TL;DR

Implemented search name in Pipeline Runs.

What changed?

Functional

• API GET /api/pipeline_runs/

  • Search pipeline run name in filter_query. Example query (not URL encoded for example):

    /api?filter_query={"and": [{"value_equals": {"key": "system/pipeline_run.name", "value": "My pipeline run name"}}]}

• API POST /api/pipeline_runs/

  • Copies the pipeline run name to the annotations table for future searching. Example pipeline run annotations table after API is called:

    pipeline_run_id	key	value
    abc123	system/pipeline_run.name	My pipeline run name

Other

• Implemented backfill_pipeline_name_annotations function to populate existing pipeline runs with name annotations

How to test?

uv run pytest tests/test_api_server_sql.py tests/test_filter_query_sql.py tests/test_database_ops.py

• Create pipeline runs with various name configurations and verify that name annotations are properly created
• Test the backfill functionality by removing annotations and running the backfill process
• Verify that filter queries using the new SystemKey.NAME work with different predicate types (exists, equals, contains, in)
• Confirm that both user annotations and system annotations coexist properly

Why make this change?

• This enables users to search and filter pipeline runs by name.
• By mirroring pipeline names as system annotations, pipeline names will be searchable via the annotations table.