from pathlib import Path
# Markdown content for Synexs AI System overview
markdown_content = “””
Synexs AI Agent System Overview
body {
font-family: “Segoe UI”, sans-serif;
line-height: 1.6;
max-width: 960px;
margin: 40px auto;
padding: 0 20px;
background-color: #111;
color: #eee;
}
h1, h2 {
color: #82cfff;
}
code {
background-color: #333;
padding: 2px 4px;
border-radius: 4px;
color: #fff176;
}
pre {
background-color: #222;
padding: 10px;
overflow-x: auto;
border-left: 4px solid #82cfff;
}
.cell {
border: 1px solid #444;
padding: 15px;
border-radius: 8px;
margin-bottom: 20px;
background-color: #1a1a1a;
}
Synexs AI Agent System Overview
🧠 Purpose
This document describes the function of each AI “Cell” in the Synexs system, including model training scripts and the autonomous loop controller.
⚙️ Agent Cells (Executed in Loop by cell_008.py)
cell_001.py — Model + Sequence Generator
Defines the LSTM-based symbolic model. Includes generate_symbolic_sequence() for sequence creation. Shared with other cells.
cell_002.py — Batch Sequence Generator
Uses generate_symbolic_sequence() to create 50 symbolic sequences. Stores as JSON in datasets/generated/.
cell_003.py — Refiner
Reads generated sequences, cleans them (e.g., removes <JUNK>), and saves them into datasets/refined/.
cell_004.py — Integrity Logger
Watches the refined folder, hashes files, and logs their fingerprints for tamper detection.
cell_005.py — Pattern Detector
Analyzes refined datasets to count token frequencies. Logs top tokens into pattern_analysis_log.json.
cell_006.py — Decision Core
Loads core_model.pth and classifies refined sequences into actions: discard, refine, replicate, mutate, flag.
cell_007.py — Action Applier
Takes decisions from cell_006 and sends sequences to the appropriate folder based on their classification.
cell_008.py — Autonomous Loop
Executes cell_001 to cell_007 in order every 60 seconds. Fully automates symbolic AI system.
🧪 Model Training Scripts
prepare_training_data.py
Builds labeled training data for symbolic decision model. Output is saved into datasets/core_training/train_data.json.
synexs_core_ai.py
Trains the symbolic classification model. Outputs the file core_model.pth, a trained PyTorch model used by cell_006.py.
📁 Notes
- core_model.pth is a saved PyTorch model file, not a Python script.
- This model is only trained once using
synexs_core_ai.py. cell_006.pyloads it for decision classification.core_model.py(if it exists) may be older or unrelated and can be cleaned up.
✅ System status: Fully Functional
“””
# Save the HTML file
html_path = Path(“/mnt/data/synexs_ai_overview.html”)
html_path.write_text(markdown_content)
html_path.name
Synexs 