Skip to main content

Gemini CLI Integration

📚 Navigation: ← Back to Agent Context Kit | Cursor Integration → | Claude Integration →

Connect Gemini CLI to DataHub so the AI agent can search your data catalog, inspect schemas, and trace lineage — right from the terminal.

Prerequisites​

DataHub Cloud​

Gemini CLI natively supports streamable HTTP with custom headers.

gemini mcp add --transport http \
  --header "Authorization: Bearer <token>" \
  datahub-cloud \
  "https://<tenant>.acryl.io/integrations/ai/mcp/"

Replace <tenant> with your DataHub Cloud tenant name and <token> with your personal access token.

Or add it directly to your settings.json (~/.gemini/settings.json for user-level, or .gemini/settings.json for project-level):

{
  "mcpServers": {
    "datahub-cloud": {
      "httpUrl": "https://<tenant>.acryl.io/integrations/ai/mcp/",
      "headers": {
        "Authorization": "Bearer <token>"
      }
    }
  }
}

Self-Hosted DataHub​

Install uv (curl -LsSf https://astral.sh/uv/install.sh | sh), then:

gemini mcp add \
  -e DATAHUB_GMS_URL="<your-datahub-url>" \
  -e DATAHUB_GMS_TOKEN="<your-datahub-token>" \
  datahub \
  uvx mcp-server-datahub@latest

Verify​

Run gemini mcp list or use the /mcp command inside a session to confirm the DataHub server is connected.

For general troubleshooting, see the MCP server guide.