LodgeiT

ClawDog Quickstart

The easiest way to run the ClawDog neurosemantic pipeline is via MyClaw. No local installation or terminal required.

✨ Optimal AI: Gemini 3

For the best experience running the ClawDog neurosemantic engine, we highly recommend using Google Gemini 3.1 Pro (or newer). It offers the exceptional reasoning capabilities and large context windows required for parsing complex accounting taxonomies.

How to get your Gemini API Key:

  1. Visit Google AI Studio.
  2. Sign in with your Google account.
  3. Click "Get API key" in the left menu and create a new key.
  4. Provide this key to your MyClaw instance when configuring your primary AI model.

1 Zero-Setup Execution

Since MyClaw is an autonomous agent running in its own pre-configured Ubuntu environment (which already has Python and SWI-Prolog installed), you can just talk to it. Paste this "Magic Prompt" into your MyClaw chat:

"Hey MyClaw, please clone https://github.com/lodgeit-labs/clawdog, install the dependencies from requirements.txt, and run pipeline.py. Tell me when the thermodynamic audit is complete and what files were generated."

What happens next?

  • MyClaw will automatically run git clone, execute pip install -r requirements.txt, and run the pipeline.
  • It will validate the sample trial balances using the Prolog physics engine.
  • It will report back to you when the JSON-LD and iXBRL outputs are ready in the outputs/ folder.

2 Map Your Own Data

Once the initial run is complete, the engine is ready. You can simply ask MyClaw to map a new CSV for you.

"I have a new trial balance CSV. Can you map the accounts using the heuristic mapper and run it through the ClawDog pipeline for me?"