Snowflake
BetaNatural-language analytics and governed SQL over Snowflake via Cortex Analyst and Cortex Search.
- Snowflake
- Cortex
- Analytics
- Authentication
- API key
- Hosting
- Vendor-hosted endpoint
How it works
Every agent request passes through Ferentin policy before it reaches Snowflake.
Ferentin connects to a Snowflake-managed MCP server object on your own account host using a Programmatic Access Token. Because the endpoint names the account, database, schema, and server, the URL is supplied at install time. Snowflake role-based access control remains the authorization boundary for every call.
- AI agentAsks for something
- Ferentin policies
- Snowflake rbac
- Programmatic access tokens
- Account scoped endpoint
- Snowflake MCP serverAuthorized connection
- Governed resultMasked, logged, returned
What you can do with Snowflake
- Cortex analyst
- Cortex search
- SQL execution
- Object management
Before you connect
- The URL shipped with this card is an EXAMPLE and will not resolve — installing without overriding the endpoint fails to connect.
- Snowflake RBAC is the authorization boundary. Ferentin sends one token per instance, so every caller on that instance acts as that single Snowflake user — use a dedicated service user with a least-privilege role.
- As of 2026-08-20 the server returns tools/call responses as a Server-Sent Events stream rather than a single JSON body; Ferentin's Streamable HTTP client handles both.
- Snowflake also supports OAuth 2.0 (Snowflake OAuth and External OAuth) for this endpoint. Ferentin uses token auth here because the OAuth authorization and token endpoints are per-account and cannot be pinned in a shared catalog card.
1 more technical note
- Network policies on the Snowflake account must permit egress from Ferentin, or the connection is refused before authentication.
About Snowflake
Snowflake is a cloud data platform providing elastic data warehousing, data sharing, and the Cortex suite of AI features — including Cortex Analyst for natural-language querying of semantic views and Cortex Search for retrieval.
Frequently asked about Snowflake
- What can AI agents do with Snowflake through Ferentin?
- Through Ferentin, agents can use Snowflake for cortex analyst, cortex search, SQL execution and object management. Every call is checked against your Ferentin policies first.
- How is Snowflake access secured?
- Requests through the Snowflake integration are covered by snowflake rbac, programmatic access tokens and account scoped endpoint. These controls are applied by Ferentin, not by Snowflake, so they hold across every connected agent.
- How do I connect Snowflake to Ferentin?
- Open the Ferentin admin console, add Snowflake as a new connection and authenticate with API key. You then choose which permissions and which tools your agents may use, and the connection goes live for everyone covered by your policies.
- Is the Snowflake integration generally available?
- The Snowflake integration is currently beta on Ferentin.
Related integrations
3
Databricks
Query the Databricks lakehouse — SQL warehouses, Unity Catalog functions, Genie spaces, and vector search.
Google BigQuery
Query and explore BigQuery data warehouses. Discover datasets/tables, inspect schema, and run read-only SQL — governed by Cloud IAM.
Elasticsearch
Search, ES|QL, and index discovery over your Elasticsearch data via Elastic Agent Builder.