Skip to main content

Module airbyte_agent_sdk.connectors.google_drive.connector

Google-Drive connector.

Classes​

AboutQuery(connector: GoogleDriveConnector) : Query class for About entity operations.

Initialize query with connector reference.

Methods​

get(self, fields: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.About : Gets information about the user, the user's Drive, and system capabilities

Args: fields: Fields to include in the response (use * for all fields) **kwargs: Additional parameters

Returns: About

ChangesQuery(connector: GoogleDriveConnector) : Query class for Changes entity operations.

Initialize query with connector reference.

Methods​

list(self, page_token: str | None = None, page_size: int | None = None, drive_id: str | None = None, include_items_from_all_drives: bool | None = None, supports_all_drives: bool | None = None, spaces: str | None = None, include_removed: bool | None = None, restrict_to_my_drive: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Change], ChangesListResultMeta] : Lists the changes for a user or shared drive

Args: page_token: Token for the page of changes to retrieve (from changes.getStartPageToken or previous response) page_size: Maximum number of changes to return (1-1000) drive_id: The shared drive from which changes are returned include_items_from_all_drives: Whether to include changes from all drives supports_all_drives: Whether the requesting application supports both My Drives and shared drives spaces: Comma-separated list of spaces to query include_removed: Whether to include changes indicating that items have been removed restrict_to_my_drive: Whether to restrict the results to changes inside the My Drive hierarchy **kwargs: Additional parameters

Returns: ChangesListResult

ChangesStartPageTokenQuery(connector: GoogleDriveConnector) : Query class for ChangesStartPageToken entity operations.

Initialize query with connector reference.

Methods​

get(self, drive_id: str | None = None, supports_all_drives: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.StartPageToken : Gets the starting pageToken for listing future changes

Args: drive_id: The ID of the shared drive for which the starting pageToken is returned supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: StartPageToken

CommentsQuery(connector: GoogleDriveConnector) : Query class for Comments entity operations.

Initialize query with connector reference.

Methods​

get(self, file_id: str, comment_id: str, include_deleted: bool | None = None, fields: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.Comment : Gets a comment by ID

Args: file_id: The ID of the file comment_id: The ID of the comment include_deleted: Whether to return deleted comments fields: Fields to include in the response (required for comments) **kwargs: Additional parameters

Returns: Comment

list(self, file_id: str, page_size: int | None = None, page_token: str | None = None, start_modified_time: str | None = None, include_deleted: bool | None = None, fields: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Comment], CommentsListResultMeta] : Lists a file's comments

Args: file_id: The ID of the file page_size: Maximum number of comments to return (1-100) page_token: Token for continuing a previous list request start_modified_time: Minimum value of modifiedTime to filter by (RFC 3339) include_deleted: Whether to include deleted comments fields: Fields to include in the response (required for comments) **kwargs: Additional parameters

Returns: CommentsListResult

DrivesQuery(connector: GoogleDriveConnector) : Query class for Drives entity operations.

Initialize query with connector reference.

Methods​

get(self, drive_id: str, use_domain_admin_access: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.Drive : Gets a shared drive's metadata by ID

Args: drive_id: The ID of the shared drive use_domain_admin_access: Issue the request as a domain administrator **kwargs: Additional parameters

Returns: Drive

list(self, page_size: int | None = None, page_token: str | None = None, q: str | None = None, use_domain_admin_access: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Drive], DrivesListResultMeta] : Lists the user's shared drives

Args: page_size: Maximum number of shared drives to return (1-100) page_token: Token for continuing a previous list request q: Query string for searching shared drives use_domain_admin_access: Issue the request as a domain administrator **kwargs: Additional parameters

Returns: DrivesListResult

FilesExportQuery(connector: GoogleDriveConnector) : Query class for FilesExport entity operations.

Initialize query with connector reference.

