Misata: synthetic data library — Claude Code MCP server: install
Builds relational synthetic datasets with matching foreign keys and exact mathematical totals. Useful for seeding databases and validating business logic locally.
claude mcp add misata -- uvx misataWhat Misata: synthetic data library does
Misata turns natural language stories, YAML schemas, or existing database connections into multi-table relational data. Instead of randomly generating numbers column by column, it works backwards from target outcomes. Order totals, churn rates, and balance sums reconcile down to the cent without machine learning models.
The engine enforces relational integrity across all generated tables. Timestamps follow chronological logic, and foreign keys remain valid throughout the set. Built-in audit and realism commands scan output CSV files for contradictions like deliveries preceding order dates.
You need uv installed in your environment to execute uvx misata. Without Python and the tool runner available, the coding assistant cannot spin up the server.
What it helps with
- populate relational tables without orphan foreign keys
- reconcile aggregate row sums to exact financial targets
- fill database columns with domain-authentic text
- audit generated datasets for chronological and logical conflicts
Example prompts for the agent
- Generate a synthetic database for an e-commerce platform with 500 customers and realistic order amounts
- Build tables from
misata.yamland verify foreign key consistency - Audit the generated CSV files in
./datafor date contradictions
How to install Misata: synthetic data library
Platform CLI
claude mcp add misata -- uvx misatauvx runs the command named after the package; if the README names another one, use uvx --from misata <command>
Platform CLI
codex mcp add misata-synthetic-data-library -- uvx misatauvx runs the command named after the package; if the README names another one, use uvx --from misata <command>
Config file
{
"mcpServers": {
"misata-synthetic-data-library": {
"command": "uvx",
"args": [
"misata"
]
}
}
}uvx runs the command named after the package; if the README names another one, use uvx --from misata <command>
Config file
{
"servers": {
"misata-synthetic-data-library": {
"type": "stdio",
"command": "uvx",
"args": [
"misata"
]
}
}
}uvx runs the command named after the package; if the README names another one, use uvx --from misata <command>
Platform CLI
gemini mcp add misata-synthetic-data-library uvx misatauvx runs the command named after the package; if the README names another one, use uvx --from misata <command>
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about Misata: synthetic data library
How to install Misata: synthetic data library in Claude Code?
Run claude mcp add misata-synthetic-data-library -- uvx misata in your terminal.
How does Misata ensure math reconciles across rows?
It solves for individual rows using a closed-form conditional sum mechanism derived from the Gamma distribution. This hits target totals exactly rather than approximating them.
What is required to run this MCP server?
The machine must have Python and the uv tool runner installed so that uvx misata can execute locally.
