Connect Snowflake with Unblocked to give agents live warehouse context

Unblocked builds and keeps current a schema wiki for the Snowflake databases you connect — tables, columns, types, and relationships — so an agent can reference real names instead of guessing at your data model.

Agents can also run a single read-only SQL query (SELECT, WITH, SHOW, or DESCRIBE) against your connected databases to check something directly, with results row-capped and reconciled with your code, pull requests, and docs.

Once connected, any MCP-compatible coding agent — Claude Code, Cursor, Windsurf, GitHub Copilot, Codex — can pull this context through a single Unblocked MCP server connection.

What you get

• Auto-generated schema wiki — Unblocked builds and refreshes a wiki of the databases, schemas, tables, and columns you connect, so agents reference real names instead of guessing.
• Live, read-only SQL queries — Agents can run a single read-only query against your connected databases to check something directly, capped and reconciled with your code.
• Database-level scoping — Choose which Snowflake databases Unblocked can see, and update the selection anytime from Settings → Data Sources → Snowflake.
• Cross-source synthesis — Warehouse schema and query results are combined with your code, pull requests, and docs into a single knowledge graph.

Need screenshots or troubleshooting?This is a concise connection summary. Follow the complete setup guide ↗

Prerequisites

• An Unblocked account (free to start).
• A Snowflake account with an OAuth security integration registered for Unblocked, and the account identifier, client ID, and secret it generates.
• A Snowflake user with read access to the databases you want to connect — queries always run read-only.
• An MCP-compatible coding agent (Claude Code, Cursor, Windsurf, etc.) to use the context in your editor.

Step 1: Connect Snowflake to Unblocked

1. In Unblocked, go to Settings → Data Sources → Connect another data source, and under Data warehouses select Snowflake.
2. In Snowflake, register an OAuth security integration for Unblocked and note the client ID and secret it generates.
3. In Unblocked, enter your Snowflake account identifier along with the client ID and secret, then click Connect.
4. Sign in to Snowflake and authorize Unblocked, then select which databases to make available and click Save Settings.

Step 2: Install the Unblocked MCP server

On macOS or Linux, one command auto-detects and configures your supported agents (Claude Code, Cursor, Windsurf, GitHub Copilot, Codex, OpenCode, VS Code):

On Windows, or when you prefer remote MCP, add the server manually using the configuration below, or follow the manual install guide ↗.

claude mcp add -s user --transport http unblocked https://getunblocked.com/api/mcpsse

{
  "mcpServers": {
    "unblocked": {
      "url": "https://getunblocked.com/api/mcpsse"
    }
  }
}

https://getunblocked.com/api/mcpsse

To give Claude Code, Cursor, GitHub Copilot, Codex, or another MCP-compatible AI coding agent access to Snowflake context, connect Snowflake to Unblocked and add the single Unblocked MCP server to your agent.

Unblocked acts as the MCP context layer for your Snowflake data, searching schema wiki, read-only SQL queries, table relationships alongside your other connected sources and returning cited results through these tools:

• context_researchMulti-source research across your Snowflake schema wiki, plus your code, pull requests, and other connected sources. Returns a synthesized answer citing the tables it used.
• query_databasesRuns a single read-only SQL query against the Snowflake databases you’ve connected — using the schema wiki to reference real table and column names — and returns the rows.

Questions you (or your agent) can ask Unblocked about your Snowflake data:

• What does the customers table look like, and how does it relate to orders?
• Which columns in the events table track feature usage?
• How many rows were added to the billing table in the last 24 hours?
• What’s the schema for our analytics database?