Hugging Face Hub MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| PORT | No | HTTP server port | 3003 |
| HF_BASE_URL | No | Base URL for Hugging Face API | https://huggingface.co |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Server capabilities have not been inspected yet.
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| hf_list_modelsA | Get information from all models in the Hub. Supports filtering by search terms, authors, tags, and more. Returns paginated results with model metadata including downloads, likes, and tags. |
| hf_get_model_infoC | Get detailed information for a specific model including metadata, files, configuration, and more. |
| hf_get_model_tagsB | Gets all available model tags hosted in the Hub, organized by type (e.g., task types, libraries, languages). |
| hf_list_datasetsB | Get information from all datasets in the Hub. Supports filtering by search terms, authors, tags, and more. Returns paginated results with dataset metadata including downloads, likes, and tags. |
| hf_get_dataset_infoB | Get detailed information for a specific dataset including metadata, files, configuration, and more. |
| hf_get_dataset_parquetB | Get the list of auto-converted parquet files for a dataset. Can specify subset (config) and split to get specific files. |
| hf_get_croissantC | Get the Croissant metadata for a dataset. Croissant is a high-level format for machine learning datasets. |
| hf_get_dataset_tagsB | Gets all available dataset tags hosted in the Hub, organized by type (e.g., task categories, languages, licenses). |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 8 tools
Each tool has a clearly distinct purpose with no overlap: there are separate tools for datasets vs. models, and within each category, tools for getting info, listing all, getting tags, and dataset-specific tools for Croissant metadata and parquet files. The descriptions make the boundaries explicit, preventing misselection.
All tools follow a consistent 'hf_' prefix with verb_noun pattern (e.g., hf_get_dataset_info, hf_list_models). The naming is uniform across the set, using snake_case and clear verbs like 'get', 'list', and 'get' for specific actions, making it predictable and readable.
With 8 tools, this server is well-scoped for interacting with the Hugging Face Hub. It covers core operations for datasets and models (info, listing, tags) plus dataset-specific features like Croissant and parquet files, avoiding bloat while providing essential functionality.
The tool set covers read operations comprehensively for datasets and models, including metadata, listings, tags, and file access. A minor gap is the lack of write operations (e.g., upload or modify datasets/models), but for a read-focused server, it handles the domain well with no dead ends.