "oper8r": {
      "serverUrl": "https://oper8r.io/api/mcp"
    }
  }
}
  1. Save the file - Windsurf will detect that authentication is required and open your browser
  2. Sign in with your oper8r account and select Allow

If the OAuth flow does not trigger automatically, you can use the mcp-remote bridge instead:

{
  "mcpServers": {
    "oper8r": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://oper8r.io/api/mcp"]
    }
  }
}

This requires Node.js to be installed. After saving, mcp-remote will open your browser for authentication.

API Keys

Create API keys for programmatic access to oper8r's capabilities.

Managing API Keys

  • Create - generate a new API key with a read scope and an optional label for identification
  • Labels - add descriptive labels to keep track of what each key is used for
  • Rotate - cycle/rotate an existing key to generate a new one while invalidating the old key
  • Usage tracking - see when each key was last used

API Keys manager showing a list of keys with labels, scopes, and last used timestamps

Saved Prompts

Create and save reusable prompts for use across oper8r sessions. Saved prompts are especially useful for standardizing instructions across your team - ensuring consistent tone, formatting, and requirements across all RFP responses.

Saved Prompt management interface

Creating Prompts

Each prompt includes:

  • Title - a descriptive name for quick identification
  • Content - the full prompt text
  • Visibility - choose between:
    • Public - visible to everyone in your organization (only you can edit)
    • Private - visible only to you

Using Prompts

Saved prompts can be applied in Chats and Documents via the Prompts selector. Click the Prompts button in the content generation settings to browse and apply your saved prompts.

Using Saves Prompts interface

External Integrations

Connect third-party services to enhance oper8r's context and capabilities. All integrations use secure OAuth-based authentication - oper8r never stores your third-party passwords.

Important

Please reach out to oper8r to get integrations turned on in your enterprise account. External integrations are not enabled by default, and are only available in enterprise accounts.

Google Calendar

Query your calendar events in real time during oper8r sessions. Google Calendar is queried live - no data is stored in oper8r.

How it works: Runtime querying - oper8r reads your calendar on the fly when you ask about upcoming meetings, scheduling conflicts, or attendee details. No data is synced or persisted.

Scope: Individual - each user connects their own Google account. Any member can connect.

Permissions required:

  • View calendar properties
  • Read calendar events

Key capabilities:

  • Live event lookup - ask about your schedule, upcoming meetings, or attendee lists during any session
  • Scheduling context - pull meeting details into chats and documents for richer proposals and follow-ups

Google Drive

Sync documents, spreadsheets, and files from Google Drive directly into oper8r's knowledge base. Google Drive performs an initial selective sync, then stays current via webhooks.

How it works: Initial selective sync - you choose which files and folders to import into oper8r's document database. After the initial import, Google Drive webhooks automatically push updates so your knowledge base stays aligned with the latest versions.

Scope: Individual - each user connects their own Google account. Any member can connect.

Permissions required:

  • Read and manage files in Google Drive
  • View file metadata

Key capabilities:

  • Selective sync - choose specific files, folders, or shared drives to import
  • Automatic updates - webhook-driven updates keep your knowledge base current without manual re-uploading
  • AI processing - imported files are processed through oper8r's AI pipeline for indexing and retrieval

HubSpot

Connect your HubSpot CRM to bring customer and deal context directly into oper8r. HubSpot data is queried live - no CRM records are stored in oper8r.

How it works: Runtime querying - oper8r reads your HubSpot data on the fly when you ask about contacts, companies, deals, or owners. No CRM data is synced or persisted.

Scope: Workspace - the connection is shared across your organization. Only admins can connect.

Permissions required:

  • Read contacts and companies
  • Read deals and pipelines
  • Access owner info

Key capabilities:

  • Read CRM records - access contacts, companies, deals, and owner data from HubSpot
  • Ask questions - query your HubSpot data through oper8r's chat and document sessions to enrich proposals with real customer context
  • Enrich RFP responses - pull in account details, deal stages, and contact information to personalize responses
  • Connection stats - view your total contacts, companies, deals, and owners at a glance

Fathom

Import call transcripts and meeting summaries from Fathom into oper8r's knowledge base for analysis and follow-up generation.

How it works: Initial sync backfill - when connected, oper8r imports your existing Fathom call history into its document database. After the initial backfill, each new call automatically receives an update to add the transcript once it becomes available.

Scope: Workspace - the connection is shared across your organization. Only admins can connect.

Permissions required:

  • Access call recordings
  • Read meeting summaries
  • View participant info

Key capabilities:

  • Automatic transcript import - call transcripts are synced into your knowledge base as they become available
  • Meeting summaries - Fathom's AI-generated summaries are imported alongside full transcripts
  • Streamlined workflow - go from recorded call to CRM update and customer follow-up email in minutes
  • Full context - oper8r analyzes the complete transcript to generate accurate, contextual outputs

Slack

Connect oper8r to your Slack workspace and bring AI-powered assistance directly into your team's communication channels. Interact with oper8r via direct messages or by mentioning it in channels.

How it works: Once installed, oper8r listens for direct messages and @mentions in channels it has been added to. Responses are grounded in your organization's knowledge base.

Scope: Workspace - the connection is shared across your organization. Only admins can connect.

Permissions required:

  • Read and send messages
  • Access user and team info
  • Read direct messages

Installation:

  1. Navigate to Data → External Integrations in oper8r
  2. Click Connect on the Slack integration
  3. Authorize oper8r in the Slack OAuth flow - this installs the oper8r bot into your workspace
  4. Invite the oper8r bot to any channels where you want it available (e.g. /invite @oper8r)

Using oper8r in Slack:

  • Direct messages - send a DM to the oper8r bot to ask questions privately, just like using oper8r's chat interface
  • Channel mentions - @mention oper8r in any channel it has been added to and it will respond in-thread with knowledge-grounded answers
  • Knowledge-grounded responses - all replies draw from your organization's indexed knowledge base

Key capabilities:

  • AI-powered Q&A - team members can ask questions and get responses grounded in your knowledge base without leaving Slack
  • Channel integration - add oper8r to channels where your team discusses RFPs, customer questions, or security questionnaires
  • DM support - interact with oper8r one-on-one for private queries and draft generation

Connection Management

Each integration is scoped based on its type:

  • Workspace integrations (HubSpot, Fathom, Slack) - connected by an admin and available to all members of the organization
  • Individual integrations (Google Calendar, Google Drive) - connected by each user for their own account

You can disconnect and reconnect integrations at any time. OAuth token refresh is handled automatically - once connected, your integration stays active without manual intervention.

← Back to Documentation