Red-Hat-AI-Innovation-TeamRed-Hat-AI-Innovation-TeamOfficialยท1 Agent Skills Included

sdg_hub

Composable pipelines for synthetic training data generation

Builds synthetic training datasets by chaining LLM, parsing, transform, and filtering blocks into YAML-defined flows. Eliminates manual prompt scripting, brittle parsing code, and one-off data pipelines that are hard to reuse. Includes pre-built flows for QA generation, text analysis, and MCP tool-use distillation, with support for 100+ model providers and resumable checkpointed runs.
npx skills add Red-Hat-AI-Innovation-Team/sdg_hub --all -g -y
Available:

Points your AI agent to the right knowledge-base documents for tasks like adding blocks, authoring flows, writing tests, and reviewing code, along with the correct commands to run them.

All Skills in This Repository (1)

Pure Emerald Level Indicators

Frequently Asked Questions

FAQPage Schema
How to install SDG Hub?โ–ผ

Run `npx skills add Red-Hat-AI-Innovation-Team/sdg_hub --all -g -y` in your terminal to install all skills in this suite globally.

How to generate synthetic training data?โ–ผ

Pick a pre-built flow or write your own YAML pipeline, set your model with flow.set_model_config(), then call flow.generate(dataset). Always run flow.dry_run() on a small sample first to validate the pipeline.

Which LLM providers does SDG Hub support?โ–ผ

It supports 100+ providers through LiteLLM, including OpenAI, Anthropic, Azure, Google Gemini, AWS Bedrock, Groq, and local models via vLLM or Ollama.

Can I build custom data generation pipelines?โ–ผ

Yes. You can chain 20+ built-in blocks in a custom YAML flow, write ad-hoc Python scripts, or connect external agent frameworks like Langflow and LangGraph.

Does SDG Hub work with Claude Code and Codex?โ–ผ

Yes. It ships as a plugin with setup-guide, data-generation, and flow-browser skills that run directly inside Claude Code and Codex CLI.

Related Repositories in Data & Analytics

View All in Data & Analyticsโ†’