Zero Chaos: Scaling Detection Engineering at the Speed of Software, with Detection As Code

Zero Chaos: Scaling Detection Engineering at the Speed of Software, with Detection As Code

The article explains how Detection as Code brings software engineering discipline to security detections by adding version control, peer review, testing, rollback, and traceability through Terraform. It also shows a Rapid7 Terraform example for an encoded PowerShell detection mapped to T1059.001, along with AI-assisted rule writing and import support for existing UI-built rules. #Rapid7 #Terraform #PowerShell #T1059.001 #IncidentCommand #InsightIDR

Keypoints

  • Detection engineering is often managed manually through UI edits, wikis, and single-user saves, unlike software engineering workflows.
  • Detection as Code applies version control, peer review, automated validation, and rollback to detection rules.
  • Inline test cases help verify that a detection fires on malicious activity and does not trigger on benign activity.
  • Terraform plan and Terraform state provide a preview and authoritative record of the detection environment.
  • The article provides a Rapid7 Terraform provider example for detecting encoded PowerShell execution with high priority alerting.
  • The example detection is mapped to MITRE ATT&CK technique T1059.001 and includes positive and negative test payloads.
  • The workflow supports importing existing UI rules and using AI tools like Claude Code, Cursor, VS Code Copilot, and Kiro to draft detections faster.

Read more: https://www.rapid7.com/blog/post/dr-scaling-engineering-detection-as-code