motleycrew.tools

MotleyTool class and tools library.

Modules

autogen_chat_tool

code

html_render_tool

image

llm_tool

mermaid_evaluator_tool

tool

class motleycrew.tools.MotleyTool(tool: BaseTool | None = None, name: str | None = None, description: str | None = None, args_schema: BaseModel | None = None, return_direct: bool = False, exceptions_to_reflect: List[Type[Exception]] | None = None, retry_config: RetryConfig | None = None)

Bases: Runnable

Base tool class compatible with MotleyAgents.

It is a wrapper for Langchain BaseTool, containing all necessary adapters and converters.

__init__(tool: BaseTool | None = None, name: str | None = None, description: str | None = None, args_schema: BaseModel | None = None, return_direct: bool = False, exceptions_to_reflect: List[Type[Exception]] | None = None, retry_config: RetryConfig | None = None)

Initialize the MotleyTool.

Parameters:

  • name – Name of the tool (required if tool is None).

  • description – Description of the tool (required if tool is None).

  • args_schema – Schema of the tool arguments (required if tool is None).

  • tool – Langchain BaseTool to wrap.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

property name

Name of the tool.

property description

Description of the tool.

property args_schema

Schema of the tool arguments.

invoke(input: str | Dict, config: RunnableConfig | None = None, **kwargs: Any) → Any

Transform a single input into an output. Override to implement.

Parameters:

  • input – The input to the Runnable.

  • config – A config to use when invoking the Runnable. The config supports standard keys like ‘tags’, ‘metadata’ for tracing purposes, ‘max_concurrency’ for controlling how much work to do in parallel, and other keys. Please refer to the RunnableConfig for more details.

Returns:

The output of the Runnable.

run(*args, **kwargs)
static from_langchain_tool(langchain_tool: BaseTool, return_direct: bool = False, exceptions_to_reflect: List[Exception] | None = None, retry_config: RetryConfig | None = None) → MotleyTool

Create a MotleyTool from a Langchain tool.

Parameters:

  • langchain_tool – Langchain tool to convert.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

Returns:

MotleyTool instance.

static from_llama_index_tool(llama_index_tool: None, return_direct: bool = False, exceptions_to_reflect: List[Exception] | None = None, retry_config: RetryConfig | None = None) → MotleyTool

Create a MotleyTool from a LlamaIndex tool.

Parameters:

  • llama_index_tool – LlamaIndex tool to convert.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

Returns:

MotleyTool instance.

static from_crewai_tool(crewai_tool: None, return_direct: bool = False, exceptions_to_reflect: List[Exception] | None = None, retry_config: RetryConfig | None = None) → MotleyTool

Create a MotleyTool from a CrewAI tool.

Parameters:

  • crewai_tool – CrewAI tool to convert.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

Returns:

MotleyTool instance.

static from_motley_agent(agent: MotleyAgentAbstractParent, return_direct: bool = False, exceptions_to_reflect: List[Exception] | None = None, retry_config: RetryConfig | None = None) → MotleyTool

Convert an agent to a tool to be used by other agents via delegation.

Parameters:

  • agent – The MotleyAgent to convert to a tool.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

Returns:

The tool representation of the agent.

static from_supported_tool(tool: MotleyTool | BaseTool | None | MotleyAgentAbstractParent, return_direct: bool = False, exceptions_to_reflect: List[Exception] | None = None, retry_config: RetryConfig | None = None) → MotleyTool

Create a MotleyTool from any supported tool type.

Parameters:

  • tool – Tool of any supported type. Currently, we support tools from Langchain, LlamaIndex, as well as motleycrew agents.

  • return_direct – If True, the tool’s output will be returned directly to the user.

  • exceptions_to_reflect – List of exceptions to reflect back to the agent.

  • retry_config – Configuration for retry behavior. If None, exceptions will not be retried.

Returns:

MotleyTool instance.

to_langchain_tool() → BaseTool

Convert the MotleyTool to a Langchain tool.

Returns:

Langchain tool.

to_llama_index_tool() → None

Convert the MotleyTool to a LlamaIndex tool.

Returns:

LlamaIndex tool.

to_autogen_tool() → Callable

Convert the MotleyTool to an AutoGen tool.

An AutoGen tool is basically a function. AutoGen infers the tool input schema from the function signature. For this reason, because we can’t generate the signature dynamically, we can only convert tools with a single input field.

Returns:

AutoGen tool function.

to_crewai_tool() → None

Description

Return type:

Crewai__BaseTool

class motleycrew.tools.RetryConfig(max_retries: int = 3, wait_time: float = 1.0, backoff_factor: float = 2.0, exceptions_to_retry: ~typing.Tuple[~typing.Type[Exception]] = (<class 'Exception'>, ))

Bases: object

Configuration for retry behavior of MotleyTool.

max_retries

Maximum number of retry attempts.

Type:

int

wait_time

Base wait time between retries in seconds.

Type:

float

backoff_factor

Multiplicative factor for exponential backoff.

Type:

float

exceptions_to_retry

Exceptions that should trigger a retry.

Type:

List[Type[Exception]]

max_retries: int = 3
wait_time: float = 1.0
backoff_factor: float = 2.0
exceptions_to_retry: Tuple[Type[Exception]] = (<class 'Exception'>,)
__init__(max_retries: int = 3, wait_time: float = 1.0, backoff_factor: float = 2.0, exceptions_to_retry: ~typing.Tuple[~typing.Type[Exception]] = (<class 'Exception'>, )) → None
exception motleycrew.tools.DirectOutput(output: Any)

Bases: BaseException

Auxiliary exception to return a tool’s output directly.

When the tool returns an output, this exception is raised with the output. It is then handled by the agent, who should gracefully return the output to the user.

__init__(output: Any)