Methods​

download(self, file_id: str, mime_type: str, range_header: str | None = None, **kwargs) ‑> AsyncIterator[bytes] : Exports a Google-NATIVE Workspace file (Docs, Sheets, Slides, Drawings -- mimeType application/vnd.google-apps.*) to a specified format. Use this ONLY for those native types: exporting a binary file (PDF, image, uploaded .docx/.xlsx) returns 403 fileNotExportable -- for those use the files download action instead. If unsure of a file's type, check its mimeType with files.get first. Common export formats:

  • application/pdf (all types)

  • text/plain (Docs)

  • text/csv (Sheets)

  • application/vnd.openxmlformats-officedocument.wordprocessingml.document (Docs to .docx)

  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet (Sheets to .xlsx)

  • application/vnd.openxmlformats-officedocument.presentationml.presentation (Slides to .pptx) Note: Export has a 10MB limit. For larger files, use the Drive UI.

    Args: file_id: The ID of the Google Workspace file to export mime_type: The MIME type of the format to export to. Common values:

  • application/pdf

  • text/plain

  • text/csv

  • application/vnd.openxmlformats-officedocument.wordprocessingml.document

  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

  • application/vnd.openxmlformats-officedocument.presentationml.presentation

    range_header: Optional Range header for partial downloads (e.g., 'bytes=0-99') **kwargs: Additional parameters

    Returns: AsyncIterator[bytes]

download_base64(self, file_id: str, mime_type: str, range_header: str | None = None, **kwargs) ‑> dict[str, typing.Any] : Exports a Google-NATIVE Workspace file (Docs, Sheets, Slides, Drawings -- mimeType application/vnd.google-apps.*) to a specified format. Use this ONLY for those native types: exporting a binary file (PDF, image, uploaded .docx/.xlsx) returns 403 fileNotExportable -- for those use the files download action instead. If unsure of a file's type, check its mimeType with files.get first. Common export formats:

  • application/pdf (all types)
  • text/plain (Docs)
  • text/csv (Sheets)
  • application/vnd.openxmlformats-officedocument.wordprocessingml.document (Docs to .docx)
  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet (Sheets to .xlsx)
  • application/vnd.openxmlformats-officedocument.presentationml.presentation (Slides to .pptx) Note: Export has a 10MB limit. For larger files, use the Drive UI. and return a JSON-safe base64 chunk.

download_local(self, file_id: str, mime_type: str, path: str, range_header: str | None = None, **kwargs) ‑> Path : Exports a Google-NATIVE Workspace file (Docs, Sheets, Slides, Drawings -- mimeType application/vnd.google-apps.*) to a specified format. Use this ONLY for those native types: exporting a binary file (PDF, image, uploaded .docx/.xlsx) returns 403 fileNotExportable -- for those use the files download action instead. If unsure of a file's type, check its mimeType with files.get first. Common export formats:

  • application/pdf (all types)

  • text/plain (Docs)

  • text/csv (Sheets)

  • application/vnd.openxmlformats-officedocument.wordprocessingml.document (Docs to .docx)

  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet (Sheets to .xlsx)

  • application/vnd.openxmlformats-officedocument.presentationml.presentation (Slides to .pptx) Note: Export has a 10MB limit. For larger files, use the Drive UI. and save to file.

    Args: file_id: The ID of the Google Workspace file to export mime_type: The MIME type of the format to export to. Common values:

  • application/pdf

  • text/plain

  • text/csv

  • application/vnd.openxmlformats-officedocument.wordprocessingml.document

  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet

  • application/vnd.openxmlformats-officedocument.presentationml.presentation

    range_header: Optional Range header for partial downloads (e.g., 'bytes=0-99') path: File path to save downloaded content **kwargs: Additional parameters

    Returns: str: Path to the downloaded file

download_text(self, file_id: str, mime_type: str, range_header: str | None = None, **kwargs) ‑> dict[str, typing.Any] : Exports a Google-NATIVE Workspace file (Docs, Sheets, Slides, Drawings -- mimeType application/vnd.google-apps.*) to a specified format. Use this ONLY for those native types: exporting a binary file (PDF, image, uploaded .docx/.xlsx) returns 403 fileNotExportable -- for those use the files download action instead. If unsure of a file's type, check its mimeType with files.get first. Common export formats:

  • application/pdf (all types)
  • text/plain (Docs)
  • text/csv (Sheets)
  • application/vnd.openxmlformats-officedocument.wordprocessingml.document (Docs to .docx)
  • application/vnd.openxmlformats-officedocument.spreadsheetml.sheet (Sheets to .xlsx)
  • application/vnd.openxmlformats-officedocument.presentationml.presentation (Slides to .pptx) Note: Export has a 10MB limit. For larger files, use the Drive UI. and return a JSON-safe UTF-8 text chunk.

FilesQuery(connector: GoogleDriveConnector) : Query class for Files entity operations.

Initialize query with connector reference.

Methods​

context_store_search(self, query: FilesSearchQuery, limit: int | None = None, cursor: str | None = None, fields: list[list[str]] | None = None) ‑> airbyte_agent_sdk.connectors.google_drive.models.AirbyteSearchResult[FilesSearchData] : Search files records from Airbyte cache.

This operation searches cached data from Airbyte syncs. Only available in hosted execution mode.

Available filter fields (FilesSearchFilter):

  • id: Unique identifier of the file in Google Drive.
  • updated_at: Timestamp of the last modification to the file.
  • file_name: Name of the file.
  • file_path: Full path of the file within the synced Drive folder.
  • mime_type: MIME type of the file.
  • content: Extracted text content of the file.

Args: query: Filter and sort conditions. Supports operators such as eq, neq, gt, gte, lt, lte, in, startswith, endswith, contains, array_contains, fuzzy, keyword, not, and, or. Example: {"filter": {"eq": {"status": "active"}}} limit: Maximum results to return (default 1000) cursor: Pagination cursor from previous response's meta.cursor fields: Field paths to include in results. Each path is a list of keys for nested access. Example: [["id"], ["user", "name"]] returns id and user.name fields.

Returns: FilesSearchResult with typed records, pagination metadata, and optional search metadata

Raises: NotImplementedError: If called in local execution mode

context_store_sql_query(self, sql: str, limit: int | None = None) ‑> airbyte_agent_sdk.connectors.google_drive.models.AirbyteSearchResult[dict[str, Any]] : Run a SQL query against files records in the Airbyte Context Store.

Only available in hosted execution mode.

Args: sql: SQL query to execute. limit: Maximum results to return.

Returns: AirbyteSearchResult containing the projected rows and query metadata.

Raises: NotImplementedError: If called in local execution mode.

create(self, name: str, mime_type: str | None = None, parents: list[str] | None = None, description: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.File : Creates a new file or folder in Google Drive (metadata only, no content). To create a folder, set mimeType to 'application/vnd.google-apps.folder'. To create a Google Doc, use 'application/vnd.google-apps.document'. To create a Google Sheet, use 'application/vnd.google-apps.spreadsheet'.

Args: name: The name of the file or folder mime_type: The MIME type of the file. Use 'application/vnd.google-apps.folder' for folders, 'application/vnd.google-apps.document' for Google Docs, 'application/vnd.google-apps.spreadsheet' for Google Sheets.

parents: The IDs of the parent folders. If not specified, the file is placed in My Drive root. description: A short description of the file **kwargs: Additional parameters

Returns: File

delete(self, file_id: str, supports_all_drives: bool | None = None, **kwargs) ‑> dict[str, typing.Any] : Permanently deletes a file owned by the user without moving it to the trash.

Args: file_id: The ID of the file to delete supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: dict[str, Any]

download(self, file_id: str, alt: str | None = None, acknowledge_abuse: bool | None = None, supports_all_drives: bool | None = None, range_header: str | None = None, **kwargs) ‑> AsyncIterator[bytes] : Downloads the raw binary content of a file (PDF, image, zip, uploaded .docx/.xlsx, etc.). The Drive alt=media query parameter is applied automatically by this action, so you normally do NOT need to pass alt -- the response is the file's bytes. (Without alt=media Drive returns file metadata JSON instead of content, so it is forced here.) This only works for binary files: for Google Workspace files (Docs, Sheets, Slides, Drawings) use the files_export action with a mimeType instead -- downloading them directly returns 403.

