Search
Getting Started
Feature guides
Integrations
Community support
Model Context Protocol (MCP) connections enable Dreambase to integrate external AI tools and data sources directly into your workspace. When you add an MCP connection, its tools are instantly available to the Dreambase Analyst Agent and enhance report generation capabilities, no manual setup required.
Adding MCP Integrations
Model Context Protocol (MCP) connections let Dreambase integrate external AI tools and data sources directly into your workspace. Once added, those tools are instantly available to the Analyst Agent for every dashboard, Skill, and Slack report, with no manual setup.
Plan availability
Pro subscribers can add any featured Connector from the Connectors catalog in one click, whether it's implemented as an MCP server or a plain API under the hood. Authentication is handled with OAuth, no manual server configuration needed.
Team subscribers get everything Pro gets, plus the MCP tab described below, where you can register a completely custom MCP server, not just the featured catalog. This is the path for internal MCP servers, niche providers, or anything not yet in the catalog.
Adding a featured MCP connector (Pro)
Go to Connectors → MCP (previously "Marketplace").
Click on the desired MCP connector.
Click Add MCP.
Adding a custom MCP connector (Team)
Go to Connectors → MCP (previously "Workspace Settings → MCP").
Click Add MCP.
Fill in:
Name — a clear, descriptive name for the connection.
Description — optional, but useful for future reference.
Transport Type — Streamable HTTP is recommended.
Server URL — the base URL of your MCP server.
Custom Headers — any auth/config headers, in JSON.
Schema Definition — optional but recommended, for stronger type checking.
Client Configuration — any additional options your server needs.
Click Test Connection to verify before saving.
Click Create to finalize and activate.
After creating
All tools exposed by the MCP server register automatically with the Analyst Agent.
New capabilities show up in dashboards and Slack integrations with no additional wiring.
Dreambase handles tool registration and context mapping for you.
Best practices & troubleshooting
Keep connection names clear and descriptive.
Use schema definitions for robust compile-time type checking.
Review and rotate custom headers regularly for security.
If tools aren't showing up, double-check the server URL and headers, and use Test Connection to diagnose connectivity.
Adding MCP Integrations
Model Context Protocol (MCP) connections let Dreambase integrate external AI tools and data sources directly into your workspace. Once added, those tools are instantly available to the Analyst Agent for every dashboard, Skill, and Slack report, with no manual setup.
Plan availability
Pro subscribers can add any featured Connector from the Connectors catalog in one click, whether it's implemented as an MCP server or a plain API under the hood. Authentication is handled with OAuth, no manual server configuration needed.
Team subscribers get everything Pro gets, plus the MCP tab described below, where you can register a completely custom MCP server, not just the featured catalog. This is the path for internal MCP servers, niche providers, or anything not yet in the catalog.
Adding a featured MCP connector (Pro)
Go to Connectors → MCP (previously "Marketplace").
Click on the desired MCP connector.
Click Add MCP.
Adding a custom MCP connector (Team)
Go to Connectors → MCP (previously "Workspace Settings → MCP").
Click Add MCP.
Fill in:
Name — a clear, descriptive name for the connection.
Description — optional, but useful for future reference.
Transport Type — Streamable HTTP is recommended.
Server URL — the base URL of your MCP server.
Custom Headers — any auth/config headers, in JSON.
Schema Definition — optional but recommended, for stronger type checking.
Client Configuration — any additional options your server needs.
Click Test Connection to verify before saving.
Click Create to finalize and activate.
After creating
All tools exposed by the MCP server register automatically with the Analyst Agent.
New capabilities show up in dashboards and Slack integrations with no additional wiring.
Dreambase handles tool registration and context mapping for you.
Best practices & troubleshooting
Keep connection names clear and descriptive.
Use schema definitions for robust compile-time type checking.
Review and rotate custom headers regularly for security.
If tools aren't showing up, double-check the server URL and headers, and use Test Connection to diagnose connectivity.