Dataset Generator

Use this when the user wants to generate, normalize, verify, deduplicate, or export training datasets for Codex, Antigravity, or Claude Code from topics, URLs, reference material, web research, or existing JSONL/CSV files. Supports SFT and DPO workflows, custom export schemas, and deterministic local pipeline scripts.

21 stars 4 forksSoftware Development

Install

skills CLI (Claude Code, Cursor, Codex and more)

npx skills add Bhanunamikaze/AI-Dataset-Generator --skill dataset-generator

Manual install

git clone --depth 1 https://github.com/Bhanunamikaze/AI-Dataset-Generator.git /tmp/AI-Dataset-Generator
mkdir -p ~/.claude/skills
cp -r /tmp/AI-Dataset-Generator/. ~/.claude/skills/dataset-generator

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Dataset Generator

This skill is a tool-native dataset pipeline for Codex, Antigravity, and Claude Code.

- Use the IDE's own tools for browsing, reading, search, and reasoning.
- Use local Python scripts for deterministic normalization, state tracking, verification, deduplication, and export.
- Do not call external LLM-provider APIs as part of this skill.

## Command surface

- `dataset generate "<request>" [--count <n>]`
- `dataset collect "<topic or query>" [--urls url1 url2] [--paths ./dir]`
- `dataset verify <path/to/file>`
- `dataset audit [<path/to/file>]`
- `dataset export --format <openai|huggingface|csv|jsonl|all> [--schema-file path] [--split 0.1]`

If `dataset generate` does not include a size, default to `500` records.
If `dataset collect` does not include `--max-results`, default to `10` results per query.

## Core architecture

- `sub-skills/` contains the cognitive instructions.
- `scripts/` contains deterministic helpers.
- `resources/internal-schema/canonical_schema.json` is the fixed pipeline backbone.
- `resources/target-schemas/` contains preset export profiles.
- `resources/templates/custom_flat_schema.json` is the starting point for custom headers.

## Fixed vs flexible schema

- The canonical internal schema is fixed.
- The final export schema is not universal and must be chosen per user request.
- For custom CSV or flat JSONL headers, create or update a schema file and pass it to `scripts/export.py`.

Read the full skill on GitHub

Details

Skill name: dataset-generator
Repository: Bhanunamikaze/AI-Dataset-Generator
License: MIT
Repo updated: 8 August 2026

FAQ

What does the Dataset Generator skill do?

How do I install the Dataset Generator skill?

skills CLI (Claude Code, Cursor, Codex and more): npx skills add Bhanunamikaze/AI-Dataset-Generator --skill dataset-generator. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Dataset Generator popular?

It lives in Bhanunamikaze/AI-Dataset-Generator, which has 21 GitHub stars and 4 forks.

Related skills

Razbor Servisa

mcdenil-skills/razbor-servisa

Метод-рефлекс: как превратить ЛЮБОЙ сторонний сервис в CLI-инструмент для агента, когда у сервиса нет удобного «разъёма». Скилл сам определяет ситуацию доступа (есть открытый API / API кривой и неполный / API нет вообще) и ведёт по нужной ветке: собрать по докам API, либо подсмотреть скрытый API прямо в браузере — режим разработчика (F12) → вкладка Network → Copy as cURL → сборка CLI под стек проекта → генерация Skill, чтобы агент сам звал команды. Обязательно используй этот скилл, когда задача касается ЧУЖОГО/СТОРОННЕГО сервиса: интеграция, автоматизация или повторение его логики (маркетплейс

Software Development

Publisher

publisher-skill/publisher

Publish HTML frontend projects to the PushWebly publishing platform, list published projects, obtain shareable project URLs, change a published app's public/private visibility, and package published HTML projects into Android APKs. Use when Codex needs to publish a local zip project, publish and build an APK, build an APK for an existing published project, toggle an app between public and private, list a user's published projects, or log in/register for this publishing platform. Codex should use HTTP APIs with curl when remote MCP tools are not mounted.

Software Development

Codex Delegator

eddiearc/codex-delegator

Automatically delegate complex, logic-intensive tasks to OpenAI Codex CLI via `codex exec --full-auto`. Claude Code uses this skill to invoke Codex for complex backend logic, intricate algorithms, or persistent bugs. Enables seamless AI-to-AI collaboration where Claude Code analyzes and Codex executes.

Software Development

Only Skill You Need

Endokelp/Only-Skill-You-Need

The first and only skill a beginner should install. Bootstraps the rest of the agent skill stack for this project, then routes every task to the current best skill for that job. Keeps a dual ledger (global defaults + project overlay) so the same use-case never gets two skills and a worse skill never replaces a better one. This skill should be used when starting a new project, setting up Claude Code / Cursor / Codex / Copilot, asking "what skills should I use", "best skills right now", "bootstrap my skills", "the only skill I need", beginning any coding task where the right skill might already

Software Development

Enterprise Architecture

gauravs19/enterprise-architecture-skill

Unified enterprise & software architecture skill grounded in four open-source standards — C4 + Structurizr DSL, ArchiMate 3.x, TOGAF ADM, and arc42 + ADRs (MADR). Use it whenever the user wants to: produce architecture diagrams as code (C4 context/container/component, Structurizr, PlantUML, Mermaid); write architecture docs or decision records (arc42, ADRs/MADRs, design docs); review or assess a design against quality attributes and EA principles; or model an enterprise (capabilities → applications → technology, app landscapes, portfolio rationalization). Trigger it even when no framework is n

Software Development

Repl Scratchpad

knot0-com/repl-scratchpad

Use instead of chaining 3+ Bash calls, grep/cat/find pipelines, or any multi-step data processing. This skill should be used when the user asks to "start a scratchpad", "use the REPL", "scratchpad mode", "persistent Python session", or when a task involves exploration, analysis, or file processing that would benefit from composing operations in code rather than making individual tool calls. It creates a persistent Python REPL via tmux where variables survive across turns and only print() output enters context — a context-efficient alternative to Bash for data-heavy work.

Software Development

Browse the most popular Claude skills, all software development skills, or write your own with the skills generator.