Sqlmesh

:sparkles: Features
* Column comments in python models (2712) treysp
* Support name inference from schema directories (2697) Themiscodes

:beetle: Bug fixes
* Add the dbt "var" macro to the "builtins" package (2723) izeigerman
* Add migration script for column_descriptions_ rename (2724) treysp
* Capture bultin macro dependencies when loading dbt projects (2725) izeigerman
* Broken build due to the hyperscript upgrade (2729) izeigerman
* Normalize column names in column_descriptions dict (2731) treysp
* Make sure to commit a transaction when fetching a dataframe using Pandas API (2732) izeigerman
* Only check destructive changes for directly modified models at plan time (2733) izeigerman

:broom: Chore
* Update UI models to subclass PydanticModel (2726) vchan

:alien: Other
* External models default yaml (2717) dbrtly

:beetle: Bug fixes
* Render audit condition when used in a WHERE(condition) clause (2713) georgesittas
* Yet another edge case related to snowflake staged file syntax (2714) georgesittas
* We can't inspect built-in routines to get their file (2720) georgesittas

:broom: Chore
* Update ibis example (2719) georgesittas
* Add missing semicolon after MODEL statement in Table Migration docs (2721) schlich

:sparkles: Features
* Databricks allow disabling spark session (2703) eakmanrq

:beetle: Bug fixes
* Avoid using concat_ws when there's only one key is present (2711) izeigerman

:broom: Chore
* Allow to skip connection testing in the 'info' command (2705) izeigerman
* Bump sqlglot to v24.1.0 (2708) georgesittas
* Include the user agent information as part of the MotherDuck connection config (2706) izeigerman

:alien: Other
* Update info magic help string (2707) treysp

:beetle: Bug fixes
* **(motherduck)** Filter on catalog name in DuckDB _get_data_objects query (2699) treysp
* Exclude private key from the state sync fingerprint calculation (2704) izeigerman

:broom: Chore
* Fix typo tobymao

:alien: Other
* Add UI guide (2700) treysp

:broom: Chore
* Ping sqlglot to 24.0.2 izeigerman

:sparkles: Features
* Snowpark (2666) tobymao
* Introduce INCREMENTAL_BY_PARTITION model kind (2687) izeigerman

:beetle: Bug fixes
* Allow partial query results in unit tests, update docs (2676) georgesittas
* Add missing reporting for builtin runs (2675) izeigerman
* For MSSQL apply the same workaround for the varchar size as we use for Redshift (2674) izeigerman
* Correct destructive change error messages (2678) treysp
* Use exp.column when casting seed columns to avoid ParseError (2686) georgesittas
* Unescape sequences in CSV settings (2690) georgesittas
* Remove remaining dynamic overwrite modes (2692) eakmanrq
* Ensure MacroEvaluator.columns_to_types returns types as DataTypes (2694) georgesittas

:broom: Chore
* Add google analytics to docs (2682) treysp
* Update unit test docs to include specifying query as fallback (2683) georgesittas
* Update docs for ui formatter option (2685) simonprydden
* Allow usage of start_* / end_* macros in SCD2 models (2691) izeigerman
* Don't store the seed's content in the state (2688) izeigerman
* Fix airflow tests izeigerman

:alien: Other
* Add csv_settings to seed models concepts doc (2679) treysp
* Clarify incremental by unique key in kinds concepts doc (2680) treysp
* Add docs for incremental by partition model kind (2695) treysp