Implementation:Openai Openai python Response Function Shell Tool Call

Knowledge Sources	Openai_Openai_python OpenAI API Reference
Domains	API_Types, Responses_API
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete response type representing a shell tool call that executes shell commands in a managed environment, provided by the openai-python SDK.

Description

ResponseFunctionShellToolCall is a Pydantic model representing a tool call that executes one or more shell commands in a managed environment. It contains an id for the call, an action field (an Action sub-model with a list of commands, optional max_output_length, and optional timeout_ms), a call_id for response mapping, an optional environment (a discriminated union of ResponseLocalEnvironment or ResponseContainerReference), a status field, a fixed type of "shell_call", and an optional created_by field. This type is auto-generated from the OpenAI OpenAPI specification by Stainless.

Usage

Import ResponseFunctionShellToolCall when processing Responses API output items that contain shell tool calls and you need to execute the requested commands and return results.

Code Reference

Source Location

• Repository: openai-python
• File: src/openai/types/responses/response_function_shell_tool_call.py

Signature

class ResponseFunctionShellToolCall(BaseModel):
    """A tool call that executes one or more shell commands in a managed environment."""

    id: str
    action: Action
    call_id: str
    environment: Optional[Environment] = None
    status: Literal["in_progress", "completed", "incomplete"]
    type: Literal["shell_call"]
    created_by: Optional[str] = None

Import

from openai.types.responses import ResponseFunctionShellToolCall

I/O Contract

Fields

Name	Type	Required	Description
id	str	Yes	The unique ID of the shell tool call.
action	Action	Yes	The shell commands and limits describing how to run the tool call.
call_id	str	Yes	The unique ID of the shell tool call generated by the model.
environment	Optional[Union[ResponseLocalEnvironment, ResponseContainerReference]]	No	The execution environment for the shell commands.
status	Literal["in_progress", "completed", "incomplete"]	Yes	The status of the shell call.
type	Literal["shell_call"]	Yes	The type of the item. Always "shell_call".
created_by	Optional[str]	No	The ID of the entity that created this tool call.

Action Sub-Type

Name	Type	Required	Description
commands	List[str]	Yes	The shell commands to execute.
max_output_length	Optional[int]	No	Maximum number of characters to return from each command.
timeout_ms	Optional[int]	No	Timeout in milliseconds for the commands.

Usage Examples

from openai.types.responses import ResponseFunctionShellToolCall

# Process shell tool calls from response output
for item in response.output:
    if isinstance(item, ResponseFunctionShellToolCall):
        print(f"Shell call ID: {item.call_id}")
        print(f"Commands: {item.action.commands}")
        if item.action.timeout_ms:
            print(f"Timeout: {item.action.timeout_ms}ms")

        # Execute the commands and provide output back
        for cmd in item.action.commands:
            result = subprocess.run(cmd, shell=True, capture_output=True, text=True)
            print(f"stdout: {result.stdout}")

Related Pages

• Environment:Openai_Openai_python_Python_3_9_Plus