Module base 

Base classes and utilities for LangChain tools.

This module provides the core tool abstractions, mirroring langchain_core.tools.base.

Structs

InjectedToolArg

Annotation for tool arguments that are injected at runtime.

InjectedToolCallId

Annotation for injecting the tool call ID.

SchemaAnnotationError

Error raised when args_schema is missing or has incorrect type annotation.

ToolDefinition

Represents a tool’s definition for LLM function calling.

ToolException

Exception thrown when a tool execution error occurs.

Enums

ArgsSchema

Represents a tool’s schema, which can be a JSON schema or a type reference.

HandleToolError

How to handle tool errors.

HandleValidationError

How to handle validation errors.

ResponseFormat

Represents the response format for a tool.

ToolInput

Input type for tools - can be a string, dict, or ToolCall.

ToolOutput

Output type for tools.

Constants

FILTERED_ARGS

Arguments that are filtered out from tool schemas.

TOOL_MESSAGE_BLOCK_TYPES

Block types that are valid in tool messages.

Traits

BaseTool

Base trait for all LangChain tools.

BaseToolkit

Base class for toolkits containing related tools.

Functions

format_output

Format tool output as appropriate.

handle_tool_error_impl

Handle a tool exception based on the configured flag.

handle_validation_error_impl

Handle a validation error based on the configured flag.

is_message_content_block

Check if object is a valid message content block.

is_message_content_type

Check if content is a valid message content type.

is_tool_call

Check if an input is a tool call dictionary.

prep_run_args

Prepare arguments for tool execution.

stringify_content

Convert content to string, preferring JSON format.

Type Aliases

DynTool

Type alias for dynamic tool reference.