Start with
docs/AUTHORING_QUEST_PACKS.md— the single guided authoring walkthrough. This file is the copy-paste validator-pattern appendix.
type: sql_assertion
mode: sync
statement: |
SELECT COUNT(*) AS cnt
FROM ${team_catalog}.${team_schema}.target_table
expect:
operator: ">="
value: 1Supported operators:
=, !=, >, >=, <, <=, contains, not_contains, is_true, is_false
type: databricks_sdk
mode: sync
check: job_exists_with_schedule
params:
name_pattern: "${team_slug}-daily-ingest"
trigger_type: CRONtype: databricks_sdk
mode: async
check: pipeline_update_completed
params:
name_pattern: "${team_slug}-customer-pipeline"
created_after: "${event_start}"type: system_table
mode: async
table: system.query.history
condition: |
executed_by IN (${team_members})
AND start_time >= ${event_start}
AND execution_status = 'FINISHED'
AND LOWER(statement_text) LIKE '%ai_query%'
expect:
min_rows: 1type: notebook
mode: async
notebook_path: /Workspace/Shared/quest_validators/check_model_quality
params:
team_catalog: "${team_catalog}"
team_schema: "${team_schema}"
endpoint_name: "${team_slug}-endpoint"
timeout_seconds: 300Notebook output contract:
{
"status": "passed",
"message": "Validation passed",
"score_delta": 200,
"evidence": {}
}type: manual
mode: async
rubric_md: |
Award pass if the team can explain the architecture clearly and justify governance choices.
points: 100