Build, train and deploy state of the art models powered by the reference open source in machine learning.
Build, train and deploy state of the art models powered by the reference open source in machine learning. On Nagent, Hugging Face is exposed as a fully-configurable artificial intelligence integration that any agent can call — 135 actions, and API_KEY, OAUTH2 authentication. No code is required to wire Hugging Face into your workflow — connect it once via the External Integrations panel and reuse it across every agent you build.
Agent builders use Hugging Face to automate the kinds of tasks artificial intelligence teams previously handled manually. Concrete examples — each one is a single agent step in Nagent — include:
Every action and trigger is paired with a structured input/output schema (visible in the sections below), so when you wire Hugging Face into Helix — our agentic agent builder — the editor knows exactly what each step expects and produces. Configure once, deploy anywhere across your Nagent agents.
Every operation an agent can call against Hugging Face, with input parameters and output schema. Drop these into any step of an agent built in Helix.
HUGGING_FACE_CHANGE_DISCUSSIONS_STATUSTool to change the status of a Hugging Face repository discussion. Use when you need to open or close discussions on models, datasets, or spaces.
Input parameters
The discussion number or ID to change the status of. This is the numeric identifier of the discussion thread.
The repository name where the discussion exists. For example, 'bert-base-uncased' or 'my-dataset'.
The new status to set for the discussion. Use 'open' to reopen a closed discussion or 'closed' to close an open discussion.
Optional comment to add when changing the status. Useful for explaining why the discussion is being closed or reopened.
The namespace (username or organization) that owns the repository. For example, 'google-bert' or 'huggingface'.
The type of repository where the discussion exists. Must be one of: models, spaces, or datasets.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CHECK_DATASET_VALIDITYTool to check whether a specific dataset is valid on Hugging Face Hub. Use when you need to determine what features (preview, viewer, search, filter, statistics) are available for a dataset.
Input parameters
Name of the dataset to check validity for. Format: 'namespace/repo-name' (e.g., 'rajpurkar/squad', 'huggingface/cifar10').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CHECK_MODELS_UPLOAD_METHODTool to check if files should be uploaded through the Large File mechanism or directly. Use when preparing to upload files to a Hugging Face model repository to determine the appropriate upload method for each file.
Input parameters
The git revision (branch, tag, or commit hash) to upload to.
The name of the model repository.
List of files to check for upload method. Maximum 1000 files.
Content of the .gitignore file for the revision. Optional, otherwise uses existing .gitignore content.
The namespace (user or organization) that owns the model repository.
Content of the .gitattributes file. Provide this if you plan to modify .gitattributes yourself when uploading LFS files. Not needed if relying on automatic LFS detection.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CHECK_SPACES_UPLOAD_METHODTool to check if files should be uploaded through the Large File mechanism or directly to Hugging Face Spaces. Use when preparing to upload files to a Hugging Face Space repository to determine the appropriate upload method for each file.
Input parameters
The git revision (branch, tag, or commit hash) to upload to.
The name of the Space repository.
List of files to check for upload method. Maximum 1000 files.
Content of the .gitignore file for the revision. Optional, otherwise uses existing .gitignore content.
The namespace (user or organization) that owns the Space repository.
Content of the .gitattributes file. Provide this if you plan to modify .gitattributes yourself when uploading LFS files. Not needed if relying on automatic LFS detection.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CLAIM_SETTINGS_PAPERS_CLAIMTool to claim authorship of a paper on Hugging Face. Use when you need to associate yourself or another user with an ArXiv paper.
Input parameters
ArXiv paper identifier being claimed. This should be the ArXiv ID of the paper you want to claim authorship for.
HF user who should receive the claim. Must be a 24-character hexadecimal string representing the Hugging Face user ID. Only provide this if you need to assign the claim to a specific user.
Author entry on the paper being claimed. Must be a 24-character hexadecimal string representing the specific author entry on the paper. Only provide this if you need to specify a particular author entry.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_ASK_ACCESSTool to request access to a gated repository on Hugging Face Hub. Use when you need to submit an access request for models, datasets, or Spaces that require approval. The fields required vary by repository.
Input parameters
The repository name of the gated model, dataset, or Space.
The namespace (organization or user) that owns the gated repository.
The type of repository (models, datasets, or spaces). Defaults to models.
A dictionary of field names and values required by the repository's access form. Common fields include "First Name", "Last Name", "Date of birth" (YYYY-MM-DD format), "Country" (country code), "Affiliation", "Job title", and terms acceptance fields. The exact fields vary by repository and are specified in the repository's gated access configuration.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_COLLECTIONTool to create a new collection on Hugging Face. Use when you need to organize and curate models, datasets, spaces, papers, or other collections into a named collection.
Input parameters
Item to be added to the collection upon creation.
The title of the collection. Must be between 1 and 60 characters.
If not provided, the collection will be public. This field will respect the organization's visibility setting.
The namespace (username or organization) under which to create the collection
Optional description for the collection. Maximum 150 characters.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DATASETS_BRANCHTool to create a new branch in a Hugging Face dataset repository. Use when you need to create a branch for versioning or experimentation with dataset changes.
Input parameters
The name of the new branch to create.
The name of the dataset repository without the namespace prefix.
The namespace (username or organization) that owns the dataset repository.
Overwrite the branch if it already exists. Defaults to false.
Create an empty branch without any files. Defaults to false.
The commit hash, tag, or branch name to start the new branch from. Defaults to 'main' if not specified.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DATASETS_COMMITTool to create a commit in a Hugging Face dataset repository. Use when you need to add, update, or delete files in a dataset. Supports both regular files and Large File Storage (LFS) for large binary files. Can optionally create a pull request instead of directly committing.
Input parameters
Revision (branch name, tag, or commit SHA) to commit to
Name of the dataset repository
List of files to add or update in the commit. Each file requires a path and either content or oldPath.
Summary of the commit (required). This is the commit message title.
Whether to create a pull request from the commit. Set to '1' or 'true' to create a PR.
List of Large File Storage (LFS) files to include in the commit
Namespace (username or organization) that owns the dataset repository
For Spaces, whether to try to hot reload the commit (only for single Python file updates). Set to '1' or 'true' to enable.
Detailed description of the commit (optional). Defaults to empty string.
Parent commit SHA (40-character hexadecimal string). Optional, defaults to HEAD of the revision.
List of files to delete from the repository
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DATASETS_PREUPLOADTool to check if files should be uploaded via Large File Storage (LFS) or directly to a Hugging Face dataset repository. Use before uploading files to determine the correct upload method for each file based on size and repository settings.
Input parameters
Revision (branch or commit) to check preupload for. Typically 'main' for the default branch
Name of the dataset repository
List of files to check for upload method. Maximum 1000 files per request
Content of .gitignore file for the revision. Optional; uses existing .gitignore if not provided
Namespace (username or organization) that owns the dataset repository
Content of .gitattributes file if you plan to modify it yourself. Only needed if managing LFS tracking manually; otherwise HF automatically updates .gitattributes
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DATASETS_TAGTool to create a tag on a Hugging Face dataset repository. Use when you need to mark a specific revision with a named tag.
Input parameters
The revision (branch or commit SHA) to create the tag from, typically 'main'.
The name of the tag to create (e.g., 'v1.0', 'release-2024').
The name of the dataset repository to create a tag for.
Optional message describing the purpose of this tag.
The namespace (username or organization) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DISCUSSIONSTool to create a new discussion on a Hugging Face repository (model, dataset, or Space). Use when you need to start a conversation, report an issue, or create a pull request discussion.
Input parameters
The name of the repository where the discussion will be created.
The title of the discussion. Must be between 3 and 200 characters.
The namespace (username or organization) that owns the repository.
The type of repository: models, spaces, or datasets. This determines where the discussion will be created.
The description/content of the discussion. Can contain markdown formatting. Maximum 65536 characters.
Whether this discussion should be created as a pull request. If true, creates a PR instead of a regular discussion.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DISCUSSIONS_COMMENTTool to create a new comment on a Hugging Face repository discussion. Use when you need to add comments or replies to discussions on models, datasets, or spaces.
Input parameters
The discussion number or ID to comment on. This is the numeric identifier of the discussion thread.
The repository name where the discussion exists. For example, 'gpt2' or 'my-dataset'.
The comment text to post on the discussion. Must be between 1 and 65536 characters.
The namespace (username or organization) that owns the repository. For example, 'openai-community' or 'huggingface'.
The type of repository where the discussion exists. Must be one of: models, spaces, or datasets.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_DISCUSSIONS_PINTool to pin or unpin a discussion on a Hugging Face repository (model, dataset, or Space). Use when you need to highlight important discussions by pinning them to the top of the list, or unpin them when they're no longer priority.
Input parameters
The discussion number/identifier to pin or unpin. This is the numeric ID shown in the discussion URL.
The name of the repository containing the discussion. For example, for 'meta-llama/Llama-2-7b', the repo is 'Llama-2-7b'.
Whether to pin (true) or unpin (false) the discussion. Pinned discussions appear at the top of the discussions list.
The namespace (username or organization) that owns the repository. For example, for 'meta-llama/Llama-2-7b', the namespace is 'meta-llama'.
