Feature/start services

.vscode/settings.json

@@ -18,5 +18,10 @@
     }
   },
   "python.terminal.activateEnvironment": false,
-  "python.terminal.activateEnvInCurrentTerminal": false
+  "python.terminal.activateEnvInCurrentTerminal": false,
+  "workbench.editorAssociations": {
+    "*.copilotmd": "vscode.markdown.preview.editor",
+    "*.docx": "default",
+    "file:/**/*.csv": "jupyter-data-wrangler"
+  }
 }

Readme.md

@@ -53,7 +53,7 @@ A complete machine learning operations (ML Ops) system demonstrating the full li
                                               │
                                               ▼
                            ┌──────────────────────────────────────────┐
-                           │      GATEWAY SERVICE (:8000)             │
+                           │   GATEWAY SERVICE (:8080 → :8000)       │
                            │  ┌────────────────────────────────────┐  │
                            │  │  1. Whitelist Check (15 domains)   │  │
                            │  │     ├─ google.com, github.com      │  │
@@ -87,7 +87,7 @@ A complete machine learning operations (ML Ops) system demonstrating the full li
                                               │
                                               ▼
                            ┌──────────────────────────────────────────┐
-                           │         MODEL SERVICE (:9000)            │
+                           │         MODEL SERVICE (:8002)            │
                            │  ┌────────────────────────────────────┐  │
                            │  │  • Feature Extraction (8 features) │  │
                            │  │  • XGBoost Inference               │  │
@@ -714,6 +714,74 @@ Final decision: ALLOW
 - **LEAN_PHISH:** Likely phishing → Final decision: BLOCK
 - **UNCERTAIN:** Unclear → Final decision: REVIEW (manual review queue)
 
+
+### How to use the optional LLM (Ollama)
+
+#### STEP 1: Verify Ollama is Running
+```
+# Check if Ollama service is running
+ollama list
+# Should show: llama3.2:1b
+
+# Test Ollama directly
+ollama run llama3.2:1b
+# Try: "Is google.com a phishing URL?"
+# If it responds, Ollama is working!
+# Type /bye to exit
+
+# Check Ollama API endpoint
+curl http://localhost:11434/api/tags
+# Should return JSON with model list
+```
+- If Ollama isn't running:
+
+```
+# Start Ollama service
+ollama serve
+# Keep this running in a terminal
+```
+
+#### STEP 2: Configure PhishGuardAI to Use Ollama
+
+
+##### Option A: Environment Variables
+
+```
+# .env file
+
+# Enable LLM judge (instead of stub)
+JUDGE_BACKEND=llm
+
+# Ollama API endpoint (default is localhost:11434)
+OLLAMA_BASE_URL=http://localhost:11434
+
+# Model name (must match your installed model)
+OLLAMA_MODEL=llama3.2:1b
+
+# Judge timeout (seconds)
+JUDGE_TIMEOUT=10
+
+# Verbose logging (to see judge calls)
+LOG_LEVEL=DEBUG
+```
+
+##### Option B: Set Environment Variables Directly
+- **Windows:**
+```
+set JUDGE_BACKEND=llm
+set OLLAMA_BASE_URL=http://localhost:11434
+set OLLAMA_MODEL=llama3.2:1b
+set LOG_LEVEL=DEBUG
+```
+
+- **Linux/Mac:**
+```
+export JUDGE_BACKEND=llm
+export OLLAMA_BASE_URL=http://localhost:11434
+export OLLAMA_MODEL=llama3.2:1b
+export LOG_LEVEL=DEBUG
+```
+
 ### SHAP Explainability Integration
 
 **Purpose:** Provide feature-level explanations for model predictions to support:
@@ -854,7 +922,7 @@ suite.expect_column_to_exist("CharContinuationRate")
 
