Get started

Hugging Face for AI agents

Hugging Face is a platform for building, training, and deploying open-source machine learning models, datasets, and spaces. Product and engineering teams use it as a building block inside their own AI features. Connect it once through Arc0, and your agent, or Claude, ChatGPT and Cursor, can use it through one MCP endpoint, limited to what each user approved.

Connect Hugging FaceSee the 135 actions ↓
AI and modelsOAuth 2.0MCP + RESThuggingface.co
AUDIT LOG · HUGGING FACEPOLICY: acme-support
09:41:07 · claude · u_8f2read
hugging_face.list_docs
List Available Documentation✓ allowed · 212ms
09:41:08 · claude · u_8f2read
hugging_face.get_whoami
Get Authenticated User Info✓ allowed · 164ms
09:41:09 · claude · u_8f2write
hugging_face.create_repo
Create Repository✓ approved · approved by user
09:41:12 · claude · u_8f2destructive
hugging_face.delete_spaces_tag
Delete Spaces Tag✕ blocked · policy: deny
EVERY HUGGING FACE CALL, ON THE RECORD
01 · USE CASES

What agents do in Hugging Face.

01

Search for a model or dataset

Search Hugging Face for a model or dataset matching a project's requirements.

02

Check a model's discussion

Pull the discussion thread on a model repository to see recent community feedback.

03

Create a discussion comment

Post a comment on a repository's discussion to ask a question or report an issue.

02 · ACTIONS

135 Hugging Face actions, graded by risk.

Every Hugging Face action is tagged read, write or destructive, so one policy covers the whole app and new actions inherit the right default.

read

85

Look things up. Allowed by default.

  • hugging_face.list_docs
    List Available Documentation
  • hugging_face.get_whoami
    Get Authenticated User Info
  • hugging_face.get_resolve
    Get Resolve
  • hugging_face.list_spaces
    List Spaces on Hugging Face Hub
  • hugging_face.search_docs
    Search Documentation
  • hugging_face.get_trending
    Get Trending Repositories
  • hugging_face.list_vendors
    List Cloud Provider Vendors
  • hugging_face.list_datasets
    List Datasets
  • hugging_face.search_papers
    Search Papers
  • hugging_face.get_discussion
    Get Discussion Details
  • hugging_face.get_model_info
    Get Model Information
  • hugging_face.get_models_jwt
    Get Models JWT
  • hugging_face.get_space_info
    Get Space Info
  • hugging_face.get_spaces_jwt
    Get Spaces JWT
  • hugging_face.list_endpoints
    List Inference Endpoints
  • hugging_face.search_dataset
    Search dataset
+ 69 MORE

write

40

Create and change things. Allow, or ask the user first.

  • hugging_face.create_repo
    Create Repository
  • hugging_face.create_webhook
    Create Webhook
  • hugging_face.create_ask_access
    Request Repository Access
  • hugging_face.create_collection
    Create Collection
  • hugging_face.create_models_tag
    Create Models Tag
  • hugging_face.create_spaces_tag
    Create Spaces Tag
  • hugging_face.create_discussions
    Create Discussion
  • hugging_face.create_datasets_tag
    Create Datasets Tag
  • hugging_face.create_papers_index
    Create Papers Index
  • hugging_face.create_models_branch
    Create models branch
  • hugging_face.create_models_commit
    Create Models Commit
  • hugging_face.create_spaces_branch
    Create spaces branch
  • hugging_face.create_spaces_commit
    Create Spaces Commit
  • hugging_face.create_papers_comment
    Create Paper Comment
  • hugging_face.create_spaces_secrets
    Create or update Space secret
  • hugging_face.update_settings_watch
    Update Watch Settings
+ 24 MORE

destructive

10

Delete, cancel or archive. Ask first, or deny outright.