The type of repository: models, spaces, or datasets. This determines which repository type contains the discussion.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_MODELS_BRANCHTool to create a new branch in a Hugging Face model repository. Use when you need to create a branch for experimenting with model changes, versioning, or creating isolated development environments.
Input parameters
The name for the new branch to be created (e.g., 'feature-branch', 'experiment-v2'). This will be the branch name in the repository.
The name of the model repository where the branch will be created. For example, for model 'meta-llama/Llama-2-7b', the repo is 'Llama-2-7b'.
The namespace (username or organization) that owns the model repository. For example, for model 'meta-llama/Llama-2-7b', the namespace is 'meta-llama'.
Overwrite the branch if it already exists. If False (default), creating a branch that already exists will result in an error.
Create an empty branch without any files. If True, startingPoint is not required. If False (default), startingPoint should be provided.
The commit hash or branch name to start the new branch from (e.g., 'main', 'dev', or a commit hash). Required unless creating an empty branch. Defaults to 'main' if not specified and emptyBranch is False.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_MODELS_COMMITTool to create a commit to a Hugging Face model repository. Use when you need to add, update, or delete files in a model repository. Supports both standard JSON and JSON-lines (NDJSON) formats. JSON-lines format is recommended for better performance.
Input parameters
The git revision (branch, tag, or commit SHA) to commit to
The name of the model repository to commit to
List of files to add or update in the commit
Summary message for the commit. This is a required field that describes the commit
Whether to create a pull request from the commit. Typically '1' or 'true' to create a PR
List of LFS (Large File Storage) files to add or update
The namespace (user or organization) that owns the model repository
For Spaces, whether to try to hot reload the commit (only for single python files updates). Typically '1' or 'true' to enable hot reload
Optional detailed description for the commit. Defaults to empty string if not provided
Content type for the request. Use application/x-ndjson (recommended) for JSON-lines format or application/json for standard JSON
Optional parent commit SHA (40-character hex string). If provided, the commit will be based on this specific commit
List of files to delete from the repository
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_MODELS_TAGTool to create a tag on a Hugging Face model repository. Use when you need to mark a specific revision with a named tag.
Input parameters
The revision (branch or commit SHA) to create the tag from, typically 'main'.
The name of the tag to create (e.g., 'v1.0', 'release-2024').
The name of the model repository to create a tag for.
Optional message describing the purpose of this tag.
The namespace (username or organization) that owns the model.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_PAPERS_COMMENTTool to create a new comment on a Hugging Face paper. Use when you need to add comments or feedback to research papers on Hugging Face.
Input parameters
The comment text to post on the paper. Must be between 1 and 65536 characters.
The unique identifier of the paper to comment on. This can be the paper ID (e.g., '2408.04619') or the full paper path.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_PAPERS_COMMENT_REPLYTool to create a reply to a comment on a Hugging Face paper. Use when you need to respond to an existing comment on a paper discussion.
Input parameters
The content of the reply comment.
The unique identifier of the paper (e.g., '2312.09323').
The unique identifier of the comment to reply to (e.g., '65844ff1ee15e3c7fc034c97').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_PAPERS_INDEXTool to index a paper from arXiv by its ID on Hugging Face. Use when you need to make a paper searchable and accessible on the platform. Note: If the paper is already indexed, only its authors can re-index it.
Input parameters
The arXiv ID of the paper to index (format: YYMM.NNNNN or YYMM.NNNN, e.g., '2411.19876' or '1234.5678'). Only the paper's authors can re-index if already indexed.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_REPOTool to create a new repository (model, dataset, or Space) on Hugging Face Hub. Use when you need to initialize a new repository for uploading models, datasets, or deploying Spaces applications.
Input parameters
SDK options for Spaces.
The name of the repository to create. This will be the repository identifier within the namespace.
The type of repository to create: model, dataset, or space. Defaults to 'model' if not specified.
Whether the repository should be private. If not specified, follows the default visibility setting for your account or organization.
The organization namespace to create the repository under. If not provided, the repository will be created under your personal namespace.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SPACES_BRANCHTool to create a new branch in a Hugging Face space repository. Use when you need to create a branch for experimenting with space changes, versioning, or creating isolated development environments.
Input parameters
The name for the new branch to be created (e.g., 'feature-branch', 'experiment-v2'). This will be the branch name in the repository.
The name of the space repository where the branch will be created. For example, for space 'huggingface/my-space', the repo is 'my-space'.
The namespace (username or organization) that owns the space repository. For example, for space 'huggingface/my-space', the namespace is 'huggingface'.
Overwrite the branch if it already exists. If False (default), creating a branch that already exists will result in an error.
Create an empty branch without any files. If True, startingPoint is not required. If False (default), startingPoint should be provided.
The commit hash or branch name to start the new branch from (e.g., 'main', 'dev', or a commit hash). Required unless creating an empty branch. Defaults to 'main' if not specified and emptyBranch is False.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SPACES_COMMITTool to create a commit in a Hugging Face Space repository. Use when you need to add, update, or delete files in a Space. Supports both JSON and NDJSON (recommended) payload formats for commits.
Input parameters
The revision (branch or commit) to commit to. Typically 'main' for the default branch
The name of the Space repository where the commit will be created. For example, for Space 'user/my-space', the repo is 'my-space'
List of files to add or update in the commit. For JSON payload format only
Summary of the commit (required). This is the commit message title
Whether to create a pull request from the commit instead of committing directly. Set to 'true' to create a PR, 'false' or omit to commit directly
List of Large File Storage (LFS) files to add or update. For JSON payload format only
For Spaces, whether to try to hot reload the commit (only works for single Python file updates). Set to 'true' to enable hot reload
The namespace (username or organization) that owns the Space. For example, for Space 'user/my-space', the namespace is 'user'
Content type for the commit request.
Detailed description of the commit (optional, defaults to empty string). This is the commit message body
The parent commit SHA (40-character hexadecimal). Optional, defaults to current HEAD of the branch
List of file paths to delete in the commit. For JSON payload format only
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SPACES_SECRETSTool to create or update a secret in a Hugging Face Space. Use when you need to add or update environment variables or sensitive configuration values for a Space. This action upserts the secret, meaning it will create a new secret if it doesn't exist or update it if it already exists.
Input parameters
The name of the secret key. Must start with a letter and contain only letters, digits, and underscores (pattern: ^\[a-zA-Z\]\[_a-zA-Z0-9\]*$). For example: 'API_KEY', 'DATABASE_URL', 'TEST_SECRET_KEY'.
The name of the Space repository where the secret will be created or updated. For example, for Space 'huggingface/my-space', the repo is 'my-space'.
The value of the secret. This will be stored securely and not visible after creation. Defaults to empty string if not provided.
The namespace (username or organization) that owns the Space. For example, for Space 'huggingface/my-space', the namespace is 'huggingface'.
Optional description of the secret to help identify its purpose. For example: 'API key for external service', 'Database connection string'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SPACES_TAGTool to create a tag on a Hugging Face space repository. Use when you need to mark a specific revision with a named tag.
Input parameters
The revision (branch or commit SHA) to create the tag from, typically 'main'.
The name of the tag to create (e.g., 'v1.0', 'release-2024').
The name of the space repository to create a tag for.
Optional message describing the purpose of this tag.
The namespace (username or organization) that owns the space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SPACES_VARIABLESTool to create or update a variable in a Hugging Face Space. Use when you need to add or update environment variables or configuration values for a Space. This action upserts the variable, meaning it will create a new variable if it doesn't exist or update it if it already exists.
Input parameters
The name of the variable key. Must start with a letter and contain only letters, digits, and underscores (pattern: ^\[a-zA-Z\]\[_a-zA-Z0-9\]*$). For example: 'APP_MODE', 'DEBUG_ENABLED', 'TEST_VAR_123'.
The name of the Space repository where the variable will be created or updated. For example, for Space 'huggingface/my-space', the repo is 'my-space'.
The value of the variable. This will be stored and can be accessed by the Space. Defaults to empty string if not provided.
The namespace (username or organization) that owns the Space. For example, for Space 'huggingface/my-space', the namespace is 'huggingface'.
Optional description of the variable to help identify its purpose. For example: 'Environment mode setting', 'Debug flag for development'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_SQL_CONSOLE_EMBEDTool to create a SQL Console embed for querying datasets on Hugging Face. Use when you need to create a shareable SQL query interface for exploring dataset splits. The embed allows users to execute SQL queries against dataset views (e.g., train, test, validation splits).
Input parameters
The SQL query to be embedded in the console. This is the query users will see and can execute.
The name of the dataset repository without the namespace prefix.
Title for the SQL console embed. Maximum 200 characters.
List of available views (splits) in the dataset that can be queried. Must contain at least one view.
Whether the SQL console embed should be private. If true, only authorized users can access it.
The type of repository. Currently only 'datasets' is supported.
The namespace (username or organization) that owns the dataset repository.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_CREATE_WEBHOOKTool to create a webhook on Hugging Face that triggers on repository or discussion events. Use when you need to receive notifications for changes to specific models, datasets, or spaces.
Input parameters
The webhook URL endpoint that will receive POST requests when events occur. Must be a valid HTTP/HTTPS URL.
