Skip to main content

msh ai review

Use AI to review an asset for risks and issues.

Usage​

msh ai review <asset_path> [--json]

Description​

Analyzes an asset and identifies:

  • Performance risks: Missing indexes, inefficient queries
  • Data quality risks: Missing tests, potential null issues
  • Glossary alignment issues: Columns not linked to glossary terms
  • Best practice violations: Common anti-patterns
  • Suggested improvements: Actionable recommendations

Options​

  • --json: Return structured review object

Examples​

Basic Review​

msh ai review assets/revenue.msh

Example Output:

Asset: revenue

Summary:
This asset calculates monthly revenue per customer. Overall structure
is good, but there are some performance and quality concerns.

Risks:
⚠️  Performance: Missing index on customer_id column
⚠️  Performance: Large JOIN without WHERE clause may be slow
⚠️  Data Quality: No tests defined for revenue calculations
⚠️  Data Quality: Potential null values in amount column

Glossary Issues:
- Column 'customer_id' not linked to 'Customer' term
- Column 'revenue' not linked to 'Revenue' metric

Performance Notes:
- Consider adding WHERE clause to filter recent data
- Consider partitioning by month for better performance

Suggested Changes:
1. Add tests for revenue calculations
2. Link columns to glossary terms
3. Add WHERE clause for incremental processing
4. Add index on customer_id

Structured Output​

msh ai review assets/revenue.msh --json

Example JSON Output:

{
  "summary": "This asset calculates monthly revenue...",
  "risks": [
    {
      "type": "performance",
      "severity": "warning",
      "message": "Missing index on customer_id column"
    },
    {
      "type": "data_quality",
      "severity": "warning",
      "message": "No tests defined for revenue calculations"
    }
  ],
  "glossary_issues": [
    {
      "column": "customer_id",
      "suggestion": "Link to 'Customer' term"
    }
  ],
  "performance_notes": [
    "Consider adding WHERE clause to filter recent data"
  ],
  "suggested_changes": [
    "Add tests for revenue calculations",
    "Link columns to glossary terms"
  ]
}

Risk Types​

Performance Risks​

  • Missing indexes
  • Inefficient queries (large JOINs, no WHERE clauses)
  • Full table scans
  • Missing partitioning

Data Quality Risks​

  • Missing tests
  • Potential null values
  • Type mismatches
  • Missing constraints

Glossary Alignment Issues​

  • Columns not linked to glossary terms
  • Missing business context
  • Unclear column names

Best Practice Violations​

  • Anti-patterns
  • Code smells
  • Security concerns

Prerequisites​

  1. AI Configuration: Must configure AI provider first

    msh config ai --provider openai --model gpt-4 --api-key env:OPENAI_API_KEY

  2. Manifest: Generate manifest before using AI commands

    msh manifest

  3. Glossary (optional): Link glossary terms for better reviews

    msh glossary link-term "Customer" --asset revenue --column customer_id

Use Cases​

  • Pre-Deployment Review: Review assets before deploying to production
  • Code Quality: Identify potential issues early
  • Performance Optimization: Find performance bottlenecks
  • Documentation: Ensure glossary alignment

Acting on Suggestions​

After reviewing suggestions, you can:

  1. Apply fixes manually: Review and implement suggested changes
  2. Use AI to fix: Let AI generate fixes automatically 
    msh ai fix assets/revenue.msh --apply
  3. Generate tests: Use AI to generate tests 
    msh ai tests assets/revenue.msh --apply