Function●Since v0.0

run_textual_app

Run the Textual application.

When server_kwargs is provided (and agent is None), the app starts immediately with a "Connecting..." banner and launches the server in the background. Server cleanup is handled automatically after the app exits.

run_textual_app(
  *,
  agent: Any = None,
  assistant_id: str | None = None,
  backend: CompositeBackend | None = None,
  auto_approve: bool = False,
  cwd: str | Path | None = None,
  thread_id: str | None = None,
  resume_thread: str | None = None,
  initial_prompt: str | None = None,
  initial_skill: str | None = None,
  mcp_server_info: list[MCPServerInfo] | None = None,
  profile_override: dict[str, Any] | None = None,
  server_proc: ServerProcess | None = None,
  server_kwargs: dict[str, Any] | None = None,
  mcp_preload_kwargs: dict[str, Any] | None = None,
  model_kwargs: dict[str, Any] | None = None
) -> AppResult

Parameters

Name	Type	Description
`agent`	`Any`	Default:`None` Pre-configured LangGraph agent (optional).
`assistant_id`	`str \| None`	Default:`None` Agent identifier for memory storage.
`backend`	`CompositeBackend \| None`	Default:`None` Backend for file operations.
`auto_approve`	`bool`	Default:`False` Whether to start with auto-approve enabled.
`cwd`	`str \| Path \| None`	Default:`None` Current working directory to display.
`thread_id`	`str \| None`	Default:`None` Thread ID for the session. `None` when `resume_thread` is provided (the TUI resolves the final ID asynchronously).
`resume_thread`	`str \| None`	Default:`None` Raw resume intent from `-r` flag. `'__MOST_RECENT__'` for bare `-r`, a thread ID string for `-r <id>`, or `None` for new sessions. Resolved asynchronously during TUI startup.
`initial_prompt`	`str \| None`	Default:`None` Optional prompt to auto-submit when session starts.
`initial_skill`	`str \| None`	Default:`None` Optional skill name to invoke when session starts.
`mcp_server_info`	`list[MCPServerInfo] \| None`	Default:`None` MCP server metadata for the `/mcp` viewer.
`profile_override`	`dict[str, Any] \| None`	Default:`None` Extra profile fields from `--profile-override`, retained so later profile-aware behavior stays consistent with the CLI override, including model selection details, offload budget display, and on-demand `create_model()` calls such as `/offload`.
`server_proc`	`ServerProcess \| None`	Default:`None` LangGraph server process for the interactive session.
`server_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for deferred `start_server_and_get_agent` call.
`mcp_preload_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for concurrent MCP metadata preload.
`model_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for deferred `create_model()` call. When provided, model creation runs in a background worker after first paint so the splash screen appears immediately.

View source on GitHub

Parameters

Name	Type	Description
`agent`	`Any`	Default:`None` Pre-configured LangGraph agent (optional).
`assistant_id`	`str \| None`	Default:`None` Agent identifier for memory storage.
`backend`	`CompositeBackend \| None`	Default:`None` Backend for file operations.
`auto_approve`	`bool`	Default:`False` Whether to start with auto-approve enabled.
`cwd`	`str \| Path \| None`	Default:`None` Current working directory to display.
`thread_id`	`str \| None`	Default:`None` Thread ID for the session. `None` when `resume_thread` is provided (the TUI resolves the final ID asynchronously).
`resume_thread`	`str \| None`	Default:`None` Raw resume intent from `-r` flag. `'__MOST_RECENT__'` for bare `-r`, a thread ID string for `-r <id>`, or `None` for new sessions. Resolved asynchronously during TUI startup.
`initial_prompt`	`str \| None`	Default:`None` Optional prompt to auto-submit when session starts.
`initial_skill`	`str \| None`	Default:`None` Optional skill name to invoke when session starts.
`mcp_server_info`	`list[MCPServerInfo] \| None`	Default:`None` MCP server metadata for the `/mcp` viewer.
`profile_override`	`dict[str, Any] \| None`	Default:`None` Extra profile fields from `--profile-override`, retained so later profile-aware behavior stays consistent with the CLI override, including model selection details, offload budget display, and on-demand `create_model()` calls such as `/offload`.
`server_proc`	`ServerProcess \| None`	Default:`None` LangGraph server process for the interactive session.
`server_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for deferred `start_server_and_get_agent` call.
`mcp_preload_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for concurrent MCP metadata preload.
`model_kwargs`	`dict[str, Any] \| None`	Default:`None` Kwargs for deferred `create_model()` call. When provided, model creation runs in a background worker after first paint so the splash screen appears immediately.

run_textual_app

Parameters

LangChain Assistant

Menu

run_textual_app

Parameters