Optional secret string used to sign webhook payloads for verification. Must contain only printable ASCII characters (spaces and characters from 0x20 to 0x7F).
List of event domains to monitor. At least one domain must be specified. Use 'repo' for repository events or 'discussion' for discussion events.
List of repositories to watch for changes. At least one repository must be specified. Events from these repositories will trigger webhook calls.
Optional job source identifier for associating the webhook with a specific job source.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_DATASETS_BRANCHTool to delete a branch from a Hugging Face dataset repository. Use when you need to remove a branch that is no longer needed. This action permanently removes the specified branch from the dataset.
Input parameters
The name of the branch to delete from the dataset repository. This should be a valid branch name (not the main/master branch in most cases).
The name of the dataset repository from which to delete the branch. This is the repository identifier without the namespace.
The namespace (user or organization) that owns the dataset. For example, '121tester' for a user or 'huggingface' for an organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_DATASETS_TAGTool to delete a tag from a Hugging Face dataset. Use when you need to remove a specific tag revision from a dataset repository.
Input parameters
The tag revision to delete from the dataset.
The repository name of the dataset.
The namespace (organization or user) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_DISCUSSIONSTool to delete a discussion from a Hugging Face repository. Use when you need to remove a discussion that is no longer needed. This action permanently removes the specified discussion from the repository.
Input parameters
The discussion number to delete. This is the unique identifier of the discussion within the repository.
The name of the repository containing the discussion to delete. This is the repository identifier without the namespace.
The namespace (user or organization) that owns the repository. For example, '121tester' for a user or 'huggingface' for an organization.
The type of repository where the discussion exists. Must be one of: models, spaces, or datasets.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_NETWORK_CIDR_LISTTool to delete a network CIDR list entry from Hugging Face Inference Endpoints. Use when you need to remove a CIDR configuration that is no longer needed. This action permanently removes the specified CIDR from the namespace's network configuration.
Input parameters
The unique identifier of the CIDR to delete. This is a hexadecimal ID that was returned when the CIDR was created.
The namespace (user or organization) that owns the CIDR configuration. For example, '121tester' for a user or 'my-org' for an organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_NOTIFICATIONSTool to delete notifications from Hugging Face. Use when you need to remove notifications either by specific discussion IDs or by applying filters to delete multiple notifications at once. Supports targeted deletion (via discussion_ids) or bulk deletion (via applyToAll with filter parameters).
Input parameters
Page number for pagination when filtering notifications. Defaults to 0 (first page).
Enum for mention filter.
Filter notifications by paper ID. Use this to delete notifications related to a specific paper. Only applies when applyToAll is true.
Filter notifications by repository name. Use this to delete notifications from a specific repository. Format should be 'namespace/repo-name'. Only applies when applyToAll is true.
Enum for repository type filter.
Filter notifications by article ID. Use this to delete notifications related to a specific article. Only applies when applyToAll is true.
Filter notifications by last update timestamp. Use this to delete notifications updated before or after a specific time. Only applies when applyToAll is true.
Filter notifications by the username of the post author. Use this to delete notifications from a specific user. Only applies when applyToAll is true.
Enum for read status filter.
When true, applies the deletion to all notifications matching the filter criteria (readStatus, repoType, repoName, etc.). Use with caution as this will delete multiple notifications at once. Required when not using discussion_ids.
Array of specific discussion IDs to delete. Each ID must be exactly 24 hexadecimal characters. Use this for targeted deletion of specific notifications. If provided, other filter parameters are ignored.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_SETTINGS_WEBHOOKSTool to delete a webhook from Hugging Face settings. Use when you need to remove a webhook configuration that is no longer needed.
Input parameters
The unique identifier of the webhook to delete. Must be a 24-character hexadecimal string.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_SPACES_BRANCHTool to delete a branch from a Hugging Face space repository. Use when you need to remove a branch that is no longer needed. This action permanently removes the specified branch from the space.
Input parameters
The name of the branch to delete from the space repository. This should be a valid branch name (not the main branch in most cases).
The name of the space repository from which to delete the branch. This is the repository identifier without the namespace.
The namespace (user or organization) that owns the space. For example, '121tester' for a user or 'huggingface' for an organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_SPACES_SECRETSTool to delete a secret from a Hugging Face space. Use when you need to remove sensitive credentials or configuration values that are no longer needed. This action permanently removes the specified secret from the space's environment variables.
Input parameters
The key name of the secret to delete from the space. This should be the exact name of the secret as it appears in the space's settings.
The name of the space repository from which to delete the secret. This is the repository identifier without the namespace.
The namespace (user or organization) that owns the space. For example, '121tester' for a user or 'huggingface' for an organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_SPACES_TAGTool to delete a tag from a Hugging Face space. Use when you need to remove a specific tag revision from a space repository.
Input parameters
The tag revision to delete from the space.
The repository name of the space.
The namespace (organization or user) that owns the space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_DELETE_SPACES_VARIABLESTool to delete a variable from a Hugging Face space. Use when you need to remove configuration values or environment variables that are no longer needed. This action permanently removes the specified variable from the space's environment.
Input parameters
The key name of the variable to delete from the space. This should be the exact name of the variable as it appears in the space's settings.
The name of the space repository from which to delete the variable. This is the repository identifier without the namespace.
The namespace (user or organization) that owns the space. For example, '121tester' for a user or 'huggingface' for an organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_FILTER_DATASET_ROWSTool to filter rows in a Hugging Face dataset split based on SQL-like query conditions. Use when you need to search or filter specific rows from a dataset based on column values, or to retrieve sorted subsets of data.
Input parameters
Name of the dataset split to filter (e.g., 'train', 'test', 'validation').
SQL WHERE clause for filtering rows. Use column names and comparison operators (e.g., 'label=0', 'score>0.5', 'text LIKE "%happy%"'). Multiple conditions can be combined with AND/OR.
Name of the dataset configuration/subset to filter. Use 'default' for datasets without explicit configs.
Maximum number of rows to return. Defaults to 100. Maximum allowed is 100.
Number of rows to skip before returning results. Use with length for pagination.
Full name of the dataset in the format 'namespace/dataset-name'. Example: 'cornell-movie-review-data/rotten_tomatoes'.
SQL ORDER BY clause for sorting results. Specify column name and optionally 'ASC' or 'DESC' (e.g., 'score DESC', 'date ASC').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GENERATE_CHAT_COMPLETIONTool to generate a response given a list of messages in a conversational context. Supports both conversational Language Models (LLMs) and Vision-Language Models (VLMs). Compatible with OpenAI SDK.
Input parameters
Random seed for deterministic sampling. Using the same seed with the same parameters should produce the same output.
Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.
The model ID to use for chat completion. Format: namespace/model or namespace/model:provider. Examples: 'meta-llama/Llama-3.2-3B-Instruct', 'Qwen/Qwen2.5-7B-Instruct-1M'.
A list of tools the model may call. Currently, only functions are supported as a tool. Use this to provide a list of functions the model may generate JSON inputs for.
An alternative to sampling with temperature, called nucleus sampling. The model considers the results of tokens with top_p probability mass. So 0.1 means only tokens comprising the top 10% probability mass are considered.
Whether to stream the response as Server-Sent Events. If true, tokens are returned as they are generated. If false (default), the full response is returned at once.
Whether to return log probabilities of the output tokens. If true, returns the log probabilities of each output token returned in the content of message.
A list of messages comprising the conversation so far. Each message has a role (system, user, assistant, tool) and content.
The maximum number of tokens that can be generated in the chat completion. If not specified, the model will generate until it reaches a natural stopping point.
Sampling temperature between 0 and 2. Higher values like 0.8 make output more random, lower values like 0.2 make it more focused and deterministic. We generally recommend altering this or top_p but not both.
Controls which (if any) tool is called by the model. 'auto' (default) means the model can pick between generating a message or calling a tool. 'none' means the model will not call a tool. 'required' means the model must call a tool. Or specify a particular function to force the model to call it.
A prompt to be appended before the tools section.
An integer between 0 and 5 specifying the number of most likely tokens to return at each token position, each with an associated log probability. logprobs must be set to true if this parameter is used.
Stream options.
The format of the response. Can be 'text' (default), 'json_object' for JSON mode, or 'json_schema' for structured output with a specific schema.
Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GENERATE_EMBEDDINGSTool to convert text into vector embeddings for feature extraction, semantic search, and similarity tasks. Use when you need numerical representations of text for ML applications.
Input parameters
The Hugging Face model ID to use for generating embeddings. Must be a valid embedding model from the Hugging Face model hub.
Array of text strings to convert into embeddings. Each string will be processed and returned as a numerical vector representation.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DAILY_PAPERSTool to retrieve daily papers from Hugging Face. Use when you need to fetch the latest AI/ML research papers shared on Hugging Face.
Input parameters
Page number for pagination (0-indexed). Defaults to 0.
Filter papers by specific date (YYYY-MM-DD format).
Sort option for daily papers.
Filter papers by specific week.
Maximum number of papers to return per page. Defaults to 50.
Filter papers by specific month (YYYY-MM format).
Filter papers by submitter username.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_CROISSANTTool to get Croissant metadata about a Hugging Face dataset. Croissant is a metadata format built on schema.org aimed at describing datasets used for machine learning. Use when you need structured metadata in JSON-LD format.
