majiayu000 You are helping the user check data quality in Snowflake tables.
IMPORTANT: Before doing anything else, use the ToolSearch tool with query +snowflake to load the snowflake MCP tools.
Follow these steps:
Step 1: Define Scope
Ask the user what to check:
- Specific table — run checks on a named table
- Schema-wide — check all tables in a schema
- Critical tables — check tables used by Sigma dashboards
If no preference, default to scanning SIGMA_ANALYTICS.PRODUCTION and SIGMA_ANALYTICS.PRODUCTION_OPERATIONS.
Step 2: Run Quality Checks
For each table, use mcp__snowflake__get_table_stats to get column-level statistics:
- Distinct count per column
- Null count and null percentage per column
- Row count
Step 3: Flag Issues
Apply these quality thresholds:
| Check | Threshold | Severity |
|---|---|---|
| Null percentage | > 50% | Warning |
| Null percentage | > 90% | Critical |
| Zero rows | 0 rows | Critical |
| Low distinct count | < 2 on non-boolean | Warning |
| All nulls | 100% null | Critical |
Step 4: Report Findings
Present a quality report:
Summary
- Tables checked: N
- Tables passing: N
- Tables with warnings: N
- Tables with critical issues: N
Details
| Table | Column | Issue | Null % | Severity |
|---|---|---|---|---|
| SCHEMA.TABLE | col_name | High null rate | 92% | Critical |
Step 5: Recommendations
For each issue:
- Suggest root cause investigation (data pipeline issue, schema change, missing data source)
- Flag if the table is used by downstream dashboards or reports
- Recommend whether to alert the data team
Error Handling
- If Snowflake MCP is unavailable, inform the user and suggest checking connection via
/jf-platform-tools:mcp-status - If a table has too many columns, report the top issues rather than every column
- If access is denied on a table, note it and continue with accessible tables