Development drift analysis of aws/aws-cdk — 7 drift signals across CI, deployment, git, and dependencies

[alpsla]

Hi CDK team,

I built Evolution Engine, an open source tool that detects development process drift by analyzing commit patterns, CI builds, deployments, and dependency signals over time.

I ran it on this repo (12,600+ commits, 2018-2026) and found some interesting patterns:

• 7 significant drift signals across CI, deployment, version control, and dependencies
• 11 correlation patterns — CI build time linked to file count and code spread, deployment cadence tied to change patterns
• CI build time spike to 70+ minutes (normally ~26 seconds) — a 111x deviation
• Release cadence gap of 16.5 days where releases normally ship every 21 hours
• Dependency count trending down (2,119 → 1,802) correlating with deployment cadence shifts

The tool also generates an investigation prompt you can paste into any AI assistant (Claude, Cursor, Copilot) to get root cause analysis — so your AI tools can help diagnose the patterns they may have created.

Full interactive report: https://codequal.dev/reports/aws-cdk

The tool runs locally (no code leaves your machine) and is open source: https://github.com/alpsla/evolution-engine

Would love to hear if the CDK team finds this kind of analysis useful. Happy to answer any questions about the methodology.

— Rosti, Evolution Engine