Input parameters
Full dataset name in format 'namespace/repo' (e.g., 'ibm-research/duorc', 'squad', 'huggingface/transformers').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_FIRST_ROWSTool to get the first 100 rows of a dataset split along with column data types and features. Use when you need to preview or sample dataset content.
Input parameters
Name of the dataset split to retrieve rows from.
Name of the dataset configuration/subset. Use 'default' for datasets without configurations.
Name of the dataset in format 'namespace/name'. For example, 'cornell-movie-review-data/rotten_tomatoes' or 'squad'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_INFOTool to get general information about a dataset including description, citation, homepage, license, and features (column schemas). Use when you need to understand dataset structure, available splits, and metadata before working with the data.
Input parameters
Name of the config/subset to retrieve information for. If not provided, returns info for the default configuration.
Name of the dataset to retrieve information for. Format: 'namespace/repo-name' (e.g., 'rajpurkar/squad', 'ibm/duorc').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_REPO_INFOTool to retrieve detailed information about a Hugging Face dataset repository. Use when you need metadata, card data, tags, downloads, likes, configurations, or other information about a specific dataset.
Input parameters
Dataset repository ID in the format author/dataset-name (e.g., 'rajpurkar/squad', 'stanfordnlp/imdb')
Git revision (branch, tag, or commit SHA) to retrieve information for. If not specified, defaults to the main branch
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_ROWSTool to retrieve a slice of rows from a Hugging Face dataset split at any given location (offset). Returns up to 100 rows at a time with complete feature type information and no truncation. Use when you need to inspect specific rows from a dataset without downloading the entire dataset.
Input parameters
Name of the dataset split to retrieve rows from. Common splits include 'train', 'test', and 'validation'.
Name of the config/subset of the dataset. Each dataset may have multiple configurations for different variations or subsets.
Number of rows to retrieve. Maximum allowed value is 100. Use smaller values for faster responses.
Starting row index (0-based) from which to begin retrieving rows. Use 0 to start from the beginning of the dataset.
Name of the dataset to retrieve rows from. Use format 'namespace/dataset_name' (e.g., 'stanfordnlp/imdb', 'ibm/duorc').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_COMPARETool to get a comparison (diff) between two revisions of a Hugging Face dataset. Use when you need to see what changed between dataset versions or commits.
Input parameters
Whether to return the raw diff output. If false or not provided, returns formatted diff.
The name of the dataset repository to compare.
The comparison specification in the format 'base..head' (e.g., 'main~1..main' for comparing the previous commit to main, or 'abc123..def456' for comparing specific commits).
The namespace (username or organization) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_SIZETool to get the size of a Hugging Face dataset including number of rows and size in bytes. Use when you need to determine dataset size, memory requirements, or storage needs for a specific dataset.
Input parameters
Name of the dataset to get size information for. Format: 'namespace/repo-name' (e.g., 'stanfordnlp/imdb', 'ibm/duorc').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_JWTTool to generate a JWT token for accessing a Hugging Face dataset repository. Use when you need authenticated access to datasets, optionally with write access for spaces in dev mode, custom expiration, or encryption.
Input parameters
The name of the dataset repository.
Enable write access for spaces in dev mode. Set to 'true' to enable write access.
Request an encrypted JWT token. Set to 'true' to receive encrypted token with key ID.
The namespace (organization or user) that owns the dataset repository.
Custom expiration time for the JWT token. Format depends on Hugging Face API specifications.
Enable inference API access. Set to 'true' to enable.
Include Pro status information in the token. Set to 'true' to include.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_LEADERBOARDTool to retrieve evaluation results ranked by score for a dataset's leaderboard. Use when you need to compare model performance on a specific dataset or task. Returns an array of leaderboard entries with model information, scores, and rankings.
Input parameters
The repository name of the dataset. For example, 'results' for the dataset 'open-llm-leaderboard/results'.
Optional task identifier to filter leaderboard results by a specific task. If not provided, returns results for all tasks.
The namespace (organization or user) that owns the dataset. For example, 'open-llm-leaderboard' for the dataset 'open-llm-leaderboard/results'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_NOTEBOOKTool to get a Jupyter notebook URL from a Hugging Face dataset repository. Use when you need to retrieve the URL for a specific .ipynb file from a dataset at a particular revision.
Input parameters
The git revision (branch, tag, or commit hash) to retrieve the notebook from.
The path to the notebook file within the repository, including the .ipynb extension.
The repository name of the dataset.
The namespace (user or organization) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_RESOLVETool to resolve and download a file from a Hugging Face dataset repository. This endpoint requires following redirections to retrieve file content or returns XET file info when Accept header is set appropriately. Use when you need to access files from datasets with proper redirection handling.
Input parameters
The git revision (commit SHA, branch name, or tag) to resolve the file from.
The path to the file within the repository (wildcard path parameter).
The name of the dataset repository.
The range in bytes of the file to download (e.g., 'bytes=0-1023' for first 1KB).
Returns json information about the XET file info if the file is a XET file. Use 'application/vnd.xet-fileinfo+json' to get XET metadata, otherwise leave empty to download file content.
The namespace (organization or user) that owns the dataset repository.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_SCANTool to retrieve the security scan status of a Hugging Face dataset repository. Use when you need to check for malware, pickle vulnerabilities, or other security issues in a dataset.
Input parameters
The name of the dataset repository to scan
The namespace (organization or user) that owns the dataset repository
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_TAGS_BY_TYPETool to retrieve all possible tags used for datasets on Hugging Face, grouped by tag type. Use when you need to discover available dataset classification tags, filter options, or metadata categories. Optionally restrict results to a single tag type.
Input parameters
Valid tag type values for filtering dataset tags.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASET_STATISTICSTool to get comprehensive statistics about a dataset split including column statistics and data distribution information. Use when you need to analyze dataset composition, understand data distributions, or get statistical summaries of dataset features.
Input parameters
Name of the split to get statistics for (e.g., 'train', 'test', 'validation').
Name of the configuration/subset of the dataset (e.g., 'mnist', 'cola', 'default').
Name of the dataset. Format: 'namespace/dataset-name' (e.g., 'ylecun/mnist', 'nyu-mll/glue').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_TREESIZETool to get the total size of a Hugging Face dataset repository at a specific revision and path. Use when you need to determine storage requirements or track repository size changes.
Input parameters
The Git revision or branch name to query (e.g., 'main', 'dev', or a commit SHA).
The path within the repository to calculate size for. Use '.' for root directory or specify a subdirectory path (e.g., 'data/train'). The size is calculated recursively for all files under this path.
The repository name of the dataset (e.g., 'gsm8k', 'wikitext').
The namespace or organization name that owns the dataset (e.g., 'squad', 'openai').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DATASETS_XET_READ_TOKENTool to get a read short-lived access token for XET from Hugging Face datasets. Use when you need temporary read access to dataset content through XET protocol.
Input parameters
The revision (branch, tag, or commit) to get the token for. Use 'main' for the default branch.
The repository name of the dataset. For example, 'squad' for the SQuAD dataset.
The namespace (organization or user) that owns the dataset. For example, 'rajpurkar' for the SQuAD dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_DISCUSSIONTool to get detailed information about a specific discussion or pull request on Hugging Face Hub. Use when you need to retrieve all comments, status changes, events, and for PRs, the diff information.
Input parameters
Repository ID in the format author/repo-name (e.g., 'google-bert/bert-base-uncased').
The type of repository: models, spaces, or datasets. This determines where to look for the discussion.
Discussion number (strictly positive integer). This is the unique identifier of the discussion within the repository.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_JOBS_HARDWARETool to retrieve available hardware configurations for Hugging Face Jobs with their specifications and pricing. Use when you need to discover compute options for running jobs.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODEL_INFOTool to retrieve detailed information about a Hugging Face model repository. Use when you need comprehensive metadata including downloads, likes, tags, configuration, files, and more.
Input parameters
The repository name of the model. For example, 'bert-base-uncased' in 'google-bert/bert-base-uncased'.
The namespace (organization or user) that owns the model repository. For example, 'google-bert' in 'google-bert/bert-base-uncased'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_COMPARETool to compare two revisions of a Hugging Face model repository. Returns a git diff showing file changes between commits. Use when you need to see what changed between model versions.
Input parameters
If true, returns raw diff output. If false or omitted, returns formatted diff.
The repository name of the model (e.g., 'bert-base-uncased', 'llama-2-7b').
Comparison specification in the format 'commit1..commit2' (two dots). Each commit is a full Git commit hash. For example: '5546055f03398095e385d7dc625e636cc8910bf2..86b5e0934494bd15c9632b12f734a8a67f723594'. Do NOT use three dots (...) as this format is not supported by the API.
The namespace or organization name that owns the model (e.g., 'google-bert', 'meta-llama', 'openai').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_JWTTool to generate a JWT token for accessing a Hugging Face model repository. Use when you need authenticated access to models, with optional write access for spaces in dev mode, custom expiration, and encryption support.
Input parameters
The repository name within the namespace.
Enable write access for spaces in dev mode. Set to 'true' to enable write access.
