Hugging Face Hub Docs
Official Hugging Face Hub docs synced from huggingface/hub-docs/docs.
Use this skill for Hugging Face Hub platform, repository, dataset, Space, inference provider, Jobs, Xet storage, organization, auth, and enterprise questions. For model loading/fine-tuning library APIs, prefer transformers-docs.
Hard Rules
- MUST search
references/ before giving Hub API, token, repository, gated access, billing, Jobs, or inference provider details.
- MUST distinguish Hugging Face Hub platform docs from Transformers library docs.
- MUST call out when guidance depends on repository type: model, dataset, Space, collection, organization, or enterprise.
- NEVER invent endpoint names, token scopes, gating behavior, repository file conventions, or inference provider parameters without checking references.
Fast Lookup
rg -n "token|auth|OAuth|gated|organization|enterprise" skills/huggingface-docs/references
rg -n "model card|dataset card|Space|repository|webhook|collection" skills/huggingface-docs/references
rg -n "Inference Provider|inference endpoint|Jobs|Xet|storage|SageMaker" skills/huggingface-docs/references
Reference Map
references/hub/ — Hugging Face Hub platform docs: repositories, models, datasets, Spaces, cards, tokens, orgs, enterprise, Jobs, billing, webhooks.references/inference-providers/ — Inference Providers docs and provider-specific configuration.references/xet/ — Xet-backed storage and large-file workflows.references/sagemaker/ — SageMaker integration docs.
Workflow
- Identify whether the question is Hub platform, inference provider, Xet/storage, SageMaker, or library-specific.
- Search the matching reference subtree.
- Prefer exact documented terms for repository type, endpoint, token permission, or provider parameter.
- Route Transformers/Diffusers/PyTorch usage questions to the relevant library-specific skill when the Hub docs are not the source of truth.
1---2name: huggingface-docs3description: Use when working with Hugging Face Hub: model/dataset/Space repositories, Hub APIs, tokens, gated models/datasets, organizations, collections, cards, inference providers, jobs, storage/Xet, enterprise features, billing, webhooks, or Hub repository troubleshooting.4---56# Hugging Face Hub Docs78Official Hugging Face Hub docs synced from [`huggingface/hub-docs/docs`](https://github.com/huggingface/hub-docs/tree/main/docs).910Use this skill for Hugging Face Hub platform, repository, dataset, Space, inference provider, Jobs, Xet storage, organization, auth, and enterprise questions. For model loading/fine-tuning library APIs, prefer `transformers-docs`.1112## Hard Rules1314- MUST search `references/` before giving Hub API, token, repository, gated access, billing, Jobs, or inference provider details.15- MUST distinguish Hugging Face Hub platform docs from Transformers library docs.16- MUST call out when guidance depends on repository type: model, dataset, Space, collection, organization, or enterprise.17- NEVER invent endpoint names, token scopes, gating behavior, repository file conventions, or inference provider parameters without checking references.1819## Fast Lookup2021```bash22rg -n "token|auth|OAuth|gated|organization|enterprise" skills/huggingface-docs/references23rg -n "model card|dataset card|Space|repository|webhook|collection" skills/huggingface-docs/references24rg -n "Inference Provider|inference endpoint|Jobs|Xet|storage|SageMaker" skills/huggingface-docs/references25```2627## Reference Map2829- `references/hub/` — Hugging Face Hub platform docs: repositories, models, datasets, Spaces, cards, tokens, orgs, enterprise, Jobs, billing, webhooks.30- `references/inference-providers/` — Inference Providers docs and provider-specific configuration.31- `references/xet/` — Xet-backed storage and large-file workflows.32- `references/sagemaker/` — SageMaker integration docs.3334## Workflow35361. Identify whether the question is Hub platform, inference provider, Xet/storage, SageMaker, or library-specific.372. Search the matching reference subtree.383. Prefer exact documented terms for repository type, endpoint, token permission, or provider parameter.394. Route Transformers/Diffusers/PyTorch usage questions to the relevant library-specific skill when the Hub docs are not the source of truth.
wenerme