build_tool.py

import subprocess
import time
import urllib.request
import sys

print("1. Starting Ollama server in the background...")
server = subprocess.Popen(["ollama", "serve"])

print("2. Waiting for server to wake up...")
ready = False
for i in range(30): # Give it up to 30 seconds
    try:
        if urllib.request.urlopen("http://127.0.0.1:7860").getcode() == 200:
            ready = True
            break
    except Exception:
        time.sleep(1)

if not ready:
    print("Error: Ollama server failed to start.")
    server.kill()
    sys.exit(1)

print("3. Server is up! Pulling base model and creating quant-agent...")
try:
    # This automatically downloads gemma3:4b and applies your Modelfile
    subprocess.run(["ollama", "create", "quant-agent", "-f", "/app/Tools.modelfile"], check=True)
    print("Model successfully baked!")
except subprocess.CalledProcessError:
    print("Error during model creation.")
    server.kill()
    sys.exit(1)

print("4. Shutting down background server to finalize Docker layer...")
server.terminate()
server.wait()
print("Build script complete. Exiting cleanly.")