dial_langchain - Dial LangChain Tools

class dial_langchain.DialToolkit¶

Bases: langchain_core.tools.BaseToolkit

All Dial tools in one bundle — phone numbers, SMS/WhatsApp, and voice calls.

Build one DialClient and construct the toolkit with it, then call get_tools() to hand the full set to an agent. Every tool shares that one client — a single connection pool for the whole agent:

from dial_sdk import DialClient, DialConfig
from dial_langchain import DialToolkit
from langchain.chat_models import init_chat_model
from langgraph.prebuilt import create_react_agent

model = init_chat_model("claude-sonnet-4-6", model_provider="anthropic")
client = DialClient(DialConfig(api_key="sk_live_..."))
toolkit = DialToolkit(client=client)
agent = create_react_agent(model, toolkit.get_tools())

get_tools() → list[langchain_core.tools.BaseTool]¶: Return every Dial tool, each sharing this toolkit’s client.

client: dial_sdk.DialClient¶

model_config¶

class dial_langchain.GetBillingTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

description: str = "Get the account's billing status: credit balance, subscription (if any), each number's billing...¶

name: str = 'get_billing'¶

class dial_langchain.GetCallTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Fetch a single call by its id, including its current status.'¶

name: str = 'get_call'¶

class dial_langchain.ListCallsTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'List recent inbound and outbound calls on the account.'¶

name: str = 'list_calls'¶

class dial_langchain.ListMessagesTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'List recent inbound and outbound messages on the account.'¶

name: str = 'list_messages'¶

class dial_langchain.ListNumbersTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

description: str = 'List all available Dial phone numbers on the account.'¶

name: str = 'list_numbers'¶

class dial_langchain.MakeCallTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Initiate an AI voice call to a phone number.'¶

name: str = 'make_call'¶

class dial_langchain.PurchaseNumberTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Provision (buy) a new Dial phone number. Billable.'¶

name: str = 'purchase_number'¶

class dial_langchain.ReplyToMessageTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Reply in-thread or react to an existing message. The reply goes out from the Dial number the...¶

name: str = 'reply_to_message'¶

class dial_langchain.SendMessageTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Send a message to a phone number, optionally with media attachments (MMS). On an iMessage...¶

name: str = 'send_message'¶

class dial_langchain.SetNumberPropertiesTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = "Update a Dial phone number's properties: its inbound instruction (the system prompt its AI...¶

name: str = 'set_number_properties'¶

class dial_langchain.StartTypingTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = "Show a typing indicator to the recipient, as if composing a message from your number. iMessage...¶

name: str = 'start_typing'¶

class dial_langchain.StopTypingTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = "Clear a typing indicator previously shown with start_typing. Delivering a message or reaction...¶

name: str = 'stop_typing'¶

class dial_langchain.WaitForMessageTool¶

Bases: _DialTool

Base for Dial LangChain tools.

Holds a shared DialClient (passed in, not constructed per call) so every tool reuses one connection pool. These tools are async-only: LangChain/LangGraph agents invoke them via _arun in their event loop, so the shared client is safe. arbitrary_types_allowed lets the Pydantic-based BaseTool carry the non-Pydantic DialClient as a field.

args_schema: type[pydantic.BaseModel]¶

description: str = 'Wait for the next inbound SMS or WhatsApp message and return it. Blocks until a message arrives...¶

name: str = 'wait_for_message'¶

dial_langchain¶

Submodules¶

Classes¶

Package Contents¶

`DialToolkit`	All Dial tools in one bundle — phone numbers, SMS/WhatsApp, and voice calls.
`GetBillingTool`	Base for Dial LangChain tools.
`GetCallTool`	Base for Dial LangChain tools.
`ListCallsTool`	Base for Dial LangChain tools.
`ListMessagesTool`	Base for Dial LangChain tools.
`ListNumbersTool`	Base for Dial LangChain tools.
`MakeCallTool`	Base for Dial LangChain tools.
`PurchaseNumberTool`	Base for Dial LangChain tools.
`ReplyToMessageTool`	Base for Dial LangChain tools.
`SendMessageTool`	Base for Dial LangChain tools.
`SetNumberPropertiesTool`	Base for Dial LangChain tools.
`StartTypingTool`	Base for Dial LangChain tools.
`StopTypingTool`	Base for Dial LangChain tools.
`WaitForMessageTool`	Base for Dial LangChain tools.