Args: file_id: The ID of the file to download alt: Applied automatically as 'media' by this action; you do not need to set it. acknowledge_abuse: Whether the user is acknowledging the risk of downloading known malware or other abusive files supports_all_drives: Whether the requesting application supports both My Drives and shared drives range_header: Optional Range header for partial downloads (e.g., 'bytes=0-99') **kwargs: Additional parameters

Returns: AsyncIterator[bytes]

download_base64(self, file_id: str, alt: str | None = None, acknowledge_abuse: bool | None = None, supports_all_drives: bool | None = None, range_header: str | None = None, **kwargs) ‑> dict[str, typing.Any] : Downloads the raw binary content of a file (PDF, image, zip, uploaded .docx/.xlsx, etc.). The Drive alt=media query parameter is applied automatically by this action, so you normally do NOT need to pass alt -- the response is the file's bytes. (Without alt=media Drive returns file metadata JSON instead of content, so it is forced here.) This only works for binary files: for Google Workspace files (Docs, Sheets, Slides, Drawings) use the files_export action with a mimeType instead -- downloading them directly returns 403. and return a JSON-safe base64 chunk.

download_local(self, file_id: str, path: str, alt: str | None = None, acknowledge_abuse: bool | None = None, supports_all_drives: bool | None = None, range_header: str | None = None, **kwargs) ‑> Path : Downloads the raw binary content of a file (PDF, image, zip, uploaded .docx/.xlsx, etc.). The Drive alt=media query parameter is applied automatically by this action, so you normally do NOT need to pass alt -- the response is the file's bytes. (Without alt=media Drive returns file metadata JSON instead of content, so it is forced here.) This only works for binary files: for Google Workspace files (Docs, Sheets, Slides, Drawings) use the files_export action with a mimeType instead -- downloading them directly returns 403. and save to file.

Args: file_id: The ID of the file to download alt: Applied automatically as 'media' by this action; you do not need to set it. acknowledge_abuse: Whether the user is acknowledging the risk of downloading known malware or other abusive files supports_all_drives: Whether the requesting application supports both My Drives and shared drives range_header: Optional Range header for partial downloads (e.g., 'bytes=0-99') path: File path to save downloaded content **kwargs: Additional parameters

Returns: str: Path to the downloaded file

download_text(self, file_id: str, alt: str | None = None, acknowledge_abuse: bool | None = None, supports_all_drives: bool | None = None, range_header: str | None = None, **kwargs) ‑> dict[str, typing.Any] : Downloads the raw binary content of a file (PDF, image, zip, uploaded .docx/.xlsx, etc.). The Drive alt=media query parameter is applied automatically by this action, so you normally do NOT need to pass alt -- the response is the file's bytes. (Without alt=media Drive returns file metadata JSON instead of content, so it is forced here.) This only works for binary files: for Google Workspace files (Docs, Sheets, Slides, Drawings) use the files_export action with a mimeType instead -- downloading them directly returns 403. and return a JSON-safe UTF-8 text chunk.

get(self, file_id: str, fields: str | None = None, supports_all_drives: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.File : Gets a file's metadata by ID

Args: file_id: The ID of the file fields: Fields to include in the response supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: File

list(self, page_size: int | None = None, page_token: str | None = None, q: str | None = None, order_by: str | None = None, fields: str | None = None, spaces: str | None = None, corpora: str | None = None, drive_id: str | None = None, include_items_from_all_drives: bool | None = None, supports_all_drives: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[File], FilesListResultMeta] : Lists the user's files. Returns a paginated list of files.

