great-expectationslisted
Install: claude install-skill testland/qa-skills
# great-expectations
## Overview
GX Core is the modern Python library for programmatic data validation
workflows. The shape is: **DataSource → DataAsset → BatchDefinition →
ExpectationSuite → ValidationDefinition → Checkpoint** ([gx-overview][1]).
This skill covers authoring expectations, running them via a
ValidationDefinition or Checkpoint, parsing the JSON result, and gating
CI on it.
[1]: https://docs.greatexpectations.io/docs/core/introduction/gx_overview
## When to use
- The repo imports `great_expectations` (Python).
- The user asks about `ExpectColumn*`, `ExpectationSuite`, `Checkpoint`,
Data Docs, or `gx.get_context()`.
- A pipeline needs row-level / column-level assertions on Pandas, SQL,
or Spark data with a programmatic interface (rather than a YAML-only
config like Soda).
- A CI workflow needs to gate a run on a validation result and surface
failures in Data Docs or Slack.
## Authoring expectations
The four key objects to compose ([gx-overview][1]):
1. **DataSource** - represents a connection to a data store (Pandas /
SQL / Spark / files).
2. **DataAsset** - a collection of records inside a DataSource (e.g. a
table, a directory of partitioned files).
3. **BatchDefinition** - slices a DataAsset into validatable batches
(whole table, partition, dataframe).
4. **ExpectationSuite** - a named collection of `Expectation` objects
that describe what the data should look like.
Expectations themselves come from the `gxe` namespace ([create-an-e