hf-cli
0The Hugging Face Hub CLI tool `hf` provides commands for interacting with the Hugging Face Hub, replacing the deprecated `huggingface-cli`. It is designed for developers who need a command line interface for downloading, uploading, and managing files in the Hub.
000Updated 2mo ago
CLI Tools
Quick InstallView Source
curl -LsSf https://hf.co/cli/install.sh | bash -s.Overview
About This Skill
The hf-cli skill integrates the Hugging Face Hub CLI directly into your AI agent, enabling full command-line access to Hub operations. Agents can download models and datasets, upload files, manage repositories and buckets, launch Spaces, and execute cloud compute jobs—all through the hf command. The skill is auto-generated from your locally installed CLI, ensuring it stays synchronized with the latest Hub features and commands. This is the recommended first skill to install when giving your agent access to Hugging Face infrastructure.
How to Use
Install using the command: `$ npx skills add https://github.com/huggingface/skills --skill hf-cli`.
Use Cases
Download datasets from the Hugging Face Hub
Manage authentication with the `hf auth` command
Upload files or entire folders to the Hub
Sync local directories with Hub repositories
Tags
Setup & Installation
Quick Install
Terminal
curl -LsSf https://hf.co/cli/install.sh | bash -s.Alternative Install (Git Clone)
git clone https://github.com/huggingface/skillsRequirements
- Claude Code or compatible AI agent
Quick Start Guide
1
Install the Skill
Copy the install command above and run it in your terminal.
2
Open Your AI Agent
Launch Claude Code, Cursor, or your preferred AI coding agent.
3
Try It Out
Use the prompt template or examples below to test the skill.
4
Customize
Adapt the skill to your specific use case and workflow.
Usage Examples
Prompt Template
Use the Hugging Face Hub CLI (`hf`) to [ACTION, e.g., download a model, upload a dataset, list files in a repo]. Replace [REPO_ID], [FILE_PATH], or [OPTIONS] with your specific details. Example: `hf download [REPO_ID] --local-dir [FILE_PATH]` or `hf upload [REPO_ID] --path [FILE_PATH] --repo-type [dataset/model]`.
Example Output
```bash # Downloading a model from the Hugging Face Hub hf download mistralai/Mistral-7B-v0.1 --local-dir ./models/mistral-7b # Uploading a dataset to a Hugging Face repo hf upload my-username/my-dataset --path ./data/processed --repo-type dataset # Listing files in a model repository hf ls t5-small --repo-type model ``` **Output:** ``` Downloading (…) tokenizer.json: 100%|██████████| 1.20M/1.20M [00:01<00:00, 1.12MB/s] Downloading (…) config.json: 100%|██████████| 1.10k/1.10k [00:00<00:00, 1.05kB/s] Downloading pytorch_model.bin: 100%|██████████| 456M/456M [02:30<00:00, 3.02MB/s] Files in `t5-small` (model): - config.json - pytorch_model.bin - spiece.model - tokenizer.json - tokenizer_config.json ```
Find the right skills for your stack
Take a free 3-minute scan and get personalized AI skill recommendations.
Take free scan