Request an encrypted token. Set to 'true' to receive an encrypted JWT token and key ID.
The namespace (organization or user) that owns the model repository.
Custom expiration time for the JWT token. Specify the duration or timestamp for token validity.
Enable inference API access with the token. Set to 'true' to enable.
Include PRO subscription status in the token. Set to 'true' to include PRO status information.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_NOTEBOOKTool to retrieve a Jupyter notebook URL from a Hugging Face model repository. Use when you need to access or display a notebook file stored in a model repository.
Input parameters
The revision, branch name, tag, or commit hash to retrieve the notebook from (e.g., 'main', 'dev', or a commit SHA).
The path to the Jupyter notebook file within the repository, including the .ipynb extension (e.g., 'Untitled.ipynb' or 'notebooks/demo.ipynb').
The repository name of the model (e.g., 'example-huggingface-model').
The namespace or organization name that owns the model repository (e.g., 'akshat-shethia').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_SCANTool to retrieve the security scan status of a Hugging Face model repository. Use when you need to check if a model has been scanned for security issues and view any detected problems.
Input parameters
The repository name of the model. For example, 'opt-125m' in 'facebook/opt-125m'.
The namespace (organization or user) that owns the model repository. For example, 'facebook' in 'facebook/opt-125m'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_TREESIZETool to get the total size of a Hugging Face model repository at a specific revision and path. Use when you need to determine storage requirements or track repository size changes.
Input parameters
The Git revision or branch name to query (e.g., 'main', 'dev', or a commit SHA).
The path within the repository to calculate size for. Use '.' for root directory or specify a subdirectory path (e.g., 'weights', 'config'). The size is calculated recursively for all files under this path.
The repository name of the model (e.g., 'bert-base-uncased', 'llama-2-7b').
The namespace or organization name that owns the model (e.g., 'google-bert', 'facebook', 'openai').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODELS_XET_READ_TOKENTool to retrieve a short-lived XET read access token for a Hugging Face model repository. Use when you need to access XET (eXtensible Tensor) data for a specific model revision.
Input parameters
The revision (branch, tag, or commit hash) to get the XET read token for. Common values include 'main', 'master', or specific commit hashes.
The repository name of the model. For example, 'bert-base-uncased' in 'google-bert/bert-base-uncased'.
The namespace (organization or user) that owns the model repository. For example, 'google-bert' in 'google-bert/bert-base-uncased'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_MODEL_TAGS_BY_TYPETool to retrieve all possible tags used for Hugging Face models, grouped by tag type. Use when you need to discover available model tags for filtering or categorization. Optionally restrict results to a specific tag type.
Input parameters
Enum for model tag types on Hugging Face.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_ORGANIZATIONS_AVATARTool to retrieve the avatar URL for a Hugging Face organization. Use when you need to get the avatar image URL for a specific organization.
Input parameters
The name of the organization. For example, 'huggingface' or 'meta-llama'.
If provided, redirect to the avatar URL instead of returning it as JSON. Set to any value to enable redirection.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_ORGANIZATIONS_MEMBERSTool to retrieve a list of members for a Hugging Face organization. Use when you need to discover who belongs to an organization, with optional filtering by search terms, email, and pagination support.
Input parameters
Name of the organization to retrieve members from. This is the organization's username or slug on Hugging Face.
Filter members by email address. This may require admin permissions depending on the organization's settings.
Maximum number of members to return. Defaults to 500.
Pagination cursor for fetching the next page of results. Use the cursor from the previous response to get more members.
Search query to filter members by username or full name. Use this to find specific members within the organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_ORGANIZATIONS_SOCIALSTool to retrieve an organization's social media handles from Hugging Face. Use when you need to find an organization's GitHub, LinkedIn, or Twitter/X profiles. Only returns handles that the organization has publicly shared on their Hugging Face profile.
Input parameters
The name of the organization to retrieve social media handles for. This is the organization's username on Hugging Face.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_RESOLVETool to resolve a file in a Hugging Face repository. Use when you need to access files from model, dataset, or space repositories. This endpoint follows redirections (302, 307) to retrieve the actual file. When Accept header is set to 'application/vnd.xet-fileinfo+json', returns JSON file metadata instead of redirecting to file content.
Input parameters
The revision (branch, tag, or commit hash) to resolve. Common values include 'main' or specific Git commit SHAs.
The file path within the repository to resolve. For example, 'config.json', 'pytorch_model.bin', or 'tokenizer.json'.
The repository name within the namespace. For example, 'bart-large' in 'facebook/bart-large'.
The range in bytes of the file to download. Use HTTP Range header format like 'bytes=0-1023' to download only specific byte ranges.
Returns JSON information about the XET file info if the file is a XET file. Set to 'application/vnd.xet-fileinfo+json' to receive JSON file metadata instead of following redirect to file content.
The namespace (organization or user) that owns the repository. For example, 'facebook' in 'facebook/bart-large'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_RESOLVE_CACHE_DATASETSTool to resolve a file from cache in a Hugging Face dataset repository. This endpoint follows redirections (302, 307) to retrieve file content or returns XET file info when Accept header is set appropriately. Use when you need to access cached files from datasets with proper redirection handling.
Input parameters
The git revision (commit SHA, branch, or tag) to resolve the file from. Note: This endpoint typically requires an exact commit SHA rather than branch names.
The path to the file within the repository (wildcard path parameter).
The name of the dataset repository.
The range in bytes of the file to download (e.g., 'bytes=0-1023' for first 1KB).
Returns json information about the XET file info if the file is a XET file. Use 'application/vnd.xet-fileinfo+json' to get XET metadata, otherwise leave empty to download file content.
The namespace (organization or user) that owns the dataset repository.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_RESOLVE_CACHE_MODELSTool to resolve and retrieve files from the Hugging Face model cache. Use when you need to access model configuration files, tokenizer files, or other JSON metadata files from a specific model repository. This endpoint returns the actual file content as JSON for JSON files (e.g., config.json, tokenizer.json, tokenizer_config.json).
Input parameters
The revision to resolve. Must be a full Git commit SHA (40 characters). Branch names like 'main' or tags are not supported by this endpoint.
The file path within the repository to resolve. For example, 'config.json', 'pytorch_model.bin', or 'tokenizer.json'.
The repository name of the model. For example, 'bert-base-uncased' in 'google-bert/bert-base-uncased'.
The range in bytes of the file to download. Use HTTP Range header format like 'bytes=0-1023'.
Optional Accept header to specify the desired response format. Common values include 'application/json', 'application/vnd.xet-fileinfo+json', or '*/*'.
The namespace (organization or user) that owns the model repository. For example, 'google-bert' in 'google-bert/bert-base-uncased'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_RESOLVE_CACHE_SPACESTool to resolve and retrieve a file from Hugging Face Spaces cache. Use when you need to download a file from a Space repository or get XET file information. This endpoint follows redirections (HTTP 302/307) to resolve the final file location.
Input parameters
The git revision (branch, tag, or commit hash) to retrieve the file from. Use a full commit SHA for exact version, or 'main' for the default branch.
The path to the file within the repository. This is a wildcard path parameter that can include subdirectories and filename.
The repository name of the Space. For example, 'gradio-user-history' for the Wauplin/gradio-user-history Space.
The range in bytes of the file to download. Format: 'bytes=start-end'. If not provided, the entire file will be retrieved.
Accept header to specify the response format. Use 'application/vnd.xet-fileinfo+json' to get JSON information about XET file info instead of the file content.
The namespace (user or organization) that owns the Space. For example, 'Wauplin' for the Wauplin/gradio-user-history Space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SETTINGS_BILLING_USAGE_JOBSTool to retrieve Jobs usage and billing information for the current subscription period from Hugging Face. Use when you need to check compute usage, costs, or job execution details for the authenticated user or organization.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SETTINGS_BILLING_USAGE_LIVETool to retrieve live billing usage stream from Hugging Face. Use when you need real-time updates on storage, inference, Zero GPU usage, and rate limits.
Input parameters
Maximum time in seconds to wait for live usage updates. Defaults to 5 seconds.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SETTINGS_BILLING_USAGE_V2Tool to retrieve user billing usage for a custom date range from Hugging Face. Use when you need to check usage statistics or resource consumption between specific dates using Unix timestamps.
Input parameters
End date of the billing period as Unix timestamp (seconds since epoch). Defines the end of the date range for which to retrieve usage data. Must be after start_date.
Start date of the billing period as Unix timestamp (seconds since epoch). Defines the beginning of the date range for which to retrieve usage data.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SETTINGS_MCPTool to retrieve MCP (Model Context Protocol) tools configuration for the authenticated user. Use when you need to discover available built-in tools and space-based tools configured in the user's Hugging Face settings.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SETTINGS_WEBHOOKSTool to retrieve a specific webhook configuration from Hugging Face settings. Use when you need to inspect webhook details, verify webhook status, or check webhook configuration for a given webhook ID.
Input parameters
The unique identifier of the webhook to retrieve. Must be a 24-character hexadecimal string.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACE_INFOTool to retrieve detailed information about a Hugging Face Space repository. Use when you need metadata, SDK type, hardware configuration, runtime status, or other information about a specific Space.
