Documentation Index

Fetch the complete documentation index at: https://docs.validatar.com/llms.txt

Use this file to discover all available pages before exploring further.

MCP Tools Reference

Prev Next

This reference documents all tools available through the Validatar MCP Server. Tools are organized into read-only operations (browsing and querying) and write operations (creating, updating, and executing).


Read-Only Tools

These tools retrieve information from your Validatar instance without making changes.


describe

Retrieve detailed information about a Validatar object.

Parameter Type Required Description
object_type enum Yes The type of object to describe (DataSource, Schema, Table, Column, Project, Folder, Test, Job)
object_id integer Yes The ID of the object
output_sections list No Specific sections to include in the response (limits output for large objects)
window string No Time window for time-sensitive data (e.g., recent job results)

Use when: You want to inspect the configuration, status, or metadata of a specific object in your catalog or project.


describe_quick_test

Get the details of a specific quick test available for a table or column.

Parameter Type Required Description
object_type enum Yes Table or Column
object_id integer Yes The ID of the table or column
template_test_id integer Yes The ID of the quick test template to describe

Use when: You want to understand what a specific quick test does before creating it, including its variables and configuration options.


list

List child objects of a parent Validatar object.

Parameter Type Required Description
object_type enum Yes The type of the parent object
object_id integer Yes The ID of the parent object
child_type enum Yes The type of child objects to list
filters object No Filter criteria to narrow results
page_size integer No Number of results per page
page_index integer No Page number for pagination

Use when: You want to browse the hierarchy — for example, listing schemas in a data source, tables in a schema, columns in a table, tests in a folder, or projects on the home page.


list_quick_tests

List available quick tests for a table or column.

Parameter Type Required Description
object_type enum Yes Table or Column
object_id integer Yes The ID of the table or column
filter_by_quality_dimension string No Filter by quality dimension (e.g., Completeness, Uniqueness)
filter_by_test_status string No Filter by whether a test already exists
filter_by_search_text string No Free-text search filter

Use when: You want to see what out-of-the-box tests are available for a specific table or column, optionally filtered by quality dimension or existing test coverage.


list_custom_metadata_options

List custom metadata fields and their options for a Validatar object.

Parameter Type Required Description
object_type enum Yes The type of object
object_id integer Yes The ID of the object
only_return_populated boolean Yes Only return fields that have values set
only_return_visible boolean Yes Only return fields visible in the UI
type_equals_filter string No Filter by field type
name_contains_filter string No Filter by field name
reference_key_contains_filter string No Filter by reference key

Use when: You want to see what custom metadata fields are available for an object and what values they currently hold.


convert_url_to_key

Convert a Validatar application URL to an object type and ID.

Parameter Type Required Description
url string Yes A Validatar application URL

Use when: You have a URL from the Validatar UI and need to determine the object type and ID for use with other MCP tools.


get_global_trust_score_settings

Retrieve the global trust score configuration.

Parameter Type Required Description
output_sections list No Sections to include: QualityDimensions, SeverityLevels, ImpactRatings

Use when: You need to understand how trust scores are calculated, what quality dimensions are configured, or what severity levels and impact ratings are available.


search_documentation_articles

Search across all articles in the Validatar documentation.

Parameter Type Required Description
search_query string Yes The search terms to look for

Use when: You want to find relevant documentation articles by keyword or topic.


search_documentation_assistive

Ask the Validatar documentation assistant a question.

Parameter Type Required Description
prompt string Yes A natural language question about Validatar

Use when: You have a question about how Validatar works and want a synthesized answer drawn from the documentation.


Write Tools

These tools create, modify, or execute operations in your Validatar instance. All write operations are tracked as AI Actions and can be reverted.


create_standard_test

Create a new standard test in a project.

Parameter Type Required Description
project_id integer Yes The project to create the test in
config object Yes Standard test configuration (includes name, folder, data sets, success conditions, and test parameters)

Use when: You want to create a one-off test that validates a specific data condition with explicitly defined source and target queries.


create_template_test

Create a new template test in a project.

Parameter Type Required Description
project_id integer Yes The project to create the test in
config object Yes Template test configuration (includes name, folder, metadata selection criteria, variables, and test logic)

Use when: You want to create a reusable test definition that automatically generates child tests based on metadata selection criteria.


create_quick_test

Create a quick test for a table or column from an available template.

