huggingface/skills provides a collection of standardized Agent Skills that package instructions, scripts, and resources together so AI coding agents can execute Hugging Face Hub operations, ML training, and evaluation. Each skill is a self-contained folder designed for specific AI/ML use cases.
Project overview
By packaging standardized instructions, scripts, and resources into self-contained Agent Skills, huggingface/skills addresses a gap where AI/ML agents previously lacked a uniform way to perform Hugging Face ecosystem tasks such as model training, dataset management, and evaluation.
Deployment
Refer to project documentation
License
Apache-2.0
Best for
Developers and AI engineers who use AI coding agents to perform Hugging Face ecosystem tasks such as model training, dataset management, and evaluation.
Key capabilities
Self-contained folders packaging instructions, scripts, and resources for specific AI/ML use cases (dataset creation, evaluation, training) following the standardized Agent Skills format.
Limitations and risks
Client marketplaces intentionally expose only the hf-cli skill to keep manifests focused; other skills must be added via CLI.
Skills must strictly follow the standardized Agent Skills format with SKILL.md and YAML frontmatter.
Getting started
Install standardized skills through plugin marketplaces or by copying folders; the documented first success path is to register the repository as a plugin marketplace or symlink the folder.
AI Search
Find projects, verify facts, compare options, or turn a complex need into an actionable plan
Try a searchA click only fills the search box; you stay in control