接入你的智能体
從發布者安裝。不是 PlugMCP 端點。
遠端 URL 還是本機指令？ 兩種安裝方式的差異
DataOps TestGen is an open-source data quality tool that profiles tables, generates tests from the profiling results, and runs them on a schedule with freshness, volume, and schema monitoring.
The open-source edition includes the full web UI, not just a library. You also get a REST API and an MCP server, so you can drive TestGen from your pipelines or an AI assistant. It is Apache 2.0 licensed and runs the same core engine and UI as TestGen Enterprise.
As of TestGen 5.134.0, TestGen connects to 14 databases:
TestGen also reads CSV, Parquet, and JSON files on a local path, in Amazon S3, or in Azure Blob Storage, with no database in front. Every source works in both the open-source and Enterprise editions.
To set up a connection, see Connect your data.
The dk-installer program installs TestGen in either Docker or pip mode.
You need 2 CPUs, 8 GB memory, 20 GB disk space. pip mode needs Python 3.9+ to run the installer. Docker mode also needs Docker 27+ and Docker Compose 5.0+.
Download the installer into a new, empty directory:
curl -o dk-installer.py 'https://raw.githubusercontent.com/DataKitchen/data-observability-installer/main/dk-installer.py'
On Windows, you can instead download dk-installer.exe and double-click it.
Install TestGen:
python3 dk-installer.py tg install
The installer checks for Docker and asks you to pick Docker or pip. Pass --docker or --pip to skip the prompt. When it finishes, it writes credentials to dk-tg-credentials.txt, generates demo data, and opens the TestGen UI in your browser. Use --no-demo to skip the demo data.
Upgrade later with python3 dk-installer.py tg upgrade. Restart with python3 dk-installer.py tg start.
For complete instructions, see the documentation:
Then walk through the Data Observability Overview Demo, and connect your own database.
TestGen ships a Model Context Protocol (MCP) server in the open-source edition. Connect Claude, Claude Code, Cursor, GitHub Copilot, or any MCP client, and ask about your data in plain English. The assistant can profile tables, generate and run tests, and triage failures. Every tool runs under your own TestGen role, so the assistant can do only what you can do.
The installer serves the API and MCP server on port 8530 by default. Change it with --api-port.
Setup instructions: docs.datakitchen.io/testgen/mcp
The core profiling, testing, and monitoring engine is free and open source, with unlimited tables and unlimited data volume. Enterprise adds team access, security, Commercial Pharma tests, source-data controls for AI, and support.
See datakitchen.io/pricing for current pricing.
Talk with other data practitioners who build with DataKitchen. Share what you know, get help, and contribute.
#support channel.For details on contributing or running the project for development, see the contributing guide.
DataOps TestGen is Apache 2.0 licensed. See LICENSE.
Open-source data quality tool: profiles tables, generates data quality tests from the profiling results, and runs them on a schedule with freshness, volume, and schema monitoring. Web UI, REST API, an
總覽
Open-source data quality tool: profiles tables, generates data quality tests from the profiling results, and runs them on a schedule with freshness, volume, and schema monitoring. Web UI, REST API, and MCP server. 14 databases plus CSV, Parquet, and JSON files. Apache 2.0.
工具
尚無工具清單。遠端伺服器會在首次健康檢查後取得工具清單。
常見問題
在本頁複製適用於 Cursor、Claude Desktop、Claude Code 或 VS Code 的 stdio 設定。它使用發布者的安裝方式，而不是 PlugMCP 代理伺服器。
這筆記錄未列出必要的環境變數。
Dataops Testgen 透過 stdio 在本機執行。
上次檢查時間：2026-11-14。 Open-source data quality tool: profiles tables, generates data quality tests from the profiling results, and runs them on a schedule with freshness, volume, and schema monitoring. Web UI, REST API, and MCP server. 14 databases plus CSV, Parquet, and JSON files. Apache 2.0.