datafrey
Talk to your database. Use to run queries, get schema context, and write SQL faster.
Open source Repository Open in the app JSON README (API)
About
Talk to your database. Use to run queries, get schema context, and write SQL faster.
Details
- Kind
- Plugins
- Topic
- Databases
- Publisher
- datafrey-ai
- Origin
- marketplace
- Category
- ferramentas
- Open pull requests
- 2
- Last push
- 2026-07-11T09:32:17Z
- Repository state
- ativo
- Language
- Python
- License
- Apache-2.0
- Added
- 2026-08-30 01:48:58
- Updated
- 2026-08-30 01:48:58
- Origin id
datafrey-ai/datafrey/datafrey
README
# DataFrey [](https://pypi.org/project/datafrey/) [](https://python.org) [](LICENSE) MCP server for your database. - **Vibecode SQL.** Ask, clean, and transform your data in Claude Code. - **5 min setup.** Connect Snowflake and install the Claude Code plugin using the CLI. - **Understands your data.** Solve complex problems using the planning agent. [Documentation](https://docs.datafrey.ai) · [Website](https://datafrey.ai) · [GitHub](https://github.com/datafrey-ai/datafrey) ## Install ```bash pip install datafrey ``` Or with [uv](https://docs.astral.sh/uv/): ```bash uv tool install datafrey ``` Then run the CLI to log in, connect your database, and install the Claude Code plugin: ```bash datafrey ``` Full walkthrough: [Quickstart](https://docs.datafrey.ai/quickstart). Learn more at [docs.datafrey.ai](https://docs.datafrey.ai). --- ## Commands See the full [CLI reference](https://docs.datafrey.ai/reference/cli) for all commands and options. | Command | Description | | --- | --- | | `datafrey login` | Authenticate with DataFrey | | `datafrey logout` | Remove stored credentials | | `datafrey status` | Show auth, database, and index status | | `datafrey db connect` | Connect a new database (interactive) | | `datafrey db list` | List connected databases | | `datafrey db drop` | Remove the connected database | | `datafrey index` | Sync the database schema index | | `datafrey index drop` | Drop the database schema index | | `datafrey client` | Set up an AI client (interactive) | | `datafrey client claude` | Configure Claude Code | | `datafrey client cursor` | Configure Cursor | | `datafrey client mcp` | Print MCP config block for any MCP-compatible client | | `datafrey doctor` | Check environment and connectivity | --- ## Supported databases - **Snowflake** -- fully supported More databases are coming soon. [Request a database](https://github.com/datafrey-ai/datafrey/issues/new) if yours is not listed. --- ## Packages This repository is a Python monorepo managed with [uv workspaces](https://docs.astral.sh/uv/concepts/workspaces/). The main entry point is the `datafrey` CLI package. | Package | PyPI | Description | | --- | --- | --- | | [`datafrey-cli`](packages/datafrey-cli/) | [](https://pypi.org/project/datafrey/) | CLI for managing databases and MCP setup | | [`datafrey-api`](packages/datafrey-api/) | [](https://pypi.org/project/datafrey-api/) | Shared Pydantic models for the API | | [`datafrey-mcp`](packages/datafrey-mcp/) | [](https://pypi.org/project/datafrey-mcp/) | MCP server bridge | | [`datafrey-mock`](packages/datafrey-mock/) | [](https://pypi.org/project/datafrey-mock/) | Mock API server for local development | --- ## Development Requires Python 3.13+ and [uv](https://docs.astral.sh/uv/). ```bash git clone https://github.com/datafrey-ai/datafrey.git cd datafrey uv sync uv run pytest ``` Format and lint: ```bash uv run ruff format packages/ uv run ruff check packages/ ``` --- ## Contributing Contributions are welcome. Please read [CONTRIBUTING.md](CONTRIBUTING.md) before submitting a pull request. For security vulnerabilities, do not open a public issue. Instead, email [slava+security@datafrey.ai](mailto:slava+security@datafrey.ai). See [SECURITY.md](SECURITY.md) for details. --- ## Telemetry DataFrey sends usage events from the CLI and the MCP server to help us improve the product. Events are linked to your DataFrey account via your WorkOS user ID (`sub`) after login — a pseudonymous identifier, not your email. **CLI** — command name, outcome, duration, error class, provider type (e.g. `snowflake`), auth method, CLI version, OS, architecture, Python version, CI flag. **MCP server** — session start (with the MCP client name, e.g. `claude-ai`, `cursor`), tool name (`plan` or `run`), outcome, duration, and error class. **Never collected (anywhere):** email, name, credentials, hostnames, account/user/role/warehouse/database names, file paths, env vars, error messages, IP addresses, prompt content, SQL, query results, or any tool arguments / responses. **Opt out** by setting either: ```bash export DATAFREY_TELEMETRY_DISABLED=1 export DO_NOT_TRACK=1 ``` The same variables disable telemetry for both the CLI and the MCP server (set them in the MCP server's environment if you self-host). When opted out, no events are sent and no telemetry identifiers are written to your local config. --- ## License Apache 2.0 -- see [LICENSE](LICENSE) for details. Copyright 2025-2026 DataFrey, Inc.