lvt-seed-data

name: lvt-seed-data description: Use when generating test data for LiveTemplate apps - covers seeding resources with realistic fake data, cleanup, and understanding data generation patterns

lvt:seed-data

Generate realistic test data for development and testing.

🎯 ACTIVATION RULES

Context Detection

This skill typically runs in existing LiveTemplate projects (.lvtrc exists).

✅ Context Established By:

1. Project context - .lvtrc exists (most common scenario)
2. Agent context - User is working with lvt-assistant agent
3. Keyword context - User mentions "lvt", "livetemplate", or "lt"

Keyword matching (case-insensitive): lvt, livetemplate, lt

Trigger Patterns

With Context: ✅ Generic prompts related to this skill's purpose

Without Context (needs keywords): ✅ Must mention "lvt", "livetemplate", or "lt" ❌ Generic requests without keywords

Overview

LiveTemplate includes a seeder that generates realistic test data based on your schema. It uses field names to generate contextually appropriate values (e.g., "email" → fake email addresses).

Key features:

• Context-aware generation (field names determine data type)
• Bulk insert with progress tracking
• Cleanup of test data
• Marks test records for easy identification

Basic Usage

# Generate 50 products
lvt seed products --count 50

# Clean up test data
lvt seed products --cleanup

# Clean up then seed fresh data
lvt seed products --cleanup --count 30

Prerequisites

Before seeding:

1. ✓ Resource generated (lvt gen resource)
2. ✓ Migrations applied (lvt migration up)
3. ✓ Database exists (app.db)

# Complete setup before seeding
lvt gen resource products name price:float
lvt migration up
lvt seed products --count 50

Commands

Seed Data

# Basic seeding
lvt seed <resource-name> --count N

# Examples
lvt seed users --count 100
lvt seed products --count 50
lvt seed tasks --count 200

Progress tracking:

Seeding products with 50 rows...
  Progress: 10/50
  Progress: 20/50
  Progress: 30/50
  Progress: 40/50
  Progress: 50/50
✅ Successfully seeded 50 rows

Total test records in products: 50

Clean Up Test Data

# Remove all test records
lvt seed <resource-name> --cleanup

# Example
lvt seed products --cleanup

Output:

Cleaning up test data for products...
✅ Deleted 50 test records from products

Clean + Reseed

# Clean old data then seed fresh
lvt seed <resource-name> --cleanup --count N

# Example - replace old test data with 100 new records
lvt seed users --cleanup --count 100

Context-Aware Generation

The seeder generates realistic data based on field names:

Auto-managed fields:

• id - Auto-generated (marked with test prefix)
• created_at - Current timestamp
• updated_at - Current timestamp

Fallback for unknown fields:

• String fields → Random text (5-10 words)
• Numeric fields → Random number (1-1000)
• Bool fields → Random true/false

Examples

User Data

# Generate 100 users
lvt seed users --count 100

Generated fields:

• name → "Alice Johnson"
• email → "alice.johnson@example.com"
• phone → "(555) 234-5678"
• age → 34
• created_at → 2024-11-04 18:30:00

Product Data

# Generate 50 products
lvt seed products --count 50

Generated fields:

• name → "Innovative Widget"
• description → "High quality product with excellent features..."
• price → 149.99
• quantity → 87
• category → "Electronics"
• status → "active"

Task Data

# Generate 200 tasks
lvt seed tasks --count 200

Generated fields:

• title → "Complete Project Documentation"
• description → "Finalize all documentation for the Q4 release..."
• status → "pending"
• priority → "high"

Test Record Identification

All test records are marked with special IDs starting with test-seed-:

-- In database
id: test-seed-00001
id: test-seed-00002
id: test-seed-00003

This allows:

• Easy cleanup (--cleanup finds these records)
• Visual identification in database browser
• Prevents mixing with real user data

