Home Run the test-data engine from a local checkout
Run the test-data engine from a local checkout
Use the command line to write fixtures to files or generate SQL in a development workflow. These instructions use repository source and do not assume an npm release exists.
Source access and license
The source repository is private and the CLI is UNLICENSED. These instructions are for authorized collaborators with repository access; they do not imply public source access or an npm release. Everyone can use the free browser tools. For source access, contact the maintainer.
Source, requirements and license status
Use a supported Node.js release (the package requires Node.js 18 or newer). Run the commands from an authorized checkout of azmiyuksel/fundata; generation needs no npm dependency install. If repository issues are available to you, report reproducible bugs there without credentials, production records or private schemas.
Inspect the checkout and discover the interface
After reviewing the source and applicable terms, obtain a local checkout and run these commands from its root:
git clone https://github.com/azmiyuksel/fundata.git cd fundata node cli/bin/fundata.js --help node cli/bin/fundata.js types node cli/bin/fundata.js templates git rev-parse HEAD
Record the revision printed by the last command. Field IDs and their options come from types; template keys come from templates. Commands below use the checkout path rather than npx fundata-cli, so their behavior does not depend on registry publication.
Write a small fixture and inspect it
node cli/bin/fundata.js generate --field id:row_number --field name:full_name --field email:email_address --rows 3 --seed qa-demo-v1 --format json --out users.json
The output file contains three synthetic users. --out writes a file; without it, the result goes to standard output. If the output path contains directories, create those directories first. Add fields with repeated --field name:type arguments; optional settings use --field age:integer:min=18,max=65.
The command line supports CSV, TSV, JSON, NDJSON, SQL and XML, with a row limit of 100,000. XLSX is a browser export, not a CLI format. Use --schema schema.json for a JSON schema file exported by the builder, or a field array with name, type, opts and blank properties. This is the tool's native schema, not an arbitrary JSON Schema document.
Reproduce data without promising permanent bytes
Keep the schema, seed, output format, SQL options and engine revision together. Set fixed date bounds: some built-in templates use the current date as a default upper bound. A seed fixes randomness, but it does not freeze changing schema defaults or source datasets. For long-lived exact fixtures, keep the reviewed generated file.
The --share option reads an embedded #s= schema payload. It does not fetch a revocable #g= cloud share. Export the native schema from the browser when you need a file-based CI input. Invalid generation inputs are reported on standard error with a nonzero exit status; let that fail the job rather than committing an empty output.
Follow the QA fixture workflow, the database seed workflow, or the local MCP setup. The methodology documents output limits.
Versioned fixtures and the browser API
Export a fixture manifest to carry the engine revision, schema, seed and every output option between the browser, CLI and MCP. Explicit CLI flags override saved options; false values remain false.
node cli/bin/fundata.js manifest --schema schema.json --out fixture.json node cli/bin/fundata.js generate --schema fixture.json --out users.json node cli/bin/fundata.js import --schema openapi.json --choices node cli/bin/fundata.js serve --schema fixture.json --route /api/users --port 3001
The optional mock binds only to 127.0.0.1 and serves a fixed JSON array over GET/HEAD. Stop with Ctrl+C. It does not persist requests or contact upstream services.
The local browser API exposes FD_GENERATOR v1 after loading fixture-contract.js, data.js, app.js and sdk.js in that order. It provides validate, generate, serialize, codegen, manifest, readManifest and saveToFile. Text streaming uses backpressure; direct file saving is offered where the browser supports it. Standard downloads remain available.
Keep fixed date bounds and the recorded engine and reference-data revisions. Named reusable datasets belong to the workspace backup; a fixture manifest does not include those lists. Keep reviewed exported files when historical bytes must survive future engine changes.
Last updated