MCP.so
Sign In

ContextfloVerifiedFeatured

@Vivek Sah

About Contextflo

Contextflo is a governed context layer that lets your whole team ask questions of your real data through the AI they already use. Connect BigQuery, Snowflake, Postgres, Redshift, Databricks, ClickHouse, or upload a CSV, then query it in Claude or ChatGPT over MCP. Table-level access control is enforced at query time and every query is audited per user, so non-technical teammates get self-serve answers without seeing data they should not.

Connection details

https://mcp.contextflo.com/mcp

Setup

claude mcp add contextflo --transport http https://mcp.contextflo.com/mcp

Tools

No tools detected

Fetch the live tool list directly from this server's endpoint using the button above.

Overview

What is Contextflo?

Contextflo is a governed context layer that connects Claude and ChatGPT to your data warehouse. Your whole team can ask questions of real company data in the AI they already use, no SQL required, and without exposing data they should not see.

What it does

  • Connects to BigQuery, Snowflake, Postgres, Redshift, Databricks, ClickHouse, and MySQL, or you upload a CSV.
  • Exposes your data over MCP, so Claude, ChatGPT, or any MCP client can query it.
  • Enforces table-level access control at query time (Users, Groups, Rules).
  • Logs every query per user for a complete audit trail.
  • Answers using context built from your schema, saved queries, metric definitions, and docs, so numbers match what your team already agreed on.
  • Includes dashboards and scheduled reports.

Built for teams

Roll it out to everyone who needs answers, sales, ops, finance, support, and keep the controls a company needs. Each person only reaches the tables you allow, you see what the whole team is asking about, and every query is logged to a per-user audit trail. Self-serve for them, visibility and control for you.

How it works

  1. Connect a data source, or upload a CSV. Setup takes about ten minutes.
  2. Set access: table-level rules per user and group.
  3. Your team asks questions from Claude or ChatGPT over MCP.
  4. Contextflo answers with your context and logs the query for audit.

Why it is different

A raw database MCP hands the model your whole database with no guardrails. Contextflo adds the layer a company actually needs:

  • Query-time, table-level access control, not all-or-nothing.
  • Per-user audit of every query.
  • Context so answers use agreed definitions instead of guessing.
  • Works across your whole stack, plus CSV uploads.

Who it is for

Teams without a dedicated data team who want self-serve answers, and data leads who need governance and audit before letting an LLM near the warehouse.

Pricing

Free for one user and one source. Paid plans from $75/user/mo. See https://contextflo.com/pricing

Frequently asked questions

What is the Contextflo remote MCP server?

The Contextflo remote MCP server is a hosted Model Context Protocol endpoint at https://mcp.contextflo.com/mcp, so AI assistants can connect to it without installing or running anything locally.

How do I connect to the Contextflo MCP server?

Add the endpoint https://mcp.contextflo.com/mcp to any MCP-compatible client such as Claude Code, Cursor, or VS Code. The setup snippets on this page configure each client in one step.

Does the Contextflo MCP server require authentication?

Yes. Contextflo uses OAuth: the first time you connect, your MCP client opens a browser window to sign in and authorize access, then reuses the credentials for future sessions.

Which transport does the Contextflo MCP server use?

Contextflo exposes a Streamable HTTP endpoint, the transport used by remote MCP servers and supported by all major MCP clients.

Comments