Deep AgentInstall

Summary

Process a user prompt through an advanced AI agent equipped with planning, filesystem access, and subagent delegation capabilities to generate a detailed response and manage tasks.

Inputs

• Source for Prompt (User Message) (required) — Choose whether to use a prompt from a connected Chat Trigger node's 'chatInput' field or define a custom prompt text via expression or static input.
• Prompt (User Message) (required) — The user message text to send to the Deep Agent. Automatically populated from the connected Chat Trigger node if selected, or manually entered if defined by user.
• Options — Settings to configure agent features including enabling filesystem tools, planning todos, subagent spawning, maximum agent steps, custom system behavior message, and verbose logging.
• Custom Subagents — Optional JSON array defining custom subagents with name, description, and system prompts. If left empty, default subagents are used.

Output shape

a single item containing a JSON object with keys 'output' (agent response text), 'todos' (array of todo tasks tracked by the agent), 'files' (list of filenames accessed or modified), 'steps' (number of internal agent steps executed), and 'toolsUsed' (names of tools invoked by the agent).

The output reflects a detailed agent interaction including task planning and tool usage. Errors during execution are returned per-item if node is set to continue on failure.