Tutorial: LLM-Assisted Command (E07)

This tutorial corresponds to the example file examples/E07_run_command_llm_assisted_example.py.

It demonstrates the core agentic capability of using an LLM to interpret a natural language command. It shows how to: - Configure the llm_assisted command processor. - Configure an embedding-based tool_lookup service to help the LLM find relevant tools. - Use genie.run_command(), which will leverage the LLM to select the correct tool and extract its parameters from the user's sentence, all in one step.

Example Code

examples/E07_run_command_llm_assisted_example.py

""" Example: genie.run_command() with LLM-Assisted Processor

---

This example demonstrates using genie.run_command() with the 'llm_assisted' command processor, using Ollama as the backing LLM.

To Run: 1. Ensure Genie Tooling is installed (poetry install --all-extras). 2. Ensure Ollama is running and 'mistral:latest' (or your chosen model) is pulled. 3. Run from the root of the project: poetry run python examples/E07_run_command_llm_assisted_example.py """ import asyncio import json import logging from typing import Optional

from genie_tooling.config.features import FeatureSettings from genie_tooling.config.models import MiddlewareConfig from genie_tooling.genie import Genie

async def run_llm_assisted_command_demo(): print("--- LLM-Assisted Command Processor Demo (Ollama) ---")

app_config = MiddlewareConfig(
    features=FeatureSettings(
        llm="ollama",
        llm_ollama_model_name="mistral:latest",
        command_processor="llm_assisted",
        # Use default compact_text_formatter for command_processor_formatter_id_alias
        # Use default embedding tool lookup with default ST embedder and compact_text_formatter
        tool_lookup="embedding",
    ),
    # Optional: Configure tool_lookup_top_k for the LLM-assisted processor
    command_processor_configurations={
        "llm_assisted_tool_selection_processor_v1": { # Canonical ID
            "tool_lookup_top_k": 3
        }
    },
    tool_configurations={
        # Ensure tools the LLM might select are enabled
        "calculator_tool": {}
    }
)

genie: Optional[Genie] = None
try:
    print("\nInitializing Genie...")
    genie = await Genie.create(config=app_config)
    print("Genie initialized!")

    command_text = "What is 123 multiplied by 4?"
    print(f"\nRunning command: '{command_text}'")

    command_result = await genie.run_command(command_text)

    print("\nCommand Result:")
    if command_result:
        print(json.dumps(command_result, indent=2, default=str))
        if command_result.get("tool_result", {}).get("error_message"):
             print(f"Tool Execution Error: {command_result['tool_result']['error_message']}")
        elif command_result.get("error"):
             print(f"Command Processing Error: {command_result['error']}")
    else:
        print("Command did not produce a result.")

except Exception as e:
    print(f"\nAn error occurred: {e}")
    logging.exception("Error details:")
finally:
    if genie:
        await genie.close()
        print("\nGenie torn down.")

if name == "main": logging.basicConfig(level=logging.INFO) # logging.getLogger("genie_tooling").setLevel(logging.DEBUG) asyncio.run(run_llm_assisted_command_demo())