Codebase Development Cost Estimator

★ 7.5 · financial-analysis

codebase-cost-estimator is a Claude Code skill that estimates the full development cost of an existing codebase using measured lines of code, architectural complexity, and team-composition overhead — not ticket volume or story points. It runs `tokei` or `cloc` to count LOC by language, converts them to developer-hours using productivity bands from `references/rates.md` (e.g. 30–50 LOC/hr for CRUD/UI, 10–20 LOC/hr for GPU/shader code), applies overhead multipliers of 1.9–2.25×, and translates hours into calendar time via organizational-efficiency factors. The output follows a structured template covering codebase metrics, market rates researched for the current year, engineering cost as a low/median/high range, and a full-team breakdown across roles (PM, QA, DevOps, design). An optional AI/Claude ROI step uses git-commit clustering to quantify speed and cost gains from AI-assisted development — useful for defending build-cost, rebuild-cost, or engineering-ROI estimates to clients and stakeholders.