MCP Server

Query metrics from any AI assistant

Metricly speaks MCP — the open standard for AI tool integration backed by Anthropic, OpenAI, Microsoft, and Google. Connect your metrics to virtually any AI assistant.

Works with your favorite AI tools

MCP is the industry standard. Connect Metricly to any compatible AI assistant.

ChatGPT

OpenAI desktop app with MCP support

Claude

Claude.ai and Claude Code

Cursor

AI-first code editor

Windsurf

Codeium AI editor

VS Code

With Copilot or Continue

Microsoft Copilot

Copilot Studio integration

And any other MCP-compatible client — the ecosystem is growing daily.

Natural language, real answers

Ask questions in plain English. Get instant answers from your semantic layer. No SQL required.

  • Queries powered by your dbt semantic layer
  • Consistent metrics across all your AI tools
  • Full context for follow-up questions
Show me revenue by region for Q4

Here's Q4 revenue by region:

North America: $2.4M (+12%)
Europe: $1.8M (+8%)
APAC: $890K (+24%)
Add this to my executive dashboard

How teams use MCP

Ask questions while coding

Query revenue, churn, or any metric without leaving your IDE. Get data context for the code you're writing.

"What's our MRR by plan tier for the last 6 months?"

Build dashboards with AI

Describe the dashboard you need. Let AI create widgets, configure visualizations, and arrange layouts.

"Create an executive dashboard with revenue, churn, and customer count KPIs"

Automate reporting workflows

Coding agents can query metrics, generate reports, and update dashboards as part of automated workflows.

"Update the weekly metrics dashboard with this week's data"

25 tools at your fingertips

Query metrics, manage dashboards, and edit your semantic layer — all through conversation.

Query Tools

  • list_metrics

    List available metrics with descriptions

  • list_dimensions

    List groupable dimensions

  • query_metrics

    Execute metric queries with filters

  • explain_metric

    Show metric calculation details

Dashboard Tools

  • list_dashboards

    List personal and team dashboards

  • get_dashboard

    Get full dashboard configuration

  • create_dashboard

    Create new dashboard

  • add_widget

    Add widget to dashboard

Semantic Layer Tools

  • get_manifest_status

    Show semantic layer status

  • list_semantic_models

    List semantic models

  • create_metric

    Create new metric definition

  • preview_metric

    Test metric before saving

View all 25 tools in the documentation

Get started in minutes

1

Authenticate with Metricly

$ metricly login
2

Add MCP server to your AI tool

Configure your AI assistant to use the Metricly MCP server. Each tool has specific setup instructions.

3

Start asking questions

Your AI assistant now has access to your metrics. Ask away.

Frequently asked questions

What is MCP (Model Context Protocol)?

MCP is an open standard for connecting AI assistants to external tools and data sources. Created by Anthropic and now backed by OpenAI, Microsoft, and Google, it allows AI assistants to securely access your metrics and dashboards.

Which AI tools support MCP?

MCP is supported by ChatGPT (desktop), Claude (Claude.ai and Claude Code), Cursor, Windsurf, VS Code with Copilot, and many other AI tools. The ecosystem is growing rapidly.

Is my data secure with MCP?

Yes. Metricly's MCP server uses OAuth authentication tied to your account. Your data never leaves your warehouse - we only execute queries when you request them through your AI assistant.

How do I set up the MCP server?

After authenticating with `metricly login`, add the MCP server configuration to your AI tool's settings. Each tool has slightly different configuration - see our documentation for specific guides.

Ready to connect your AI assistant?

Start querying your metrics from ChatGPT, Claude, Cursor, or any MCP client.