Alation

Render an Alation data product spec as a branded card with contract status, owner, and recordsets. Optionally accepts include_citations=true to render a citation strip beneath the card.

Render an Alation document hub page (experiment plan, launch notes, incident, policy) as a branded citation card. Use when the model cites a document by id.

Render an Alation glossary term as a branded citation card. Use whenever the model cites a metric definition (MRR, Activation, Churn, etc.) to give the analyst owner, status, canonical source column, and a deep link.

Get Custom Fields Definitions: Get comprehensive custom field definitions for the Alation instance. This tool retrieves custom fields, built-in metadata fields, and built-in search facets with structured metadata and usage guidance. It's designed to help with data dictionary creation, metadata updates, and understanding field capabilities. Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. Returns: ToolReturn: JSON formatted response containing: - custom_fields: List of custom field definitions with essential properties - built_in_fields: List of built-in metadata fields (title, description, steward) - built_in_search_facets: List of built-in search facets (author, owner, etc.) - usage_guide: Comprehensive guidance for using fields in various contexts Raises: ToolExecutionError: When the custom field API fails deterministically.

Get Data Product Raw Specification: Get the raw data product specification in the exact format expected by the update API. This returns the complete JSON specification that can be modified and passed to the update_data_product_spec tool. This differs from the get_data_schema tool, because that adds sample values and simplifies the data product for use in question-answering and SQL-generation. This provides the raw JSON. Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. data_product_id: The ID of the data product to fetch. data_product_version: The version of the data product to fetch. If None, extracted from input_payload or uses the default version. Returns: ToolReturn with the complete raw data product specification as a JSON string.

Get Data Schema: Get the schema for the set of tables the user is asking about. The schema can be useful in a variety of downstream applications. Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. This contains the data product already. data_product_id: The ID of the data product to get the schema for. data_product_version: The version of the data product to get the schema for. If None, extracted from input_payload or uses the default version. pre_exec_sql: An arbitrary SQL query to run in the session before the standard queries. auth_id: ID of the authentication credentials to use. auth_mode: Authentication mode ('datasource_service_account' or None for user credentials). allow_fallback_auth: Whether to allow falling back to service account creds. This is a temporary workaround to avoid having to specify an auth_id and will be removed in the near future. query: An optional query to use for semantic search of sample values.

Get Filter Context: List/enumerate objects from the Alation data catalog by filters (no keyword search). WHEN TO USE THIS TOOL: Use ONLY when there is NO name, keyword, or concept to search for - pure enumeration. If the question includes any search term, use get_search_context instead (supports filters). Examples: - "List all tables in schema FINANCE" -> filter by schema_name - "Show all PII-tagged tables" -> filter by flag_types or tag_ids - "What reports are in the Sales folder?" -> filter by parent_folder - "List columns with data type VARCHAR" -> filter by data_type DO NOT USE when there's a name/keyword to search for (e.g., "Find sales tables" or "Find columns named X in schema Y") - use get_search_context instead. SUPPORTED OBJECT TYPES: table, column, schema, datasource, query, documentation, bi_report, bi_field, bi_folder, bi_datasource Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. question: Natural language question describing what to enumerate and filter criteria. base_signature: Optional pre-built signature. If complete (has fields_required), uses it directly. If partial, agent completes it while respecting scope. ids_by_object_type: Optional dict mapping object types to lists of IDs (e.g. {"table": [1, 2, 3]}). When provided, skips search and fetches context directly by IDs. Returns: JSON with "signature" (the filter parameters used) and "context" (matching catalog objects with metadata like name, description, URL, schema, and custom fields).

Get Search Context: Search the Alation data catalog by name, keyword, or business concept. WHEN TO USE THIS TOOL: Use this tool whenever the question mentions ANY name, keyword, or concept to search for. This includes combined search + filter queries like "Find columns named X in schema Y". Examples: - "Find sales tables" -> search for "sales" - "What columns are related to revenue?" -> search for "revenue" - "Find columns named customer_id in schema FINANCE" -> search + filter - "What tables contain column order_date?" -> search for column, include parent objects DO NOT USE when there is no search term at all (pure enumeration like "List all tables in schema X") - use get_filter_context instead. SUPPORTED OBJECT TYPES: table, column, schema, datasource, query, documentation, bi_report, bi_field, bi_folder, bi_datasource Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. question: Natural language question about the data catalog. Include what you're searching for and any filters to apply. base_signature: Optional pre-built signature defining allowed object types. MUST be a nested dict mapping object type to its config, e.g.:: {"table": {"search_phrases": ["sales"], "fields_required": ["name"], "limit": 100}, "column": {"search_phrases": ["revenue"]}} Do NOT flatten to {"object_types": [...], "search_phrases": [...], "limit": ...} - that shape will be rejected. Returns: JSON with "signature" (the search parameters used) and "context" (matching catalog objects with metadata like name, description, URL, schema, and custom fields).

Data Product Search: All listed data products in the marketplace pertaining to the query. Only returns data products that are listed in the marketplace. Args: ctx (RunContext[AAIRunContext]): The run context containing dependencies. Automatically passed in. search_term: This can be the user question directly. It will be used to search for relevant data products. limit: Maximum number of results to return. Default is 5. marketplace_id: The marketplace ID to search in. If not provided, the default marketplace will be used. Returns: ToolReturn with the list of relevant data products.

Catalog Context Search: Execute the agent with the provided arguments.