| name | data-validator |
| description | Validate and transform data files including CSV, JSON, and XML. Use when asked to check data formats, validate schemas, find data quality issues, or convert between data formats. Includes executable scripts for automated validation. |
| license | MIT |
| compatibility | Requires python3 for script execution |
| metadata | {"author":"contoso-data-engineering","version":"1.0"} |
| allowed-tools | run_skill_script |
Data Validator Skill
You validate and transform structured data files. Follow these instructions carefully.
Capabilities
- Schema validation — Check that data conforms to an expected schema (required fields, types, constraints).
- Format detection — Identify whether input is CSV, JSON, XML, or another format.
- Quality checks — Find missing values, duplicates, outliers, and inconsistencies.
- Data transformation — Convert between formats or normalize field values.
Available Scripts
This skill includes executable scripts in the scripts/ directory:
| Script | Purpose | Arguments |
|---|
validate_csv.py | Validate CSV structure and data quality | --file <path> --schema <path> |
validate_json.py | Validate JSON against a JSON Schema | --file <path> --schema <path> |
summarize_data.py | Generate summary statistics for a dataset | --file <path> --format csv|json |
How to Use Scripts
When the user provides data to validate:
- Determine the data format (CSV, JSON, etc.).
- Use
run_skill_script with the appropriate script and arguments.
- Interpret the script output and present findings clearly.
Script Output Format
All scripts return JSON:
{
"status": "valid | invalid | warning",
"errors": [{"row": 1, "field": "email", "message": "Invalid format"}],
"warnings": [{"row": 5, "field": "age", "message": "Possible outlier: 200"}],
"summary": "Processed 100 rows, 2 errors, 1 warning"
}
Response Format
## Validation Results
**Format:** [CSV/JSON/XML]
**Rows processed:** [count]
**Status:** [Valid / Invalid / Warnings]
### Errors (must fix)
| Row | Field | Issue |
|-----|-------|-------|
### Warnings (review recommended)
| Row | Field | Issue |
|-----|-------|-------|
### Summary
[Brief assessment and recommended actions]
For schema definition templates, load the references/SCHEMA_GUIDE.md resource.
