Side project · in progress — this page is a work log as it comes together.
IaC Sentinel
LLM-assisted security linter
Scans infrastructure-as-code (Terraform, Docker, CI config) for misconfigurations and insecure defaults, then explains each finding — and the fix — in plain language.
- Python
- Static analysis
- LLM
Context
IaC Sentinel is an LLM-assisted security linter for infrastructure-as-code — it scans Terraform, Docker, and CI configuration for misconfigurations and insecure defaults, then explains each finding in plain language.
The problem
Infrastructure misconfigurations are a leading cause of breaches, and existing linters tend to be noisy and cryptic — they flag a rule ID without explaining the risk or the fix, so findings get ignored.
Approach
Static analysis over IaC files (Terraform, Docker, CI config) surfaces misconfigurations and insecure defaults; an LLM layer then explains each finding — the risk and the remediation — in language a developer can act on immediately. Python core.
Securing it
A security tool has to be trustworthy itself: it reads potentially sensitive config, so the intended design is local, self-contained analysis where code never leaves the user's environment and scanned config isn't retained.
What shipped
Planned — on the roadmap, building on the same security-first approach as the rest of this work.
Links