| name | validate-data-product |
| description | Instructions on how to validate an existing or newly created data product in Cortex Framework V7 with mandatory granular status reporting. |
Validating a Data Product
When asked to validate a data product, you MUST follow these steps and provide a status for every sub-step. Your task is not complete until the Final Validation Ledger is populated.
The following steps are mandatory:
- Step 1: Syntax & Compilation Check (Status Required)
- Step 2: Testing & Linting (Status Required)
- Step 3: Structural and Syntax Validation (Status Required)
- Step 4: Field and Data Foundation Validation (Status Required)
- Step 5: Readability & Configuration (Status Required)
- Step 6: Pipeline Execution (Status Required)
Please refer to validation_steps.md for the detailed validation steps to follow.
📊 Mandatory Validation Reporting
At the end of every validation task (successful or otherwise), you MUST generate a comprehensive Data Product Validation Report as an artifact in the standard brain artifacts directory.
- Use the template located at: validation_report_template.md
- Save this artifact with the exact filename
validation_report.md for user review.
- Fill in all placeholders (e.g. GCP parameters, validation statuses, detailed log outputs) accurately based on your run.
- Ensure the Final Validation Ledger inside is fully populated.
- Refer to the generated report in your final response to the user.
