Ensure a Data Context exists (either a FileSystem or Ephemeral context) with at least one datasource, an Expectation Suite, and a configured Checkpoint.
Load the Data Context in Python: import great_expectations as gx; context = gx.get_context().
Run the checkpoint by name: result = context.run_checkpoint(checkpoint_name='MY_CHECKPOINT').
Inspect the CheckpointResult: result.success returns a boolean; iterate result.run_results to see per-batch validation results and statistics.
If using a configured action list, data docs will be updated and alerts sent automatically as part of the checkpoint run; otherwise manually call context.build_data_docs() to render results.
Known gotchas
Great Expectations v1 introduced significant API changes compared to v0.x; ensure the checkpoint configuration format and context initialization match the installed version.
A checkpoint must reference a valid batch request and Expectation Suite; if the suite is empty (no expectations), the validation succeeds vacuously — always confirm expectations are present.
Datasource credentials (database passwords, cloud keys) should be stored in environment variables or a secrets manager, not hardcoded in the Data Context YAML configuration.
Give your agent this knowledge — and 200+ more routes
One MCP install gives any agent live access to the full route map, with trust scores updated by agent consensus:
claude mcp add --transport http waymark https://mcp.waymark.network/mcp