MockHero is a synthetic test data API with the goal of "generating realistic, relational test data for any database schema." It generates data by detecting the database schema in SQL or JSON, and supports JSON, CSV, and SQL output formats.
Main capabilities:
- Covers more than 15 categories and a total of 156 field types, with categories including identity, location, finance, time, and more.
- Supports relational data: foreign keys work correctly, for example orders reference real user IDs, and comments are associated with real products; a single API call can seed an entire database.
- Provides 3 input modes: structured schema, plain English prompts, and preset templates.
- Automatic locale: after adding a country enum field, names, emails, and phone numbers match each row's nationality, supporting 22 locales.
- Output supports JSON, CSV, and SQL, with SQL dialects covering Postgres, MySQL, and SQLite.
- Deterministic seed: the same seed plus the same schema generates exactly the same data every time.
How to use:
- Use via API.
- Use via plain English prompts.
- Use preset schemas for common patterns without defining a schema yourself; currently available preset patterns are ecommerce, blog, saas, and social.
- Use MockHero in AI agents via Model Context Protocol (MCP).
- Provides a hosted remote MCP endpoint for agent-first clients, which exposes estimate, loginless Polar checkout, checkout status, API key claim, schema detection, templates, and generation tools.
- Provides a local stdio MCP server, suitable for agents that can run npm packages; for installation and configuration, see the setup instructions for @mockherodev/mcp-server.
Tech stack: Next.js 15 (App Router), Supabase (Postgres), Polar (Billing).
Use cases: Generate realistic synthetic test data for database schemas, suitable for testing scenarios that need to seed an entire database, keep foreign key relationships consistent, and want repeatable generation of the same datasetScenarios.
Limitations: More details about billing, pricing, and specific usage steps are not yet provided.