Args: page_size: Maximum number of files to return per page (1-1000) page_token: Token for continuing a previous list request q: Query string for searching files order_by: Sort order (e.g., 'modifiedTime desc', 'name') fields: Fields to include in the response spaces: Comma-separated list of spaces to query (drive, appDataFolder) corpora: Bodies of items to search (user, drive, allDrives) drive_id: ID of the shared drive to search include_items_from_all_drives: Whether to include items from all drives supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: FilesListResult

update(self, file_id: str, name: str | None = None, description: str | None = None, mime_type: str | None = None, add_parents: str | None = None, remove_parents: str | None = None, supports_all_drives: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.File : Updates a file's metadata. Use addParents/removeParents query parameters to move a file between folders.

Args: name: The new name of the file description: A new description for the file mime_type: The new MIME type of the file file_id: The ID of the file to update add_parents: Comma-separated list of parent IDs to add remove_parents: Comma-separated list of parent IDs to remove supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: File

FilesUploadQuery(connector: GoogleDriveConnector) : Query class for FilesUpload entity operations.

Initialize query with connector reference.

Methods​

create(self, name: str, file_content: str, mime_type: str | None = None, parents: list[str] | None = None, description: str | None = None, file_mime_type: str | None = None, upload_type: str | None = None, supports_all_drives: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.File : Uploads a new file to Google Drive with both metadata and file content. The file content must be base64-encoded in the file_content parameter. Suitable for files up to 5MB. For larger files, use the Drive UI.

Args: name: The name of the file file_content: Base64-encoded file content to upload mime_type: The MIME type for the file metadata in Google Drive parents: The IDs of the parent folders description: A short description of the file file_mime_type: The MIME type of the actual file content (e.g., 'application/pdf', 'image/png'). Defaults to 'application/octet-stream'. upload_type: The type of upload request (must be 'multipart') supports_all_drives: Whether the requesting application supports both My Drives and shared drives **kwargs: Additional parameters

Returns: File

GoogleDriveConnector(auth_config: GoogleDriveAuthConfig | AirbyteAuthConfig | BaseModel | None = None, on_token_refresh: Any | None = None) : Type-safe Google-Drive API connector.

Auto-generated from OpenAPI specification with full type safety.

Initialize a new google-drive connector instance.

Supports both local and hosted execution modes:

  • Local mode: Provide connector-specific auth config (e.g., GoogleDriveAuthConfig)
  • Hosted mode: Provide AirbyteAuthConfig with client credentials and either connector_id or workspace_name

Args: auth_config: Either connector-specific auth config for local mode, or AirbyteAuthConfig for hosted mode on_token_refresh: Optional callback for OAuth2 token refresh persistence. Called with new_tokens dict when tokens are refreshed. Can be sync or async. Example: lambda tokens: save_to_database(tokens) Examples:

Local mode (direct API calls)

connector = GoogleDriveConnector(auth_config=GoogleDriveAuthConfig(access_token="...", refresh_token="...", client_id="...", client_secret="..."))

Hosted mode with explicit connector_id (no lookup needed)

connector = GoogleDriveConnector( auth_config=AirbyteAuthConfig( airbyte_client_id="client_abc123", airbyte_client_secret="secret_xyz789", connector_id="existing-source-uuid" ) )

Hosted mode with lookup by workspace_name

connector = GoogleDriveConnector( auth_config=AirbyteAuthConfig( workspace_name="user-123", organization_id="00000000-0000-0000-0000-000000000123", airbyte_client_id="client_abc123", airbyte_client_secret="secret_xyz789" ) )

Class variables​

connector_name : The type of the None singleton.

connector_version : The type of the None singleton.

sdk_version : The type of the None singleton.

Static methods​

agent_tool(role: AgentToolRole | None = None, *, inspect_tool: str | None = None, docs_tool: str | None = None, max_output_chars: int | None | Unset = UNSET, framework: FrameworkName = 'none', internal_retries: int = 0, should_internal_retry: Callable[[Exception, tuple[Any, ...], dict[str, Any]], bool] | None = None, exhausted_runtime_failure_message: Callable[[Exception, tuple[Any, ...], dict[str, Any]], str | None] | None = None) ‑> Callable[[~_F], ~_F] : Decorator for new user-written connector tool functions.

