Describe your checks in the Commons rules format, or start from the rule library. Use the data quality check to test them on your own file first.
Pick your platform below. The weave writes the same rules in that platform's own format, ready to paste or load. Anything that won't translate is listed, never dropped quietly.
Use the directory to see each platform's open-source code, its MCP server and its AI agents for cataloguing and metadata, with links to check every claim.
Everything runs in your browser. Nothing you paste leaves your device.
Choose a platform
The same format as rules/*.yaml in the repository. Value sets are expanded for you.
Output
Weaves also run from the command line: python3 -m dqkit export rules.yaml --to glue. The Great Expectations and R validate exports are on the check page.
Worked out by running each weave, so it always matches what the weaves really produce. Native means the platform's own rule type. SQL means a SQL expression or custom SQL rule. A dash means run that check with dqkit.
The Model Context Protocol (MCP) is an open standard that lets AI assistants call tools in other systems. A catalogue's MCP server lets an assistant search assets, read owners and classifications, trace lineage and check quality, using the catalogue's own permissions.
| Vendor | MCP server | What it can do | Source |
|---|
The Commons has its own MCP server too: python3 -m dqkit.mcp_server gives assistants the rule checks, scoring and weaves. See the AI page.