Input parameters
Space repository ID in the format author/space-name (e.g., 'Qwen/Qwen3-TTS', 'stabilityai/stable-diffusion')
Git revision (branch, tag, or commit SHA) to retrieve information for. If not specified, defaults to the main branch
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_COMPARETool to compare two revisions of a Hugging Face Space repository. Returns a git diff showing file changes between commits. Use when you need to see what changed between Space versions.
Input parameters
If true, returns raw diff output. If false or omitted, returns formatted diff.
The repository name of the Space (e.g., 'Qwen3-TTS', 'chatgpt-demo').
Comparison specification in the format 'commit1..commit2' (two dots). Each commit is a full Git commit hash. For example: 'bb80b9adea7bc1d818bc635e839db60b7d4aa8f1..8a132844625e28e09f36427c30070276dfd9b2ed'. Do NOT use three dots (...) as this format is not supported by the API.
The namespace or organization name that owns the Space (e.g., 'Qwen', 'openai', 'gradio').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_EVENTSTool to stream status updates for a Hugging Face Space using SSE protocol. Use when you need to monitor Space build stages, runtime status, or receive real-time updates about Space state changes.
Input parameters
The repository name of the Space.
Maximum time in seconds to wait for status updates. Defaults to 5 seconds.
The namespace (username or organization) that owns the Space.
Optional session UUID to filter events for a specific session.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_JWTTool to generate a JWT token for accessing a Hugging Face space repository. Use when you need authenticated access to spaces, with optional write access for spaces in dev mode, custom expiration, and encryption support.
Input parameters
The name of the space repository.
Enable write access for spaces in dev mode. Set to 'true' to enable write access.
Request an encrypted token. Set to 'true' to receive an encrypted JWT token and key ID.
The namespace (organization or user) that owns the space repository.
Custom expiration time for the JWT token. Specify the duration or timestamp for token validity.
Enable inference API access with the token. Set to 'true' to enable.
Include PRO subscription status in the token. Set to 'true' to include PRO status information.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_METRICSTool to get live metrics for a specific Space in a streaming fashion, with SSE protocol, such as current Zero-GPU usage. Use when you need real-time monitoring of Space resource utilization.
Input parameters
The repository name of the Space.
Maximum time in seconds to wait for metrics updates. Defaults to 5 seconds.
The namespace (username or organization) that owns the Space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_NOTEBOOKTool to retrieve a Jupyter notebook URL from a Hugging Face space repository. Use when you need to access or display a notebook file stored in a space.
Input parameters
The revision, branch name, tag, or commit hash to retrieve the notebook from (e.g., 'main', 'dev', or a commit SHA).
The path to the Jupyter notebook file within the repository, including the .ipynb extension (e.g., 'notebooks/automatic_mask_generator_example.ipynb').
The repository name of the space (e.g., 'SAM2-Video-Predictor').
The namespace or organization name that owns the space repository (e.g., 'fffiloni').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_RESOLVETool to resolve and retrieve a file from a Hugging Face Space repository. Use when you need to download a file from a Space or get XET file information. This endpoint follows redirections (HTTP 302/307) to resolve the final file location.
Input parameters
The git revision (branch, tag, or commit hash) to retrieve the file from. Use 'main' for the default branch or a specific commit SHA for exact version.
The file path within the repository to resolve. This wildcard path parameter can include subdirectories and filename.
The repository name of the Space. For example, 'stable-diffusion-3-medium' for the stabilityai/stable-diffusion-3-medium Space.
The range in bytes of the file to download. Format: 'bytes=start-end'. If not provided, the entire file will be retrieved.
Accept header to specify the response format. Use 'application/vnd.xet-fileinfo+json' to get JSON information about XET file info instead of the file content.
The namespace (user or organization) that owns the Space. For example, 'stabilityai' for the stabilityai/stable-diffusion-3-medium Space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_SCANTool to retrieve the security scan status of a Hugging Face space repository. Use when you need to check for malware, pickle vulnerabilities, or other security issues in a space.
Input parameters
The name of the space repository to check security status for
The namespace (organization or user) that owns the space repository
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_TREESIZETool to get the total size of a Hugging Face space repository at a specific revision and path. Use when you need to determine storage requirements or track repository size changes.
Input parameters
The Git revision or branch name to query (e.g., 'main', 'dev', or a commit SHA).
The path within the repository to calculate size for. Use '.' for root directory or specify a subdirectory path (e.g., 'static/images'). The size is calculated recursively for all files under this path.
The repository name of the space (e.g., 'hello_world', 'chatbot').
The namespace or organization name that owns the space (e.g., 'gradio', 'huggingface').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_XET_READ_TOKENTool to retrieve a short-lived XET read access token for a Hugging Face Space repository. Use when you need to access XET (eXtensible Tensor) data for a specific Space revision.
Input parameters
The revision (branch, tag, or commit hash) to get the XET read token for. Common values include 'main', 'master', or specific commit hashes.
The repository name of the Space. For example, 'Ace-Step-v1.5' in 'ACE-Step/Ace-Step-v1.5'.
The namespace (organization or user) that owns the Space repository. For example, 'ACE-Step' in 'ACE-Step/Ace-Step-v1.5'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_SPACES_XET_WRITE_TOKENTool to retrieve a short-lived XET write access token for a Hugging Face space repository. Use when you need to upload or write XET data to a specific space revision.
Input parameters
The revision (branch, tag, or commit hash) to get the XET write token for. Common values include 'main', 'master', or specific commit hashes.
The repository name of the space. For example, 'ComfyUI' in 'SpacesExamples/ComfyUI'.
Pass 1 to enable PR creation mode for users without direct write access to the repository. When set, changes will be submitted as a pull request instead of direct commits.
The namespace (organization or user) that owns the space repository. For example, 'SpacesExamples' in 'SpacesExamples/ComfyUI'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_TRENDINGTool to retrieve trending repositories from Hugging Face. Use when you need to discover popular models, datasets, or spaces that are currently trending on the platform.
Input parameters
Repository type filter for trending repositories.
Maximum number of trending repositories to return. Defaults to 10.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_USERS_AVATARTool to retrieve the avatar URL for a Hugging Face user. Use when you need to get the avatar image URL for a specific user.
Input parameters
If provided, redirect to the avatar URL instead of returning it as JSON. Set to any value to enable redirection.
The username of the user whose avatar URL to retrieve. For example, 'julien-c' or 'lysandre'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_USERS_OVERVIEWTool to retrieve a comprehensive overview of a Hugging Face user's profile. Use when you need to get user statistics, organizations, activity counts, and profile information.
Input parameters
Username of the Hugging Face user to retrieve overview for. This is the user's handle on Hugging Face.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_USERS_SOCIALSTool to retrieve a user's social media handles from Hugging Face. Use when you need to find a user's GitHub, LinkedIn, Twitter/X, or Bluesky profiles. Only returns handles that the user has publicly shared on their Hugging Face profile.
Input parameters
The username of the Hugging Face user to retrieve social media handles for. This is the user's username on Hugging Face.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_GET_WHOAMITool to get information about the authenticated Hugging Face user including username, email, organizations, and token details. Use when you need to identify the current user from an access token or retrieve user profile information.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_HANDLE_DATASETS_USER_ACCESS_REQUESTTool to handle a user's access request to a gated Hugging Face dataset. Use this to accept, reject, or update the status of access requests for repositories with gated access. Either 'user' or 'userId' must be provided to identify the user.
Input parameters
The repository name of the gated dataset. For example, 'test-gated-dataset-cancel'.
The username of the user whose access request is being handled. Either 'user' or 'userId' must be provided, but not both.
The decision on the access request. Use 'accepted' to grant access, 'rejected' to deny access, or 'pending' to keep the request pending.
The user ID (24-character hexadecimal string) of the user whose access request is being handled. Either 'user' or 'userId' must be provided, but not both.
The namespace (organization or user) that owns the dataset. For example, '121tester' or 'huggingface'.
The reason for rejecting the access request. Maximum 200 characters. Only applicable when status is 'rejected'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_COLLECTIONSTool to list collections on the Hugging Face Hub. Use when you need to discover collections of models, datasets, spaces, or papers. Collections are curated groups of repositories organized by users.
Input parameters
Search query to filter collections by title or description. Use keywords to find relevant collections.
Filter collections by item ID (repo_id, paper id, or collection slug). Use this to find collections containing a specific item.
Enum for sorting options for collections.
Maximum number of collections to return per page. Defaults to 10.
Filter collections by owner username. Use this to get collections created by a specific user or organization.
Pagination cursor for fetching the next page of results. Use the cursor from the previous response to get more collections.
Comma-separated list of fields to expand in the response. Use this to get additional details about collections.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASET_PARQUET_FILESTool to get the list of Parquet files for a dataset. Use when you need to download or access dataset files in Parquet format. Returns URLs to download Parquet files with metadata about splits, configurations, and file sizes.
Input parameters
Name of the dataset to get Parquet files for. Format: 'namespace/repo-name' (e.g., 'rajpurkar/squad', 'huggingface/cifar10').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASET_PATHS_INFOTool to list detailed information about specific paths in a Hugging Face dataset repository. Use when you need to get metadata about files or directories in a dataset, including size, type, commit history, and security scan status.
