CCM — AI-era code cost analyzer

Code-volume and inference-cost accounting for AI-assisted development teams

Problem

Classic cloc, scc and COCOMO models no longer fit AI-driven development: output speed, review effort and model-call costs are disconnected from human-effort estimates. Teams want to know net AI output and per-change cost but end up stitching numbers together by hand.

Solution

A CLI scans the repo for code deltas and change frequency, joins model API usage logs, and produces change-vs-cost-vs-effort reports. It breaks down by project, date and model, and estimates how much code was AI-generated versus human-edited.

Users

Small AI-coding teams using Cursor/Claude Code, tech leads, indie developers

MVP

CLI scans repo and outputs delta stats; imports API usage CSV/JSON; cost report by project and model; AI-generated vs human-edited ratio estimate; exportable HTML dashboard

Revenue

Free open-source CLI plus a $19–49/mo team tier with hosted reports, history comparison and multi-repo aggregation

Why now

As AI coding tools spread, managers increasingly ask about AI ROI, yet existing code counters ignore model costs; a CLI can be built solo without any backend.