🤖 SRE AI Agent

This project is a simple SRE AI Agent designed for beginners to understand Site Reliability Engineering (SRE), incident detection, and AIOps fundamentals.

The agent runs as a single Python file and simulates how SRE teams detect incidents, assess severity, and suggest remediation actions.

---

🎯 Project Goal

To demonstrate how an SRE-style system:

• Collects service metrics
• Detects incidents using simple rules
• Explains what went wrong
• Suggests safe remediation steps

This project focuses on thinking like an SRE, not complex tooling.

---

🧠 What This Agent Does

• Simulates metrics such as latency, error rate, CPU, and traffic
• Detects incidents based on SRE thresholds
• Classifies severity
• Provides remediation suggestions
• Runs in recommendation-only mode (no auto-fix)

---

🏗️ Architecture (Simplified)

Metrics → Incident Detection → Decision Logic → Human-Readable Output

---

⚙️ Prerequisites

• Python 3.8 or higher
• Basic understanding of command line
• Git (optional)

Check Python version:

python3 --version

---

🚀 How to Run

Clone the repository:

git clone git@github.com:shakirbhattt/sre-ai-agent.git

cd sre-ai-agent

---

🏃 Run the SRE AI Agent:

python main.py

---

Example Output

📊 Metrics Collected:

• latency_p95: 1.8s
• error_rate: 12%
• cpu: 40%
• traffic_rpm: 1200

🚨 INCIDENT DETECTED Severity: HIGH Reason: High error rate and latency breach

Suggested Actions:

• Rollback last deployment
• Check downstream dependencies
• Scale service temporarily