Function●Since v0.0

create_cli_agent

Create a CLI-configured agent with flexible options.

This is the main entry point for creating a deepagents CLI agent, usable both internally and from external code (e.g., benchmarking frameworks).

create_cli_agent(
  model: str | BaseChatModel,
  assistant_id: str,
  *,
  tools: Sequence[BaseTool | Callable | dict[str, Any]] | None = None,
  sandbox: SandboxBackendProtocol | None = None,
  sandbox_type: str | None = None,
  system_prompt: str | None = None,
  interactive: bool = True,
  auto_approve: bool = False,
  interrupt_shell_only: bool = False,
  shell_allow_list: list[str] | None = None,
  enable_ask_user: bool = True,
  enable_memory: bool = True,
  enable_skills: bool = True,
  enable_shell: bool = True,
  checkpointer: BaseCheckpointSaver | None = None,
  mcp_server_info: list[MCPServerInfo] | None = None,
  cwd: str | Path | None = None,
  project_context: ProjectContext | None = None,
  async_subagents: list[AsyncSubAgent] | None = None
) -> tuple[Pregel, CompositeBackend]

Parameters

Name	Type	Description
`model`*	`str \| BaseChatModel`	LLM model to use (e.g., `'anthropic:claude-sonnet-4-6'`)
`assistant_id`*	`str`	Agent identifier for memory/state storage
`tools`	`Sequence[BaseTool \| Callable \| dict[str, Any]] \| None`	Default:`None` Additional tools to provide to agent
`sandbox`	`SandboxBackendProtocol \| None`	Default:`None` Optional sandbox backend for remote execution (e.g., `ModalSandbox`). If `None`, uses local filesystem + shell.
`sandbox_type`	`str \| None`	Default:`None` Type of sandbox provider (`'agentcore'`, `'daytona'`, `'langsmith'`, `'modal'`, `'runloop'`). Used for system prompt generation.
`system_prompt`	`str \| None`	Default:`None` Override the default system prompt. If `None`, generates one based on `sandbox_type`, `assistant_id`, and `interactive`.
`interactive`	`bool`	Default:`True` When `False`, the auto-generated system prompt is tailored for headless non-interactive execution. Ignored when `system_prompt` is provided explicitly.
`auto_approve`	`bool`	Default:`False` If `True`, no tools trigger human-in-the-loop interrupts — all calls (shell execution, file writes/edits, web search, URL fetch) run automatically. If `False`, tools pause for user confirmation via the approval menu. See `_add_interrupt_on` for the full list of gated tools.
`interrupt_shell_only`	`bool`	Default:`False` If `True`, all HITL interrupts are disabled; shell commands are validated inline by `ShellAllowListMiddleware` against the configured allow-list instead. Used in non-interactive mode with a restrictive shell allow-list to avoid splitting traces into multiple LangSmith runs. Has no effect when `auto_approve` is `True` (interrupts are already disabled) or when `shell_allow_list` is `SHELL_ALLOW_ALL`.
`shell_allow_list`	`list[str] \| None`	Default:`None` Explicit restrictive shell allow-list forwarded from the CLI process. When provided (and `interrupt_shell_only` is `True`), used directly instead of reading `settings.shell_allow_list` (which may not be set in the server subprocess environment).
`enable_ask_user`	`bool`	Default:`True` Enable `AskUserMiddleware` so the agent can ask clarifying questions. Disabled in non-interactive mode.
`enable_memory`	`bool`	Default:`True` Enable `MemoryMiddleware` for persistent memory
`enable_skills`	`bool`	Default:`True` Enable `SkillsMiddleware` for custom agent skills
`enable_shell`	`bool`	Default:`True` Enable shell execution via `LocalShellBackend` (only in local mode). When enabled, the `execute` tool is available.
`checkpointer`	`BaseCheckpointSaver \| None`	Default:`None` Optional checkpointer for session persistence. When `None`, the graph is compiled without a checkpointer.
`mcp_server_info`	`list[MCPServerInfo] \| None`	Default:`None` MCP server metadata to surface in the system prompt.
`cwd`	`str \| Path \| None`	Default:`None` Override the working directory for the agent's filesystem backend and system prompt.
`project_context`	`ProjectContext \| None`	Default:`None` Explicit project path context for project-sensitive behavior such as project `AGENTS.md` files, skills, subagents, and MCP trust.
`async_subagents`	`list[AsyncSubAgent] \| None`	Default:`None` Remote LangGraph deployments to expose as async subagent tools. Loaded from `[async_subagents]` in `config.toml` or passed directly.

