This Repository is meant to serve as a demonstration. This code should not be used in production as is.
A comprehensive AWS infrastructure monitoring and analysis tool that uses CloudWatch MCP tools to discover resources, analyze performance, and generate actionable insights.
Run both agents with a single command:
python demo.pyOr run them individually:
python agent.py- Quick text-based analysis
- Writes findings to
operational.md
python natural_language_observability_agent.py --region us-east-2- Comprehensive HTML dashboard with interactive Plotly charts
- Real CloudWatch metrics visualization
- Working AWS Console deep links
- Industry postmortem analysis
pip install -r requirements.txt- Real CloudWatch Data: Actual metrics from EC2, RDS, Lambda, EKS, etc.
- Interactive Charts: Plotly visualizations of CPU utilization, service distribution
- Working Deep Links: Direct links to your AWS Console resources
- Industry Context: Postmortem analysis for similar incidents
- Beautiful Reports: Interactive HTML dashboards
operational.md- Text analysis from basic agentreports/*.html- Interactive dashboard from advanced agent