DataFrey
MCP server for Snowflake with text-to-SQL agent
Last verified:
What is DataFrey?
DataFrey is an MCP server for your database that lets you ask questions, generate SQL, and work with data directly inside Claude Code. It is positioned as a “/db” skill for vibecoding SQL, so users can ask, clean, transform, and document data without leaving their coding flow.
The setup is designed to be fast: the docs say you can connect Snowflake and install the Claude Code plugin using the CLI in about 5 minutes. The website shows a simple install path with pip install datafrey && datafrey, followed by use in Claude Code with commands like /db top 10 customers by revenue.
A key feature is that DataFrey understands your data and uses a planning agent to help solve more complex problems. The docs also emphasize that it can help you write SQL faster, which suggests it is meant to reduce the friction of querying and shaping data for analytics work.
The main audience appears to be data practitioners and developers who already use Claude Code and want a faster, database-native workflow. It is especially relevant for Snowflake users who want to interact with their warehouse in plain language and keep credentials protected, since the site says connections are encrypted at rest and in transit.
DataFrey pricing
Pricing model: Freemium
The website does not list public pricing, free tier details, or paid plan names. The docs and homepage focus on setup and usage, showing installation via pip install datafrey and use in Claude Code, but they do not describe plan tiers, included features by tier, or subscription costs.
DataFrey pros
- Works inside Claude Code
- Uses a simple /db skill
- Fast 5-minute setup
- CLI-based installation
- Supports Snowflake connection
- Helps write SQL faster
- Can ask the database in natural language
- Supports cleaning data
- Supports transforming data
- Supports documenting data
- Uses a planning agent
- Handles complex problems
- Fits a vibecoding workflow
- Encrypted at rest
- Encrypted in transit
DataFrey cons
- Focused on Claude Code
- Requires CLI setup
- Appears Snowflake-first
- Not presented as a standalone UI
- Best suited to database users
- May require existing database access
- Docs emphasize workflow, not broad integrations
- Limited website detail on pricing
Frequently asked questions about DataFrey
What is DataFrey?
DataFrey is an MCP server for your database that lets you ask questions and work with data from Claude Code. The docs describe it as a way to vibecode SQL, meaning you can ask, clean, transform, and document data in a more natural workflow.
What problem does DataFrey solve?
DataFrey is meant to reduce the friction of querying and manipulating warehouse data. Instead of writing every SQL query manually, you can use the /db skill in Claude Code to ask the database directly and speed up analysis and transformation work.
How do you install DataFrey?
The docs show a CLI setup path using pip install datafrey && datafrey. The website also says setup is about 5 minutes and includes connecting Snowflake and installing the Claude Code plugin through the CLI.
Where do you use DataFrey?
DataFrey is used inside Claude Code. The website shows an example command like /db top 10 customers by revenue, which indicates the interaction happens through Claude Code rather than a separate web app.
Which database does DataFrey support?
The docs explicitly mention connecting Snowflake. The homepage also frames the product as a smart MCP server for your database, but Snowflake is the only database called out clearly on the website.
What can DataFrey do besides querying?
The docs say you can ask, clean, and transform your data, and they also mention documenting data. That suggests it is meant to support several parts of an analytics workflow, not just SQL generation.
Does DataFrey help with complex tasks?
Yes. The site says it understands your data and can solve complex problems using a planning agent. That implies it can break down harder data tasks into steps rather than only handling simple one-off queries.
Is DataFrey secure?
The homepage says your credentials and connections are encrypted at rest and in transit. That is the main security detail publicly emphasized on the website.
How long does setup take?
The docs and homepage both describe setup as taking about 5 minutes. The quick install flow is presented as a fast way to connect Snowflake and start using the Claude Code plugin.
Who is DataFrey for?
DataFrey is aimed at developers and data practitioners who already use Claude Code and want a faster way to work with databases. The positioning around /db, SQL, and Snowflake makes it most relevant for technical users working directly with warehouse data.