View source on GitHub

create_cli_agent( model: str | BaseChatModel, assistant_id: str, *, tools: Sequence[BaseTool | Callable | dict[str, Any]] | None = None, sandbox: SandboxBackendProtocol | None = None, sandbox_type: str | None = None, system_prompt: str | None = None, interactive: bool = True, auto_approve: bool = False, interrupt_shell_only: bool = False, shell_allow_list: list[str] | None = None, enable_ask_user: bool = True, enable_memory: bool = True, enable_skills: bool = True, enable_shell: bool = True, checkpointer: BaseCheckpointSaver | None = None, mcp_server_info: list[MCPServerInfo] | None = None, cwd: str | Path | None = None, project_context: ProjectContext | None = None, async_subagents: list[AsyncSubAgent] | None = None ) -> tuple[Pregel, CompositeBackend]

Parameters

Name	Type	Description
`model`*	`str \| BaseChatModel`	LLM model to use (e.g., `'anthropic:claude-sonnet-4-6'`)
`assistant_id`*	`str`	Agent identifier for memory/state storage
`tools`	`Sequence[BaseTool \| Callable \| dict[str, Any]] \| None`	Default:`None` Additional tools to provide to agent
`sandbox`	`SandboxBackendProtocol \| None`	Default:`None` Optional sandbox backend for remote execution (e.g., `ModalSandbox`). If `None`, uses local filesystem + shell.
`sandbox_type`	`str \| None`	Default:`None` Type of sandbox provider (`'agentcore'`, `'daytona'`, `'langsmith'`, `'modal'`, `'runloop'`). Used for system prompt generation.
`system_prompt`	`str \| None`	Default:`None` Override the default system prompt. If `None`, generates one based on `sandbox_type`, `assistant_id`, and `interactive`.
`interactive`	`bool`	Default:`True` When `False`, the auto-generated system prompt is tailored for headless non-interactive execution. Ignored when `system_prompt` is provided explicitly.
`auto_approve`	`bool`	Default:`False` If `True`, no tools trigger human-in-the-loop interrupts — all calls (shell execution, file writes/edits, web search, URL fetch) run automatically. If `False`, tools pause for user confirmation via the approval menu. See `_add_interrupt_on` for the full list of gated tools.
`interrupt_shell_only`	`bool`	Default:`False` If `True`, all HITL interrupts are disabled; shell commands are validated inline by `ShellAllowListMiddleware` against the configured allow-list instead. Used in non-interactive mode with a restrictive shell allow-list to avoid splitting traces into multiple LangSmith runs. Has no effect when `auto_approve` is `True` (interrupts are already disabled) or when `shell_allow_list` is `SHELL_ALLOW_ALL`.
`shell_allow_list`	`list[str] \| None`	Default:`None` Explicit restrictive shell allow-list forwarded from the CLI process. When provided (and `interrupt_shell_only` is `True`), used directly instead of reading `settings.shell_allow_list` (which may not be set in the server subprocess environment).
`enable_ask_user`	`bool`	Default:`True` Enable `AskUserMiddleware` so the agent can ask clarifying questions. Disabled in non-interactive mode.
`enable_memory`	`bool`	Default:`True` Enable `MemoryMiddleware` for persistent memory
`enable_skills`	`bool`	Default:`True` Enable `SkillsMiddleware` for custom agent skills
`enable_shell`	`bool`	Default:`True` Enable shell execution via `LocalShellBackend` (only in local mode). When enabled, the `execute` tool is available.
`checkpointer`	`BaseCheckpointSaver \| None`	Default:`None` Optional checkpointer for session persistence. When `None`, the graph is compiled without a checkpointer.
`mcp_server_info`	`list[MCPServerInfo] \| None`	Default:`None` MCP server metadata to surface in the system prompt.
`cwd`	`str \| Path \| None`	Default:`None` Override the working directory for the agent's filesystem backend and system prompt.
`project_context`	`ProjectContext \| None`	Default:`None` Explicit project path context for project-sensitive behavior such as project `AGENTS.md` files, skills, subagents, and MCP trust.
`async_subagents`	`list[AsyncSubAgent] \| None`	Default:`None` Remote LangGraph deployments to expose as async subagent tools. Loaded from `[async_subagents]` in `config.toml` or passed directly.

create_cli_agent

Parameters

LangChain Assistant

Menu

create_cli_agent

Parameters