Use this when a tool needs a custom body or the framework lacks a native strategy. Instead of baking the full entity/action reference into the docstring, it instructs the agent to call this connector's inspect and docs tools before executing. Tool failures raise :class:airbyte_agent_sdk.AirbyteToolError by default (framework="none", no auto-detection) — pass framework=... to translate to a supported framework's signal instead.

Decorate three functions per connector — execute, inspect and docs. The role is inferred from each function's signature (extra parameters are allowed); a signature matching more than one role, a generic (*args, **kwargs) wrapper, or a callable whose signature cannot be read must pass the role explicitly:

  • (entity, action, ...) -> "execute"
  • (section, ...) -> "read_skill_docs"
  • () -> "inspect_connector"

Usage: connector = GoogleDriveConnector(...)

@GoogleDriveConnector.agent_tool() async def execute(entity: str, action: str, params: dict | None = None): return await connector.execute(entity=entity, action=action, params=params or {})

@GoogleDriveConnector.agent_tool() async def inspect_connector(): return await connector.inspect_connector()

@GoogleDriveConnector.agent_tool() async def read_skill_docs(section: str | None = None): return await connector.read_skill_docs(section)

Args: role: "execute" | "inspect_connector" | "read_skill_docs". None (default) infers the role from the decorated function's signature; an explicit role validates the canonical parameters are present (functions accepting **kwargs, or callables whose signature cannot be read, pass validation). inspect_tool: Exact registered name of the sibling inspect tool, woven into the execute docstring for tighter steering. Defaults to generic phrasing. docs_tool: Exact registered name of the sibling docs tool (see inspect_tool). max_output_chars: Max serialized output size before failing. Defaults per role: execute -> DEFAULT_MAX_OUTPUT_CHARS, docs tools -> None. framework: Translation target for tool failures. Defaults to "none" (raise AirbyteToolError); never auto-detects. internal_retries: How many transient runtime failures (429/5xx, network, timeout) to retry silently before surfacing. Forwarded to :func:airbyte_agent_sdk.translation.translate_exceptions. should_internal_retry: Optional predicate (error, args, kwargs) -> bool further restricting which retryable errors are safe for this specific tool. Forwarded to :func:airbyte_agent_sdk.translation.translate_exceptions. exhausted_runtime_failure_message: Optional callback (error, args, kwargs) -> str | None invoked after internal retries are exhausted or skipped. Forwarded to :func:airbyte_agent_sdk.translation.translate_exceptions.

tool_utils(func: _F | None = None, *, update_docstring: bool = True, max_output_chars: int | None = 100000, framework: FrameworkName | None = None, internal_retries: int = 0, should_internal_retry: Callable[[Exception, tuple[Any, ...], dict[str, Any]], bool] | None = None, exhausted_runtime_failure_message: Callable[[Exception, tuple[Any, ...], dict[str, Any]], str | None] | None = None) ‑> ~_F | Callable[[~_F], ~_F] : Deprecated. Add connector-specific documentation and runtime safeguards to one tool.

Kept for backwards compatibility with existing single-tool integrations; it is not removed and does not warn at runtime, but new code should use build_connector_tools or agent_tool below.

For new agents, prefer build_connector_tools. It returns progressive inspect_connector, read_skill_docs, and execute tools so the agent can load only the connector guidance it needs:

from airbyte_agent_sdk import build_connector_tools
from pydantic_ai import Agent

tools = build_connector_tools(connector, framework="pydantic_ai")
agent = Agent("openai:gpt-4o", tools=tools.as_list())

When a new integration needs custom tool bodies or a framework without native support, use agent_tool instead.

