Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.timbal.ai/llms.txt

Use this file to discover all available pages before exploring further.

Background tasks enable runnables to execute asynchronously, allowing your agent to continue working on other tasks while long-running operations complete in the background. A unique task ID is assigned to each background task, which can be used to check status and retrieve results.

​
Configuring Background Mode

The background_mode parameter controls when a runnable (tool) executes asynchronously. It has three values:
  • "never" (default)
  • "always"
  • "auto": The LLM decides whether to run it in the background or not.
from timbal import Tool

long_running_tool = Tool(
    handler=slow_operation,
    background_mode="always"
)
The built-in Bash tool is an example of a background task. When configured with background_mode="auto", the Agent determines whether to run the shell command in the background or synchronously. 
from timbal import Agent
from timbal.tools import Bash

agent = Agent(
    name="task_manager",
    model="openai/gpt-4o-mini",
    tools=[Bash("*")]
)

# Start a background task
result = await agent(prompt="Install project dependencies in the background").collect()

# Check status later
status_result = await agent(prompt="What is the status of the task?").collect()