Skip to main content

Vague

A constraint and probability-based programming language for describing and generating realistic data.

The Problem​

Traditional fake data tools require manual wiring of constraints, relationships, and edge cases. Generating realistic test data that respects business rules is tedious and error-prone. Validation logic often lives separately from generation logic, leading to inconsistencies.

The Solution​

Vague treats ambiguity as a first-class primitive. Declare the shape of valid data and let the runtime determine how to populate it. One schema serves dual purposes: generating test data and validating production data. Think "OpenAPI meets property-based testing."

Key Features​

  • Probabilistic distribution - Weighted probability across options
  • Constraint encoding - Hard constraints ensure data consistency
  • Cross-record references - Link records to filtered instances of other collections
  • Computed fields - Aggregation functions like sum(), count(), avg()
  • Conditional logic - Fields exist only when conditions are met
  • Statistical distributions - Gaussian, lognormal, exponential, Poisson, Beta

Example​

schema Customer {
  name: string,
  status: 0.8: "active" | 0.2: "inactive"
}

schema Invoice {
  customer: any of customers,
  amount: decimal in 100..10000,
  assume amount > 0
}

dataset TestData {
  customers: 50 of Customer,
  invoices: 200 of Invoice
}

Integration​

  • OpenAPI - Import schemas from specs and populate examples
  • CSV I/O - Generate and validate CSV data
  • Schema inference - Reverse-engineer Vague schemas from existing JSONJSON A lightweight data-interchange format that is easy for humans to read and write and easy for machines to parse and generate. JSON is commonly used for transmitting data in web applications. or CSV
  • Validation - Validate generated data against OpenAPI specifications

Usage​

npm install -g vague-lang
vague generate schema.vague -o output.json
vague generate schema.vague -f csv --seed 123