Legacy: one generated-description tool​

Existing integrations can keep using tool_utils for one broad execute tool with the connector's full generated catalog in its description:

from fastmcp import FastMCP

connector = GoogleDriveConnector()
mcp = FastMCP("Connector Agent")

@mcp.tool()
@GoogleDriveConnector.tool_utils
async def execute(entity: str, action: str, params: dict):
    ...

Configure documentation, output limits, framework translation, and retries when needed:

@mcp.tool()
@GoogleDriveConnector.tool_utils(update_docstring=False, max_output_chars=None)
async def execute(entity: str, action: str, params: dict):
    ...

@mcp.tool()
@GoogleDriveConnector.tool_utils(framework="pydantic_ai", internal_retries=2)
async def execute(entity: str, action: str, params: dict):
    ...

This decorator composes translate_exceptions for runtime wrapping, output-size checks, framework signal translation, and optional internal retries, then adds connector-specific docstring augmentation.

Args: update_docstring: When True, append connector capabilities to __doc__. max_output_chars: Max serialized output size before raising. Use None to disable. framework: One of "pydantic_ai" | "langchain" | "openai_agents" | "mcp" | "none". Defaults to None, which auto-detects each framework's canonical import in order and falls back to "none" with a warning when no supported framework is installed. Explicit always wins, and an explicit framework whose package is missing raises RuntimeError. internal_retries: How many transient runtime failures (429/5xx, network, timeout) to retry silently before surfacing. Default 0. Forwarded to airbyte_agent_sdk.translation.translate_exceptions. should_internal_retry: Optional predicate (error, args, kwargs) -> bool further restricting which retryable errors are safe for this specific tool. Forwarded to airbyte_agent_sdk.translation.translate_exceptions. exhausted_runtime_failure_message: Optional callback (error, args, kwargs) -> str | None. Invoked after internal retries are exhausted or were skipped because should_internal_retry returned False. Forwarded to airbyte_agent_sdk.translation.translate_exceptions.

Instance variables​

connector_id: str | None : Get the connector/source ID (only available in hosted mode).

Returns: The connector ID if in hosted mode, None if in local mode.

Methods​

check(self) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveCheckResult : Perform a health check to verify connectivity and credentials.

Executes a lightweight list operation (limit=1) to validate that the connector can communicate with the API and credentials are valid.

Returns: GoogleDriveCheckResult with status ("healthy" or "unhealthy") and optional error message

Example: result = await connector.check() if result.status == "healthy": print("Connection verified!") else: print(f"Check failed: {result.error}")

close(self) : Close the connector and release resources.

entity_schema(self, entity: str) ‑> dict[str, typing.Any] | None : Get the JSON schema for an entity.

Args: entity: Entity name (e.g., "contacts", "companies")

Returns: JSON schema dict describing the entity structure, or None if not found.

Example: schema = connector.entity_schema("contacts") if schema: print(f"Contact properties: {list(schema.get('properties', {}).keys())}")

execute(self, entity: str, action: "Literal['list', 'get', 'create', 'update', 'delete', 'download', 'context_store_search', 'context_store_sql_query']", params: Mapping[str, Any] | None = None, *, select_fields: list[str] | None = None, exclude_fields: list[str] | None = None, skip_truncation: bool = True) ‑> Any : Execute an entity operation with full type safety.

This is the recommended interface for blessed connectors as it:

  • Uses the same signature as non-blessed connectors
  • Provides full IDE autocomplete for entity/action/params
  • Makes migration from generic to blessed connectors seamless

Args: entity: Entity name (e.g., "customers") action: Operation action (e.g., "create", "get", "list") params: Operation parameters (typed based on entity+action) select_fields: Optional allowlist of dot-notation fields to include exclude_fields: Optional blocklist of dot-notation fields to remove skip_truncation: Disable long-text truncation for collection actions

Returns: Typed response based on the operation

Example: customer = await connector.execute( entity="customers", action="get", params={"id": "cus_123"} )

inspect_connector(self) ‑> dict[str, typing.Any] : Inspect this connector's hosted metadata/readiness and resolve its docs skill id.

