Echeo

I'm getting into this because human-in-the-loop and agentic labor have to become symbiotic with their tooling. I want businesses to know the true value of their labor.

the thesis

Writing the code was never the expensive part.

42%

of new code is AI-generated, and climbing. The code gets written faster than ever. The bill for it doesn't arrive at the pull request. It arrives later, in maintenance, when someone has to understand, change, and keep alive the thing the model produced in a second. Most teams can't see that bill until it's overdue.

Maintenance is where the real bill arrives.echeo.ai

what it would measure

Make code health something you can actually read.

Echeo's whole idea is to turn a vague worry into three concrete readings, so an engineering lead can point at a module and say why it's risky, not just feel that it is:

• The Slop DetectorCognitive load per module. How much a human has to hold in their head to safely touch this code.
• The Churn DashboardWhat keeps getting rewritten. Code that never settles is code that's still costing you.
• Repo Survivability ScoreHow hard is this repo to change? The honest version of "predict roadmap risk": flag what will fight you before you promise a date on it.