Open-source AI coding agent skill that finds and fixes infra security misconfigs

I built an open-source skill for AI coding agents (Claude Code, Cursor, Windsurf, etc.) that scans your infrastructure configs for security issues, explains them in plain language, and offers to fix them.

One-liner install:

  curl -fsSL https://raw.githubusercontent.com/coguardio/misconfiguration-detection-skill/master/install.sh | bash

Then type /misconfiguration-detection in your agent.

It wraps CoGuard (https://coguard.io) and scans Terraform, Kubernetes, Helm, Docker, CloudFormation, cloud configs (AWS/Azure/GCP), databases, web servers, CI/CD pipelines, and more. Results come back with severity ratings, file paths, line numbers, and specific fixes.

The interesting angle: developers are generating more infrastructure code with AI agents than ever, but there's no security review step built into that workflow. This plugs that gap; your agent writes the Terraform, then immediately checks whether it's actually secure.

It also supports compliance frameworks (SOC2, HIPAA, STIG) if you're in audit season.

Repo: https://github.com/coguardio/misconfiguration-detection-skil... Video demo: https://www.youtube.com/watch?v=851QsRDuoS4 Free to use. MIT licensed. Happy to answer questions about the approach.