revanth-3995/rl-devsecops-openenv-test

RL-Based DevSecOps OpenEnv Environment

🚀 Overview

This project implements a DevSecOps simulation environment compatible with OpenEnv for reinforcement learning-based security decision-making.

It simulates a CI/CD pipeline where an agent performs security-related actions such as detecting secrets, triaging vulnerabilities, and making deployment decisions based on risk levels.

🏗️ Architecture

┌────────────────────────────┐
│      RL Agent / Policy     │
└────────────┬───────────────┘
             │ action
             ▼
┌────────────────────────────┐
│   DevSecOps Gym Env        │
│ (task + severity state)    │
└────────────┬───────────────┘
             │ reward + next state
             ▼
┌────────────────────────────┐
│   FastAPI Server (app.py)  │
│  /reset  /step  /state     │
└────────────┬───────────────┘
             │ HTTP
             ▼
┌────────────────────────────┐
│ Hugging Face Space (Docker)│
└────────────────────────────┘

📂 Project Structure

• —devsecops_env.py: Core Gymnasium environment defining state, actions, and rewards for the DevSecOps simulation.
• —app.py: FastAPI application exposing the environment via HTTP endpoints (/reset, /step, /state).
• —inference.py: Dummy inference script to demonstrate how an agent interacts with the environment.
• —openenv.yaml: OpenEnv metadata specification.
• —Dockerfile: Containerizes the FastAPI server for easy deployment (e.g., Hugging Face Spaces).

🧠 Tasks Implemented

1. Secret Scanning

• —Detect exposed credentials in code
• —Actions: DETECT, REPORT, APPROVE

2. CVE Triage

• —Analyze vulnerability severity and prioritize fixes
• —Actions: BLOCK, PATCH, ESCALATE, APPROVE

3. Pipeline Security Audit

• —Multi-stage decision-making across pipeline stages
• —Actions: BLOCK, PATCH, ESCALATE, APPROVE, ROLLBACK

⚙️ Action Space

["DETECT", "REPORT", "BLOCK", "PATCH", "ESCALATE", "APPROVE", "ROLLBACK"]

📊 Observation Space

{
  "task": "int (0–2)",
  "severity": "float (1–10)"
}

🛠️ Installation & Usage

1. Local Setup

Create a virtual environment and install dependencies:

python -m venv venv
source venv/bin/activate  # On Windows use: venv\Scripts\activate
pip install -r requirements.txt

2. Running FastAPI Server

The application exposes a REST API to interact with the environment:

uvicorn app:app --host 0.0.0.0 --port 7860

Endpoints:

• —POST /reset: Resets the environment state.
• —POST /step: Takes an action (e.g., {"action": 0}) and returns the next state, reward, and done flag.
• —GET /state: Retrieves the current state.

3. Running Dummy Inference

Test the environment locally using the dummy agent:

python inference.py

4. Docker Deployment

You can also run this environment via Docker:

docker build -t devsecops-env .
docker run -p 7860:7860 devsecops-env