Parameter Type Required Description
object_type enum Yes Table or Column
object_id integer Yes The ID of the table or column
template_test_id integer Yes The ID of the quick test template to use
variables object Yes Variable values required by the template
custom_fields object Yes Custom field values to assign to the created test

Use when: You want to quickly add test coverage to a table or column using one of the predefined quick test templates.


create_data_source

Create a new data source.

Parameter Type Required Description
config object Yes Data source configuration (includes name, connection details, template, and settings)

Use when: You want to register a new database connection in Validatar for testing and profiling.


update_data_source

Update an existing data source configuration.

Parameter Type Required Description
data_source_id integer Yes The ID of the data source to update
config object Yes Updated data source configuration

Use when: You need to modify connection details, settings, or other properties of an existing data source.


execute_standard_tests

Execute one or more standard tests.

Parameter Type Required Description
object_id_list list of integers Yes List of standard test IDs to execute (not template test IDs)

Use when: You want to run specific tests and see their results. Only standard tests (including child tests materialized from templates) can be executed directly.


execute_data_source_script

Execute a SQL or Python script against a data source.

Parameter Type Required Description
data_source_id integer Yes The data source to execute against
script string Yes The SQL or Python script to run
timeout_seconds integer No Maximum execution time (default applies if omitted)
max_rows integer No Maximum number of rows to return

Use when: You want to run ad-hoc queries to explore data, validate assumptions, or troubleshoot test results.


update_custom_metadata

Update custom metadata field values for a Validatar object.

Parameter Type Required Description
object_type enum Yes The type of object to update
object_id integer Yes The ID of the object
field_values object Yes Field values to set, structured as {"field_name": {"values": ["value1", "value2"]}}

Use when: You want to tag or classify objects with custom metadata — for example, assigning data domains, owners, or sensitivity levels.


create_folder

Create a new folder in a project.

Parameter Type Required Description
project_id integer Yes The project to create the folder in
parent_folder_id integer Yes The parent folder ID
name string Yes Name for the new folder

Use when: You want to organize tests into a folder structure within a project.


rename_folder

Rename an existing folder.

Parameter Type Required Description
project_id integer Yes The project containing the folder
folder_id integer Yes The folder to rename
name string Yes The new folder name

Use when: You want to update a folder's display name.


move_folder

Move a folder to a different parent.

Parameter Type Required Description
project_id integer Yes The project containing the folder
folder_id integer Yes The folder to move
parent_folder_id integer Yes The new parent folder ID

Use when: You want to reorganize the folder hierarchy within a project.


delete_folder

Delete a folder from a project.

Parameter Type Required Description
project_id integer Yes The project containing the folder
folder_id integer Yes The folder to delete
delete_child_tests boolean Yes Whether to also delete tests inside the folder

Use when: You want to remove an unused folder. Set delete_child_tests to true to remove contained tests, or false to move them to the parent folder.


create_project

Create a new project.

Parameter Type Required Description
config object Yes Project configuration (includes name and settings)

Use when: You want to set up a new project for organizing and executing tests.


update_project

Update an existing project's configuration.

Parameter Type Required Description
config object Yes Updated project configuration

Use when: You need to modify project settings or properties.


revert_action

Revert a change made by a previously executed AI Action.

Parameter Type Required Description
action_key string Yes The revert action key returned by the original write operation
comment string Yes A comment explaining why the action is being reverted
reason enum Yes Reason for reverting: AiDidNotDoWhatWasRequested, ChangeNoLongerDesired, Testing, or Other

Use when: An AI-initiated change needs to be undone — for example, a test was created incorrectly or a metadata update needs to be rolled back.


Object Types

The following object types are used across tools that accept an object_type parameter:

Object Type Description
Home The top-level entry point
Catalog The data catalog
DataSource A configured database connection
Schema A schema within a data source
Table A table within a schema
Column A column within a table
Project A test project
Folder A folder within a project
Test A standard or template test
Job A test execution job
Report A report definition
GlobalSettings Instance-wide settings
Marketplace The Validatar marketplace

Custom Field Value Format

Several tools accept or return custom field values. The format is consistent:

{
  "field_name": {
    "values": ["value1", "value2"]
  },
  "another_field": {
    "values": ["single_value"]
  }
}

Each field accepts a list of values — even single-value fields use an array.


Related