Designed and implemented a fully automated AI agent behavior dataset pipeline using a multi-stage architecture (task generation → agent execution → dual labeling → quality validation → HuggingFace publish), with daily GitHub Actions scheduling and zero manual intervention after deployment.

agentic-workflow data-factory reward-model-training supervisor synthetic-dataset-generation
1 Open Issue Need Help Last updated: Jul 11, 2026

Open Issues Need Help

View All on GitHub
Add an AGENT.MD file about 2 hours ago
good first issue

Designed and implemented a fully automated AI agent behavior dataset pipeline using a multi-stage architecture (task generation → agent execution → dual labeling → quality validation → HuggingFace publish), with daily GitHub Actions scheduling and zero manual intervention after deployment.

Python
#agentic-workflow#data-factory#reward-model-training#supervisor#synthetic-dataset-generation