Get LLM instance by type. Returns cached instance if available.

Build and return the agent workflow graph.

Factory function to create a logged version of any tool class. Args: base_tool_class: The original tool class to be enhanced with lo

Run the agent workflow with the given user input. Args: user_input: The user's query or request debug: If True, enables debug lev

Create a ChatOpenAI instance with the specified configuration

Enable debug level logging for more detailed execution information.

Helper method to log tool operations.

Create a ChatDeepSeek instance with the specified configuration

Enable debug level logging for more detailed execution information.

Run the agent workflow with the given user input. Args: user_input_messages: The user request messages debug: If True, enables de

Run the browser task asynchronously.

Override _run method to add logging.

Run the browser task synchronously.

Use this to execute bash command and do necessary operations.

Node for the browser agent that performs web browsing tasks.

Chat endpoint for LangGraph invoke. Args: request: The chat request req: The FastAPI request object for connection state che

Node for the coder agent that executes Python code.

Coordinator node that communicate with customers.

Use this to crawl a url and get a readable content in markdown format.

A decorator that logs the input parameters and output of a tool function. Args: func: The tool function to be decorated Returns

Planner node that generate the full plan.

Use this to execute python code and do data analysis or calculation. If you want to see the output of a value, you should print it out with `print

Reporter node that write a final report.

Node for the researcher agent that performs research tasks.

Supervisor node that decides which agent should act next.

Test bash tool when command fails

Test bash tool when an unexpected exception occurs

Test bash tool with a command that produces output

Test crawling with a valid URL.

Test that crawler can be properly initialized.

Test that crawler output can be converted to markdown.

Test that debug logging is properly enabled.

Test basic workflow execution.

Test workflow execution with empty input.

Test bash tool with a successful command execution