Input parameters
Git revision (branch, tag, or commit hash) to query
Dataset repository name
List of paths to get information about, or a single path string. Paths are relative to the repository root.
Whether to expand the response with last commit and security file status information. Can be a boolean or an object with expansion options.
Namespace (user or organization) that owns the dataset
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASETSTool to list datasets on the Hugging Face Hub. Use when you need to discover or search for datasets. Supports filtering by author, search query, tags, and sorting by various properties.
Input parameters
Enum for sorting properties.
Limit the number of datasets returned. Defaults to 100 if not specified.
Filter datasets by an author or organization. Use this to get datasets from a specific user or org.
Pagination cursor for fetching the next page of results. Use the next_cursor value from a previous response.
Filter based on tags. Use this to find datasets with specific tags like task types, languages, or licenses.
Filter based on substrings for dataset names and their usernames. Use this to search for datasets by keyword.
Enum for sort direction.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASETS_COMMITSTool to list commits from a Hugging Face dataset repository. Use when you need to retrieve the commit history for a specific dataset branch or revision.
Input parameters
Page number for pagination. Starts from 0 for the first page.
The revision (branch, tag, or commit hash) to list commits from.
The repository name of the dataset.
Maximum number of commits to return per page. Default is 50.
The namespace (username or organization) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASET_SPLITSTool to get the list of subsets and splits of a dataset. Returns the available configurations and splits for a given dataset on the Hub. Use when you need to understand the structure of a dataset before querying specific splits.
Input parameters
Name of the dataset to retrieve splits for. Format: 'namespace/repo-name' (e.g., 'stanfordnlp/imdb', 'huggingface/squad').
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASETS_REFSTool to list all references (branches, tags, converts, pull requests) in a Hugging Face dataset repository. Use when you need to retrieve available references for a specific dataset.
Input parameters
The repository name of the dataset.
The namespace (organization or user) that owns the dataset.
Whether to include pull requests in the response. Set to true to include pull requests.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DATASETS_USER_ACCESS_REQUESTTool to list access requests for a gated Hugging Face dataset repository. Use when you need to view pending, accepted, or rejected access requests for datasets with restricted access.
Input parameters
The repository name of the gated dataset
Cursor for pagination - returns requests after this cursor
Maximum number of access requests to return (default: 1000)
Cursor for pagination - returns requests before this cursor
Filter access requests by status: pending, accepted, or rejected
The namespace or organization name that owns the dataset
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DISCUSSIONSTool to list discussions for a Hugging Face repository. Use when you need to retrieve discussions or pull requests for a specific model, dataset, or space.
Input parameters
Page number for pagination. Starts from 0 for the first page.
The repository name to list discussions from.
Sort type enum.
Discussion type enum.
Filter discussions by author username.
Search query to filter discussions by title or content.
Discussion status enum.
The namespace (username or organization) that owns the repository.
Type of repository to list discussions from: models, spaces, or datasets.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_DOCSTool to retrieve the list of available documentation from Hugging Face. Use when you need to discover available documentation resources.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_ENDPOINTSTool to list Hugging Face Inference Endpoints for a specific user or organization. Use when you need to retrieve endpoints, optionally filtered by tags or name.
Input parameters
Filter endpoints by tags (comma-separated). Use this to find endpoints with specific tags.
Maximum number of endpoints to return per page. Defaults to 20 if not specified.
Pagination cursor to fetch the next or previous page of results. Use the nextCursor or prevCursor from the previous response.
Filter endpoints by name substring. Use this to search for endpoints containing specific text in their name.
User or organization name to list endpoints for. This is the owner of the endpoints you want to retrieve.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_MODELS_COMMITSTool to list commits from a Hugging Face model repository. Use when you need to retrieve the commit history for a specific model branch or revision.
Input parameters
Page number for pagination. Starts from 0 for the first page.
The revision (branch, tag, or commit hash) to list commits from.
The repository name of the model.
Maximum number of commits to return per page. Default is 50.
List of fields to expand in the response. Use 'formatted' to include formatted commit messages.
The namespace (username or organization) that owns the model.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_MODELS_PATHS_INFOTool to list detailed information about specific paths in a Hugging Face model repository. Use when you need to get metadata about files or directories in a model, including size, type, commit history, and security scan status.
Input parameters
Git revision (branch, tag, or commit hash) to query
Model repository name
List of paths to get information about, or a single path string. Paths are relative to the repository root.
Whether to expand the response with last commit and security file status information. Can be a boolean or an object with expansion options.
Namespace (user or organization) that owns the model
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_MODELS_REFSTool to list all references (branches, tags, converts, and optionally pull requests) in a Hugging Face model repository. Use when you need to retrieve version control information for a specific model.
Input parameters
Repository name within the namespace (e.g., 'bert-base-uncased')
Namespace or organization name (e.g., 'google-bert', 'facebook')
Whether to include pull requests in the response. Defaults to false.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_NOTIFICATIONSTool to list notifications for the authenticated Hugging Face user. Use when you need to retrieve user notifications, optionally filtered by read status, repository type, author, or other criteria.
Input parameters
Page number for pagination. Defaults to 0.
Enum for mention filter.
Filter notifications by paper ID.
Filter notifications by repository name.
Enum for repository type filter.
Filter notifications by article ID.
Filter notifications by last update timestamp.
Filter notifications by the author of the post.
Enum for notification read status filter.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_REPO_FILESTool to get the file tree of a Hugging Face repository with pagination support. Use when you need to browse files and folders in any repository type (model, dataset, or space), explore repository structure, or discover available files.
Input parameters
Path within the repository to list. Use empty string or '.' for root directory. Supports nested paths.
Maximum number of items to return. Defaults to 1000, or 100 when expand=true. Use with cursor for pagination.
Pagination cursor from a previous response to fetch the next page of results. Omit for the first page.
If true, returns associated commit data for each entry and security scanner metadata. Defaults to false if not specified.
Repository ID in the format author/repo-name. Example: 'google-bert/bert-base-uncased' for a model.
Git revision (branch, tag, or commit SHA) to list. Use 'main' for default branch.
If true, returns the tree recursively including all subdirectories. Defaults to false if not specified.
Repository type: model, dataset, or space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SETTINGS_WEBHOOKSTool to list all webhooks configured in Hugging Face settings. Use when you need to retrieve webhook configurations for the authenticated user's account.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACESTool to list Spaces on the Hugging Face Hub with filtering options. Use when you need to discover or search for ML demo applications hosted on Hugging Face.
Input parameters
Property to use when sorting results. Common values include 'lastModified', 'likes', 'trending', or 'created'.
Maximum number of spaces to return in the response. Must be between 1 and 500.
Filter spaces by a specific author or organization name.
Filter based on tags. You can specify tags to find spaces by SDK type (e.g., 'gradio', 'streamlit'), task type, or other characteristics.
Filter based on substrings for repos and their usernames. Use to search for specific space names or authors.
Direction in which to sort results. Use '1' for ascending order or '-1' for descending order. Must be used together with the sort parameter.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACES_COMMITSTool to list commits from a Hugging Face Space repository. Use when you need to retrieve the commit history for a specific Space branch or revision.
Input parameters
Page number for pagination. Starts from 0 for the first page.
The revision (branch, tag, or commit hash) to list commits from.
The repository name of the Space.
Maximum number of commits to return per page. Default is 50.
List of fields to expand in the response. Use 'formatted' to include formatted commit messages.
The namespace (username or organization) that owns the Space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACES_HARDWARETool to retrieve available hardware configurations for Hugging Face Spaces with their specifications and pricing. Use when you need to discover compute options for running spaces.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACES_LFS_FILESTool to list LFS (Large File Storage) files from a Hugging Face Space repository. Use when you need to retrieve large files stored in a Space using Git LFS.
Input parameters
Filter for Xet-specific files. Used for Xet-enabled repositories.
The repository name of the Space.
Maximum number of LFS files to return. Default is 1000.
Cursor for pagination. Use the cursor returned from a previous request to fetch the next page of results.
The namespace (username or organization) that owns the Space.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACES_PATHS_INFOTool to list detailed information about specific paths in a Hugging Face space repository. Use when you need to get metadata about files or directories in a space, including size, type, commit history, and security scan status.
Input parameters
Git revision (branch, tag, or commit hash) to query
Space repository name
List of paths to get information about, or a single path string. Paths are relative to the repository root.
Whether to expand the response with last commit and security file status information. Can be a boolean or an object with expansion options.
Namespace (user or organization) that owns the space
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_SPACES_REFSTool to list all references (branches, tags, converts, pull requests) in a Hugging Face space repository. Use when you need to retrieve available references for a specific space.
Input parameters
The repository name of the space.
The namespace (organization or user) that owns the space.
Whether to include pull requests in the response. Set to true to include pull requests.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_LIST_VENDORSTool to list available cloud provider vendors for Hugging Face Inference Endpoints. Use when you need to discover available infrastructure options across AWS, Azure, and GCP. Returns vendors with their regions and compute instance types for deploying models.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_SEARCH_DATASETTool to search text in a dataset split on Hugging Face. Searches in columns of type string, even if values are nested in a dictionary. Use when you need to find specific text or patterns within a dataset's content. Returns matching rows with their data.
