Skip to main content
Connecting Snowflake lets the agent join what customers said to what your business data records: revenue, plan, usage, churn, product events.

​
What the agent gets

One tool, sql_exec_tool, which runs SQL against the databases and schemas the MCP server you created exposes. That narrow scope is deliberate: you decide at setup time, in Snowflake, exactly what the agent can query. Rippit does not choose.

​
Setup

Snowflake MCP servers are created inside your account, so there is no shared endpoint.
1

Create the MCP server in Snowflake

Use CREATE MCP SERVER to define the server and what it exposes. Its URL follows this shape:
2

Create an OAuth security integration

Snowflake does not let Rippit register itself, so create one confidential OAuth security integration pointing at the redirect URI Rippit shows in the connect dialog. Copy that URI exactly.
3

Connect in Rippit

Paste the MCP server URL and the integration’s client ID and secret. Rippit checks the URL points at a snowflakecomputing.com host.

​
Identity and scope

The connection uses the Snowflake identity of the person who authorized it, so every query runs under that user’s role and inherits their grants, masking policies, and row access policies. An organization-wide connection shares that identity with everyone in your Rippit organization. The clean approach is to authorize with a purpose-built Snowflake user whose role grants exactly the reporting access the agent should have, rather than an analyst’s personal account.

​
Tools