  • hugging_face.delete_spaces_tag
    Delete Spaces Tag
  • hugging_face.delete_discussions
    Delete discussion
  • hugging_face.delete_datasets_tag
    Delete Dataset Tag
  • hugging_face.delete_notifications
    Delete notifications
  • hugging_face.delete_spaces_branch
    Delete space branch
  • hugging_face.delete_spaces_secrets
    Delete space secret
  • hugging_face.delete_datasets_branch
    Delete dataset branch
  • hugging_face.delete_spaces_variables
    Delete space variable
  • hugging_face.delete_network_cidr_list
    Delete network CIDR list
  • hugging_face.delete_settings_webhooks
    Delete Settings Webhook
03 · HOW IT WORKS

Hugging Face in three steps.

  1. 01Your users connect Hugging FaceThey sign in to Hugging Face on Arc0 Connect, under your brand, and approve the access you ask for.
  2. 02You set the rulesReads run, writes like “create Repository” can wait for the user, and “delete Spaces Tag” can be denied outright.
  3. 03Any agent can actYour agent calls Hugging Face through the Arc0 SDK or MCP, and so can Claude, ChatGPT and Cursor. Every call lands on the audit log.
POLICY.TS
await arc0.policies.set('hugging_face', {
  read: 'allow',
  write: 'ask',        // create_repo
  destructive: 'deny',  // delete_spaces_tag
})

# Claude Code: the same connection, one URL
$ claude mcp add --transport http arc0 \
    https://mcp.arc0.ai/u/u_8f2
04 · AUTH AND DATA

How Hugging Face connects.

Users sign in to Hugging Face on Arc0 Connect and approve the scopes you request. Build with Arc0’s Hugging Face OAuth app, or bring your own so the Hugging Face consent screen names you. Tokens refresh automatically, and you can export them whenever you want.

The same Hugging Face connection serves your agent over MCP and your own backend over REST and the proxy, so a user connects once. How Arc0 handles credentials →

AUTH
API key · OAuth 2.0
CREDENTIALS
Per-tenant encrypted vault
MODEL SEES
Results only, never credentials
AUDIT LOG
Every call, on every plan
05 · WORKS WITH

Use Hugging Face from any agent.

Claude
ChatGPT
Cursor
Codex
VS Code
OpenAI Agents SDK
Claude Agent SDK
Vercel AI SDK
Mastra
LangGraph
06 · RELATED

More ai and models apps.

Perplexity AI9 actionsAffinda119 actionsAgent Mail5 actionsAI/ML API30 actionsAivoov1 actionsAll Images AI7 actionsAltText.ai10 actionsAltTextLab2 actions
ALL AI AND MODELS APPS →
07 · FAQ

Hugging Face and Arc0, answered.

Q01

Can I use Hugging Face with Claude, ChatGPT or Cursor?

Yes. Connect Hugging Face to Arc0 once, then add your Arc0 MCP URL to Claude, ChatGPT, Cursor, Claude Code or any other remote-MCP client. Each assistant only gets the Hugging Face actions you allow.

Q02

How do users connect Hugging Face?

Users sign in to Hugging Face on Arc0 Connect and approve the scopes you request. Build with Arc0’s Hugging Face OAuth app, or bring your own so the Hugging Face consent screen names you. Tokens refresh automatically, and you can export them whenever you want.

Q03

Which Hugging Face actions can my agent take?

135 in total: 85 read, 40 write and 10 destructive, such as “create Repository”. Your policies decide which of them each agent may call.

Q04

Can I stop my agent from deleting things in Hugging Face?

Yes. Actions like “delete Spaces Tag” are graded destructive. Set destructive actions to deny, or to ask so the user approves each one, and blocked calls still show up on the audit log.

Q05

Can my own backend call Hugging Face too?

Yes. The same Hugging Face connection is available over REST and through the Arc0 proxy, so your product and your agent share one connection per user.

Get started

Plug Hugging Face into your agent.

Your users connect Hugging Face once, under your brand. Your agent gets 135 actions behind your policies, with every call on the record.

Connect Hugging FaceBrowse all apps →

Free to build · MCP + REST · Audit log on every plan