Input parameters
Text to search for in the dataset. Searches in all columns of type string, even if values are nested in a dictionary.
Name of the dataset split to search.
Name of the configuration/subset of the dataset to search.
Length of the slice (number of rows to return). Maximum value is 100. Defaults to API default if not specified.
Offset of the slice for pagination. Specifies the starting position in the results. Defaults to 0 if not specified.
Name of the dataset to search. Use format 'owner/dataset-name' for user datasets or just 'dataset-name' for official datasets.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_SEARCH_DOCSTool to search Hugging Face documentation across all products and libraries. Use when you need to find information about HF tools, models, datasets, or API usage.
Input parameters
Search query string to find relevant documentation pages. Use keywords related to the topic you want to learn about.
Maximum number of search results to return. Defaults to 10.
Number of results to skip for pagination. Use with limit to page through results. For example, offset=10 with limit=10 returns results 11-20.
Enum for Hugging Face product types to search documentation.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_SEARCH_PAPERSTool to perform hybrid semantic/full-text search on papers in Hugging Face. Use when you need to find research papers by keywords, topics, or authors.
Input parameters
Search query string to find relevant papers. Use keywords, paper titles, or topics. Supports hybrid semantic and full-text search.
Maximum number of papers to return. If not specified, the API will use its default limit.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_SQUASH_DATASET_COMMITSTool to squash all commits in a dataset ref into a single commit with the given message. Use when consolidating commit history into a single commit. WARNING: This operation is irreversible.
Input parameters
The ref (branch or revision) to squash. This operation is irreversible and will squash all commits in this ref into a single commit.
The name of the dataset repository to squash commits in.
The commit message for the squashed commit. Maximum length is 500 characters.
The namespace (organization or user) that owns the dataset.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_SQUASH_SPACES_COMMITSTool to squash all commits in a space ref into a single commit with the given message. Use when consolidating commit history into a single commit. WARNING: This operation is irreversible.
Input parameters
The ref (branch or revision) to squash. This operation is irreversible and will squash all commits in this ref into a single commit. Typically 'main' or a branch name.
The name of the space repository where commits will be squashed. For example, for space '121tester/test-curl-space', the repo is 'test-curl-space'.
The commit message for the squashed commit. Maximum length is 500 characters. If not provided, a default message will be used.
The namespace (username or organization) that owns the space repository. For example, for space '121tester/test-curl-space', the namespace is '121tester'.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_DATASETS_SETTINGSTool to update settings for a Hugging Face dataset repository. Use when you need to configure visibility, discussions, gating, or access control for a dataset.
Input parameters
The name of the dataset repository
Gating configuration for the dataset. Controls access restrictions and gating behavior. Can be a boolean to enable/disable gating, or an object with gating configuration
Whether the dataset is private. Set to true to make private, false to make public
The namespace (organization or username) that owns the dataset
Enum for discussion sorting options.
Whether discussions are disabled for this dataset. Set to true to disable discussions, false to enable them
Enum for gated notification modes.
Email address to receive notifications about gated access requests. Must be a valid email format
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_DISCUSSIONS_TITLETool to change the title of an existing discussion on a Hugging Face repository (model, dataset, or Space). Use when you need to update or correct a discussion's title.
Input parameters
The discussion number/identifier to update the title for.
The name of the repository where the discussion is located.
The new title for the discussion. Must be between 3 and 200 characters.
The namespace (username or organization) that owns the repository.
The type of repository: models, spaces, or datasets. This determines where the discussion is located.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_MODELS_SETTINGSTool to update settings for a Hugging Face model repository. Use when you need to modify repository configuration such as privacy, discussions, or gated access settings.
Input parameters
The name of the model repository to update settings for.
Whether the model repository should be private (requires authentication to access) or public.
The namespace (username or organization) that owns the model repository.
Sorting order for discussions.
Whether to disable discussions for this model repository. Set to true to disable, false to enable.
Notification mode for gated access requests.
Email address to receive notifications about gated access requests. Must be a valid email format.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_SETTINGS_NOTIFICATIONSTool to update notification settings for the authenticated Hugging Face user. Use when you need to enable or disable various notification types such as announcements, discussions, paper digests, or product updates.
Input parameters
Notification settings to update. Only include the specific notification preferences you want to change.
Prepaid amount to be provided when enabling launch_prepaid_credits notification. Maximum length is 24 characters.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_SETTINGS_WATCHTool to update watch settings for your Hugging Face account. Use when you want to get notified about discussions on organizations, users, or repositories. You can add new items to watch and/or remove items from your watch list in a single request.
Input parameters
Array of items to start watching. Each item must specify an 'id' and 'type'. You will receive notifications when discussions happen on these items. Leave empty if you only want to remove items from watch list.
Array of items to stop watching. Each item must specify an 'id' and 'type'. You will no longer receive notifications for discussions on these items. Leave empty if you only want to add items to watch list.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_SETTINGS_WEBHOOKSTool to update an existing webhook in Hugging Face settings. Use when you need to modify webhook configuration such as watched entities, event domains, target URL, or job settings.
Input parameters
Model for webhook job configuration to execute when webhook triggers.
URL endpoint where webhook POST requests will be sent when events occur. Must be a valid HTTP/HTTPS URL.
Secret string used to sign webhook requests for verification. Must contain only printable ASCII characters.
List of event domains that trigger the webhook. At least one domain is required. 'repo' for repository events, 'discussion' for discussion events.
List of entities to watch (users, organizations, or repositories). At least one item is required. Events from these entities will trigger the webhook.
The unique identifier of the webhook to update (24-character hexadecimal string). Get this from the list webhooks action.
Source identifier for the job. Used to track the origin of the job execution.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_SPACES_SETTINGSTool to update settings for a Hugging Face Spaces repository. Use when you need to modify repository configuration such as privacy, discussions, or gated access settings.
Input parameters
The name of the Spaces repository to update settings for.
Gating configuration for the Spaces repository. Controls access restrictions and gating behavior. Can be a boolean to enable/disable gating, or an object with gating configuration.
Whether the Spaces repository should be private (requires authentication to access) or public.
The namespace (username or organization) that owns the Spaces repository.
Sorting order for discussions.
Whether to disable discussions for this Spaces repository. Set to true to disable, false to enable.
Notification mode for gated access requests.
Email address to receive notifications about gated access requests. Must be a valid email format.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_SQL_CONSOLE_EMBEDTool to update an existing SQL console embed for a Hugging Face dataset. Use when you need to modify the SQL query, title, or privacy settings of an existing embed.
Input parameters
The slug identifier of the SQL console embed to update (not the _id). This is returned as 'slug' when creating an embed
The SQL query string to execute in the console embed
The name of the dataset repository
The title of the SQL console embed. Maximum 200 characters
Whether the SQL console embed is private. Set to true to make it private, false to make it public
The namespace (username or organization) that owns the dataset repository
The type of repository. Currently only 'datasets' is supported
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
HUGGING_FACE_UPDATE_WEBHOOK_STATUSTool to enable or disable a webhook on Hugging Face. Use when you need to temporarily deactivate a webhook without deleting it, or reactivate a previously disabled webhook.
Input parameters
The action to perform on the webhook: 'enable' to activate the webhook or 'disable' to deactivate it.
The unique identifier of the webhook to enable or disable. Must be a 24-character hexadecimal string.
Output
Data from the action execution
Error if any occurred during the execution of the action
Whether or not the action execution was successful or not
No publicly available marketplace agent is found using this tool yet. There are 59 agents privately built on Nagent that already use Hugging Face.
Build on Nagent
Connect Hugging Face to any Nagent agent in minutes — no API key management, no boilerplate. Just configure and deploy.
The five questions agent builders ask before adopting a new integration.
Open the External Integrations panel inside Nagent (app.nagent.ai/externalIntegration), find Hugging Face, and click "Connect Now." You'll authenticate with API_KEY, OAUTH2 — Nagent handles credential storage and refresh automatically. Once connected, Hugging Face is available to any agent in your workspace.
No. Nagent provides no-code integration for every tool. Once Hugging Face is connected, you configure its 135 actions directly in the agent builder UI — no API calls, no boilerplate, no schema management.
Helix — Nagent's agentic agent builder — lets you drop Hugging Face steps into any workflow visually. Pick an action (e.g., one of those listed above), fill in the inputs (Helix knows the required vs. optional schema for each parameter), and connect it to upstream/downstream steps. Triggers run as the entry point of an agent, so when a Hugging Face event fires, the agent kicks off automatically.
Every Hugging Face action and trigger ships with a fully-typed schema — input parameters with name, type, required flag, and description, plus the output payload shape. The schemas are documented in the sections above. Helix uses these schemas to validate your configuration at build time and to type-check the data flowing between steps.
Yes. While Hugging Face ships with 135 pre-built artificial intelligence actions, you can layer custom logic around them inside Helix — pre/post-processing steps, conditional branches, retries, or stitching Hugging Face together with other connected tools. For deeper customization, talk to our team about Nagent's Agentic AI Lab — forward-deployed engineers who build Hugging Face-based workflows tailored to your business.