 **Multi-Stage Build:**
 ```dockerfile
-# gateway.Dockerfile  (BRANCH: feature/docker-slim-gateway)
+# gateway.Dockerfile  
 
 # ---- build stage: install runtime deps into a venv ----
 FROM python:3.11-slim AS builder
@@ -1067,71 +1135,109 @@ pip install -e ".[dev]"
 
 ### Running the Services
 
+**Option 1: Using Docker Compose (Recommended)**
+```bash
+# Start all services
+docker compose up -d
+
+# Check service health
+docker ps
+
+# View logs
+docker compose logs -f gateway
+docker compose logs -f model-svc
+```
+
+**Option 2: Running Locally (Development)**
+
 **Terminal 1: Model Service**
 ```bash
-python -m model_svc.main
+python -m src.model_svc.main
 # Wait for: ✓ Model Service Ready
-# Listening on http://localhost:9000
+# Listening on http://localhost:8002
 ```
 
 **Terminal 2: Gateway Service**
 ```bash
 # Windows
-set MODEL_SVC_URL=http://localhost:9000
+set MODEL_SVC_URL=http://localhost:8002
+
 # Linux/Mac
-export MODEL_SVC_URL=http://localhost:9000
+export MODEL_SVC_URL=http://localhost:8002
 
-python -m gateway.main
+python -m src.gateway.main
 # Listening on http://localhost:8000
 ```
 
 ### Testing the System
 
+**When using Docker (recommended):**
+
 **1. Whitelist Test**
 ```bash
-curl -X POST http://localhost:8000/predict \
-  -H "Content-Type: application/json" \
-  -d '{"url":"https://github.com"}'
+curl -X POST http://localhost:8080/predict ^
+  -H "Content-Type: application/json" ^
+  -d "{\"url\":\"https://github.com\"}"
 # → {"decision":"ALLOW","reason":"domain-whitelist","source":"whitelist"}
 ```
 
 **2. Phishing Detection**
 ```bash
-curl -X POST http://localhost:8000/predict \
-  -H "Content-Type: application/json" \
-  -d '{"url":"https://phishing.top"}'
+curl -X POST http://localhost:8080/predict ^
+  -H "Content-Type: application/json" ^
+  -d "{\"url\":\"https://phishing.top\"}"
 # → {"p_malicious":1.0,"decision":"BLOCK","reason":"policy-band"}
 ```
 
 **3. Short Domain Routing**
 ```bash
-curl -X POST http://localhost:8000/predict \
-  -H "Content-Type: application/json" \
-  -d '{"url":"https://npm.org"}'
+curl -X POST http://localhost:8080/predict ^
+  -H "Content-Type: application/json" ^
+  -d "{\"url\":\"https://npm.org\"}"
 # → {"decision":"ALLOW","reason":"judge-short-domain-lean-legit"}
 ```
 
 **4. SHAP Dashboard**
-Open browser: `http://localhost:8000/explain`
+Open browser: `http://localhost:8080/explain`
 
 **5. Stats Monitoring**
 ```bash
-curl http://localhost:8000/stats
+curl http://localhost:8080/stats
 # → {"policy_decisions":{...},"final_decisions":{...},"judge_verdicts":{...}}
 ```
 
 ### Docker Deployment
 
+#### Using standalone Docker
+
 ```bash
 # Build gateway image
 docker build -f docker/gateway.Dockerfile -t phishguard-gateway:latest .
 
-# Run container
-docker run --rm -p 8000:8000 \
-  -e MODEL_SVC_URL=http://host.docker.internal:9000 \
+# Build model service image
+docker build -f docker/model.Dockerfile -t phishguard-model:latest .
+
+# Run model service first
+docker run -d --name phishguard-model -p 8002:8002 phishguard-model:latest
+
+# Run gateway service
+docker run --rm -p 8080:8000 \
+  -e MODEL_SVC_URL=http://host.docker.internal:8002 \
+  -e THRESHOLDS_JSON=configs/dev/thresholds.json \
+  -e JUDGE_BACKEND=stub \
   phishguard-gateway:latest
 ```
 
+#### Using Docker Compose
+
+```bash
+# Build and run all services
+docker compose up -d
+
+# Or build and run specific service
+docker compose up gateway -d
+```
+
 ---
 
 ## 🗂️ Repository Structure