Auggie CLI icon

Auggie CLI

Use Auggie CLI to execute AI-powered coding tasks with powerful agentic capabilities

Execute

AI-generated

Summary

Execute a command using Auggie CLI with a specified prompt, within an optional project directory and configurable execution settings.

Inputs

  • Prompt (required) — The instruction or prompt to send to Auggie CLI, supporting expressions to dynamically use previous node data.
  • Project Path — Directory path where Auggie CLI will run; defaults to current working directory if empty.
  • Output Mode — Determines the verbosity and format of Auggie CLI output: 'Print' (default) returns full output, 'Quiet' returns only final output, and 'Compact' returns tool calls, results, and final response each on one line.
  • Model — Selects the AI model to use for processing the prompt, with option between Claude Sonnet 4.5 (default), Claude Sonnet 4, and GPT-5.
  • Timeout — Maximum execution time in seconds before aborting the Auggie CLI command; defaults to 300 seconds.
  • Additional Options — Includes 'Debug Mode' to enable debug logging during execution and 'Custom Arguments' to supply extra command-line arguments to the Auggie CLI command.

Output shape

a single item containing an object with the keys: result (string output from Auggie CLI), success (boolean indicating execution success), duration_ms (execution time in milliseconds), command (full Auggie CLI command run), exitCode (process exit code), and error (error message or null).

The output object is paired with the input item. On failure, if 'Continue On Fail' is enabled, error details include error message, type (e.g., 'timeout'), and stack trace if available; otherwise, the operation throws a detailed node error.

Examples

Example 1: Run Auggie CLI with a Python code generation prompt in default mode

Prompt set to a Python function request; output mode is 'Print'; project path empty to use current directory; model is default Claude Sonnet 4.5; timeout default 300 seconds; debug off; no custom args.

Example 2: Run Auggie CLI in quiet mode with debug logging enabled and a custom working directory

Prompt given via expression; output mode set to 'Quiet'; project path specified; debug mode true for detailed logs; a timeout set per user input; using Claude Sonnet 4 model; custom arguments passed to CLI command.

Discussion