{
  "markdown": "# DataFrey\n\n[![PyPI version](https://img.shields.io/pypi/v/datafrey?style=flat-square)](https://pypi.org/project/datafrey/)\n[![Python 3.13+](https://img.shields.io/badge/python-3.13%2B-blue?style=flat-square)](https://python.org)\n[![License: Apache 2.0](https://img.shields.io/badge/license-Apache%202.0-blue?style=flat-square)](LICENSE)\n\nMCP server for your database.\n\n- **Vibecode SQL.** Ask, clean, and transform your data in Claude Code.\n- **5 min setup.** Connect Snowflake and install the Claude Code plugin using the CLI.\n- **Understands your data.** Solve complex problems using the planning agent.\n\n[Documentation](https://docs.datafrey.ai) · [Website](https://datafrey.ai) · [GitHub](https://github.com/datafrey-ai/datafrey)\n\n## Install\n\n```bash\npip install datafrey\n```\n\nOr with [uv](https://docs.astral.sh/uv/):\n\n```bash\nuv tool install datafrey\n```\n\nThen run the CLI to log in, connect your database, and install the Claude Code plugin:\n\n```bash\ndatafrey\n```\n\nFull walkthrough: [Quickstart](https://docs.datafrey.ai/quickstart). Learn more at [docs.datafrey.ai](https://docs.datafrey.ai).\n\n---\n\n## Commands\n\nSee the full [CLI reference](https://docs.datafrey.ai/reference/cli) for all commands and options.\n\n| Command | Description |\n| --- | --- |\n| `datafrey login` | Authenticate with DataFrey |\n| `datafrey logout` | Remove stored credentials |\n| `datafrey status` | Show auth, database, and index status |\n| `datafrey db connect` | Connect a new database (interactive) |\n| `datafrey db list` | List connected databases |\n| `datafrey db drop` | Remove the connected database |\n| `datafrey index` | Sync the database schema index |\n| `datafrey index drop` | Drop the database schema index |\n| `datafrey client` | Set up an AI client (interactive) |\n| `datafrey client claude` | Configure Claude Code |\n| `datafrey client cursor` | Configure Cursor |\n| `datafrey client mcp` | Print MCP config block for any MCP-compatible client |\n| `datafrey doctor` | Check environment and connectivity |\n\n---\n\n## Supported databases\n\n- **Snowflake** -- fully supported\n\nMore databases are coming soon. [Request a database](https://github.com/datafrey-ai/datafrey/issues/new) if yours is not listed.\n\n---\n\n## Packages\n\nThis 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.\n\n| Package | PyPI | Description |\n| --- | --- | --- |\n| [`datafrey-cli`](packages/datafrey-cli/) | [![PyPI](https://img.shields.io/pypi/v/datafrey?style=flat-square&label=datafrey)](https://pypi.org/project/datafrey/) | CLI for managing databases and MCP setup |\n| [`datafrey-api`](packages/datafrey-api/) | [![PyPI](https://img.shields.io/pypi/v/datafrey-api?style=flat-square&label=datafrey-api)](https://pypi.org/project/datafrey-api/) | Shared Pydantic models for the API |\n| [`datafrey-mcp`](packages/datafrey-mcp/) | [![PyPI](https://img.shields.io/pypi/v/datafrey-mcp?style=flat-square&label=datafrey-mcp)](https://pypi.org/project/datafrey-mcp/) | MCP server bridge |\n| [`datafrey-mock`](packages/datafrey-mock/) | [![PyPI](https://img.shields.io/pypi/v/datafrey-mock?style=flat-square&label=datafrey-mock)](https://pypi.org/project/datafrey-mock/) | Mock API server for local development |\n\n---\n\n## Development\n\nRequires Python 3.13+ and [uv](https://docs.astral.sh/uv/).\n\n```bash\ngit clone https://github.com/datafrey-ai/datafrey.git\ncd datafrey\nuv sync\nuv run pytest\n```\n\nFormat and lint:\n\n```bash\nuv run ruff format packages/\nuv run ruff check packages/\n```\n\n---\n\n## Contributing\n\nContributions are welcome. Please read [CONTRIBUTING.md](CONTRIBUTING.md) before submitting a pull request.\n\nFor 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.\n\n---\n\n## Telemetry\n\nDataFrey 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.\n\n**CLI** — command name, outcome, duration, error class, provider type (e.g. `snowflake`), auth method, CLI version, OS, architecture, Python version, CI flag.\n\n**MCP server** — session start (with the MCP client name, e.g. `claude-ai`, `cursor`), tool name (`plan` or `run`), outcome, duration, and error class.\n\n**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.\n\n**Opt out** by setting either:\n\n```bash\nexport DATAFREY_TELEMETRY_DISABLED=1\nexport DO_NOT_TRACK=1\n```\n\nThe 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.\n\n---\n\n## License\n\nApache 2.0 -- see [LICENSE](LICENSE) for details.\n\nCopyright 2025-2026 DataFrey, Inc.\n",
  "bytes": 5105,
  "sha": "e947552a70c6554134263d8efdda15ed69be4c2359cc3a77587ad7cb48475d32",
  "repo_slug": "datafrey-ai/datafrey",
  "fonte": "repo",
  "truncated": false,
  "api": "https://agentalog.com/api/listings/plg_datafrey_ai_datafrey_datafrey_805a8ad1/readme"
}