Call this before read_skill_docs in the normal hosted flow. For local/offline connectors this returns a local-mode payload with a warning instead of a hosted inspection.

Example: info = await connector.inspect_connector() print(info["docs_skill_id"])

list_entities(self) ‑> list[dict[str, typing.Any]] : Get structured data about available entities, actions, and parameters.

Returns a list of entity descriptions with:

  • entity_name: Name of the entity (e.g., "contacts", "deals")
  • description: Entity description from the first endpoint
  • available_actions: List of actions (e.g., ["list", "get", "create"])
  • parameters: Dict mapping action -> list of parameter dicts

Example: entities = connector.list_entities() for entity in entities: print(f"{entity['entity_name']}: {entity['available_actions']}")

read_skill_docs(self, section: str | None = None) ‑> str : Read this connector's usage docs, rendered to text.

Omit section for the outline and general guidance; pass an exact section id from the outline for full details. For local/offline connectors the full generated docs are returned and section is ignored.

Example: outline = await connector.read_skill_docs() details = await connector.read_skill_docs(section="entity:contacts")

PermissionsQuery(connector: GoogleDriveConnector) : Query class for Permissions entity operations.

Initialize query with connector reference.

Methods​

get(self, file_id: str, permission_id: str, supports_all_drives: bool | None = None, use_domain_admin_access: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.Permission : Gets a permission by ID

Args: file_id: The ID of the file permission_id: The ID of the permission supports_all_drives: Whether the requesting application supports both My Drives and shared drives use_domain_admin_access: Issue the request as a domain administrator **kwargs: Additional parameters

Returns: Permission

list(self, file_id: str, page_size: int | None = None, page_token: str | None = None, supports_all_drives: bool | None = None, use_domain_admin_access: bool | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Permission], PermissionsListResultMeta] : Lists a file's or shared drive's permissions

Args: file_id: The ID of the file or shared drive page_size: Maximum number of permissions to return (1-100) page_token: Token for continuing a previous list request supports_all_drives: Whether the requesting application supports both My Drives and shared drives use_domain_admin_access: Issue the request as a domain administrator **kwargs: Additional parameters

Returns: PermissionsListResult

RepliesQuery(connector: GoogleDriveConnector) : Query class for Replies entity operations.

Initialize query with connector reference.

Methods​

get(self, file_id: str, comment_id: str, reply_id: str, include_deleted: bool | None = None, fields: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.Reply : Gets a reply by ID

Args: file_id: The ID of the file comment_id: The ID of the comment reply_id: The ID of the reply include_deleted: Whether to return deleted replies fields: Fields to include in the response (required for replies) **kwargs: Additional parameters

Returns: Reply

list(self, file_id: str, comment_id: str, page_size: int | None = None, page_token: str | None = None, include_deleted: bool | None = None, fields: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Reply], RepliesListResultMeta] : Lists a comment's replies

Args: file_id: The ID of the file comment_id: The ID of the comment page_size: Maximum number of replies to return (1-100) page_token: Token for continuing a previous list request include_deleted: Whether to include deleted replies fields: Fields to include in the response (required for replies) **kwargs: Additional parameters

Returns: RepliesListResult

RevisionsQuery(connector: GoogleDriveConnector) : Query class for Revisions entity operations.

Initialize query with connector reference.

Methods​

get(self, file_id: str, revision_id: str, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.Revision : Gets a revision's metadata by ID

Args: file_id: The ID of the file revision_id: The ID of the revision **kwargs: Additional parameters

Returns: Revision

list(self, file_id: str, page_size: int | None = None, page_token: str | None = None, **kwargs) ‑> airbyte_agent_sdk.connectors.google_drive.models.GoogleDriveExecuteResultWithMeta[list[Revision], RevisionsListResultMeta] : Lists a file's revisions

Args: file_id: The ID of the file page_size: Maximum number of revisions to return (1-1000) page_token: Token for continuing a previous list request **kwargs: Additional parameters

Returns: RevisionsListResult

Was this page helpful?