State Machine Generator Pipeline

This project generates a State Machine DSL from high-level mission requirements using a 5-step pipeline integrated with Gemma 3.

Prerequisites

• Python 3.8+
• A machine with a GPU is recommended for Gemma 3 inference (steps 2 & 3), but it can run on CPU (slowly).
• Hugging Face Access Token (for Gemma 3 models). You may need to log in via huggingface-cli login.

Installation

1. Clone/Navigate to this directory.
2. Install Dependencies:

   pip install -r requirements.txt

Usage

Run the pipeline steps sequentially:

Step 1: Mission Context

Merges your input (input_mission.json) with the security template.

python3 step1_mission_context.py

Step 2: Canonical Context (Gemma 3)

Uses LLM to analyze the context and documents.

python3 step2_canonical_context.py

Step 3: Overlay Generation (Gemma 3)

Uses LLM to generate the state machine logic with Pydantic validation.

python3 step3_overlay.py

Step 4: Compiler

Compiles and validates the overlay into the final DSL.

python3 step4_compiler.py

Step 5: Runtime Simulation

Simulates the execution of the generated state machine.

python3 step5_runtime.py

Configuration

• Input: Modify input_mission.json to change the mission parameters.
• Templates: Modify templates/generic_template.json for base security settings.
• Documents: Update data/documents.txt to change the knowledge base.