Amin Ghadersohi a26a383386 feat(mcp): add semantic layer MCP tools ...

Adds a semantic-layer surface to the MCP service for LLM-driven metric
discovery and querying across both built-in SqlaTable datasets and
external SemanticView-backed sources:

- list_metrics: unified metric discovery with inline compatible
  dimensions and pagination, sorted deterministically.
- get_table: query a dataset/semantic view by metric + dimension name,
  with time_column/time_range validation.
- get_compatible_dimensions / get_compatible_metrics: progressive
  query-builder support, resolving per-metric compatibility for
  external views.

Includes shared utilities (query_utils.validate_names,
response_utils.format_data_columns), DAO permission filtering
(SemanticViewDAO.find_accessible), and unit test coverage for all
four tools plus the shared utilities and DAO methods.

2026-07-07 03:44:45 +00:00

..

api_test.py

…

dao_test.py

feat(mcp): add semantic layer MCP tools

2026-07-07 03:44:45 +00:00

decorators_test.py

…

labels_test.py

…

mapper_test.py

fix(semantic layers): start/end ranges (#41590)

2026-07-04 16:46:26 -07:00

models_test.py

…

schemas_test.py

…

types_test.py

…