Generate realistic dummy datasets for testing with customizable columns, constraints, and output formats (CSV, JSON, SQL, Python script). Use when creating test data, building mock datasets, or generating sample data for development and demos.
The dummy-dataset skill generates realistic synthetic test data with customizable columns, constraints, and output formats. It solves the everyday development and demo problem of needing believable sample data (customer feedback, transactions, user profiles, and similar) without using real user data, producing either ready-to-run generator scripts or direct data files.
The skill takes parameters for product/system name, dataset type, row count, specific columns, output format, and business constraints, then follows an eight-step process: identify the data domain, define column specifications, set row count, choose an output format, apply realistic value patterns, enforce business constraints and relationships, generate the output, and validate quality. It supports CSV, JSON, SQL INSERT statements, and executable Python generator scripts, and includes a Python template plus a worked customer-feedback example showing skewed rating distributions and conditional rules (for example, bug reports only appearing with low ratings). Deliverables include the generated data or script, validation of constraint compliance, and quick-start usage instructions.
Target users are developers, QA engineers, product managers, and anyone building demos or populating test environments who needs authentic-looking but non-sensitive data. Typical use cases include seeding databases for development, creating mock datasets for prototypes, and producing sample data for presentations. The generated code uses only standard-library operations (csv, json, random, datetime) and writes local files, with no network or credential handling.
CSV for tabular/spreadsheet use, JSON for APIs and NoSQL, SQL INSERT statements for relational databases, and an executable Python generator script for custom or large datasets.
Yes. You can pass constraints such as skewed value distributions and conditional relationships (for example, bug-category rows only with low ratings), which the skill applies during generation.
Row count is a parameter (default 100); the Python-script output is intended for generating custom or large datasets on demand.
No. It produces fully synthetic dummy data intended for testing, development, and demos, which avoids exposing real personal information.
Standard-library modules only (csv, json, random, datetime); it writes a local file such as a CSV and prints how many records were generated.
Quick Setup:
.claude/skills/Repository
phuryn/pm-skills