Development Workflow

# 1. Create and setup resource
lvt gen resource products name price:float description
lvt migration up

# 2. Seed initial test data
lvt seed products --count 50

# 3. Develop and test with realistic data
lvt serve
# Browse to http://localhost:8080/products

# 4. Need fresh data?
lvt seed products --cleanup --count 50

# 5. Need more data?
lvt seed products --count 100
# Now 150 total (50 + 100)

Common Patterns

Testing Pagination

# Generate enough data to test pagination
lvt seed products --count 100

# Default page size is 20, so 100 records = 5 pages

Testing Search

# Generate diverse data for search testing
lvt seed products --count 50
lvt seed users --count 30
lvt seed tasks --count 100

Clean State

# Start fresh for new feature
lvt seed products --cleanup
lvt seed users --cleanup
lvt seed tasks --cleanup

# Reseed with specific amounts
lvt seed products --count 20
lvt seed users --count 10

Demo Data

# Prepare demo with realistic amounts
lvt seed users --cleanup --count 15
lvt seed products --cleanup --count 30
lvt seed orders --cleanup --count 50

Choosing Record Counts

Development (fast feedback):

• Small: 10-20 records (test basic CRUD)
• Medium: 50-100 records (test pagination)
• Large: 500-1000 records (test performance)

Testing:

• Unit tests: 5-10 records (specific scenarios)
• E2E tests: 20-50 records (realistic flows)
• Load tests: 1000+ records (stress testing)

Demo:

• Small dataset: 10-30 records (easy to navigate)
• Diverse data: Multiple resources with 15-20 each
• Representative: Enough to show features without overwhelming

Rule of thumb:

• Start small (20-50 records)
• Increase as needed for specific tests
• Use --cleanup to reset between scenarios

Common Issues

❌ Resource Not Found

# Error: resource 'product' not found in schema

# Cause: Resource doesn't exist in schema.yaml
# The seeder reads your app's schema.yaml to find resource definitions

# Fix 1: Check spelling (must match exactly)
lvt seed products --count 50  # not 'product' (case-sensitive)

# Fix 2: Generate resource first
lvt gen resource products name price:float
lvt migration up
lvt seed products --count 50

# Fix 3: Verify schema.yaml contains the resource
cat schema.yaml | grep products

❌ Database Not Found

# Error: database not found

# Cause: Haven't run migrations yet

# Fix: Run migrations first
lvt migration up
lvt seed products --count 50

❌ No Flags Specified

# Error: either --count or --cleanup must be specified

# Cause: Forgot to specify what to do

# Fix: Add --count or --cleanup
lvt seed products --count 50  # seed data
lvt seed products --cleanup   # cleanup data

❌ Foreign Key Constraints

# Error: FOREIGN KEY constraint failed

# Cause: Seeding child table without parent data

# Fix: Seed parent table first
lvt seed posts --count 20      # parent
lvt seed comments --count 100  # child (references posts)

Field Name Best Practices

For better generated data, use descriptive field names:

# ✅ Good - context-aware generation
lvt gen resource users email phone_number full_name

# ❌ Less good - generic generation
lvt gen resource users field1 field2 field3

Examples:

• Use email not user_email_address
• Use phone not contact_number
• Use price not item_cost
• Use description not desc

The seeder recognizes common patterns in field names.

Cleanup Strategy

During development:

# Clean and reseed often
lvt seed products --cleanup --count 50

Before demo:

# Clean all, then seed specific amounts
lvt seed users --cleanup --count 10
lvt seed products --cleanup --count 20

Testing edge cases:

# Small dataset
lvt seed products --cleanup --count 5

# Large dataset
lvt seed products --cleanup --count 500

Integration with Testing

# Before running tests
lvt seed products --cleanup --count 30
go test ./app/products

# After tests
lvt seed products --cleanup

Quick Reference

Remember

✓ Seed parent resources before children (foreign keys) ✓ Use descriptive field names for better data ✓ Test records marked with "test-seed-" prefix ✓ --cleanup only removes test records, not real data ✓ Seeding adds to existing data unless you use --cleanup

✗ Don't forget to run migrations before seeding ✗ Don't seed without checking foreign key dependencies ✗ Use resource name from schema.yaml, not database table name

• Resource name: products (as defined in schema.yaml)
• Table name: products (database table, should match but use resource name for seeding)