What I shipped this week on khaga.dev — Fix It buttons, health scores, and 24/7 alerts
Two weeks ago I launched khaga.dev — a free AI tool that diagnoses AWS, GCP, Azure, and Kubernetes infrastructure in seconds using Claude AI.
This week I shipped 5 major features based on early feedback. Here's what's new.
⚡ Fix It button
The most requested feature. Every finding now has a Fix It button.
Click it and you get:
- The exact shell command to fix the issue
- A context-aware safety checklist (different for kubectl vs aws vs terraform)
- One-click copy to clipboard
- "Copy & Open Remediate" to run it directly in Khaga
Example for a CRITICAL finding on a Dockerfile:
Fix: sed -i 's/FROM ubuntu:latest/FROM ubuntu:22.04/' Dockerfile
No more Googling at 2am.
📊 Infrastructure Health Score
The dashboard now shows a 0-100 health score per cloud provider on login.
- 🟢 70-100 = Healthy
- 🟡 40-69 = Degraded
- 🔴 0-39 = Critical
Scored based on the severity and frequency of findings across your last 500 diagnoses. Click any score card to jump straight to that provider's diagnosis.
🌍 Multi-region AWS scan
Previously Khaga only scanned one AWS region. Most people have resources spread across multiple regions and were missing issues.
Now there's a "Scan all regions" checkbox on the AWS panel. Check it and Khaga scans all 6 major regions in parallel:
us-east-1, us-west-2, eu-west-1, ap-south-1, ap-southeast-1, ap-northeast-1
All findings get tagged with their region.
🔔 24/7 Automated Alerts
Khaga now runs in the background and alerts you when critical issues are detected — without you having to manually trigger a diagnosis.
Set your scan frequency in Settings:
- Disabled
- Every hour
- Daily
- Weekly
Alerts arrive via Slack (Block Kit formatted) or email (HTML template via Resend) with a "View in Khaga" button.
⚡ Streaming responses
Predictive analysis and Compliance used to show a spinner for 20-30 seconds. Users were abandoning before results loaded.
Now results stream in — the UI shows an animated progress indicator within 1-2 seconds and snaps to the final result when done.
What's next
- GitHub Actions integration (scan on every deploy)
- Slack bot (
/khaga scan aws) - Team accounts
All of this is free at khaga.dev. No credit card, no setup beyond adding your cloud credentials.
If you run AWS/K8s/GCP without a dedicated SRE team — this is built for you.
What feature would make you actually use this daily? Drop it in the comments.
Built with Flask, PostgreSQL, Claude AI (Anthropic), deployed on Railway.
#devops #aws #kubernetes #showdev #webdev
Top comments (0)