What problem does it solve? Code quality is spread across many tools — type checkers, linters, test runners, dead code detectors — and no single number tells you if the codebase is getting better or worse. This Skill runs every available check, scores each category, and produces one composite dashboard with trend tracking over time. ## Core Features & Use Cases - Tool auto-detection: Detects tsc, biome, eslint, ruff, pytest, cargo test, go test, knip, and shellcheck from project config files, or reads a pinned ## Health Stack section from CLAUDE.md. - Weighted composite scoring: Scores type check (25%), lint (20%), tests (30%), dead code (15%), and shell lint (10%) on a 0-10 scale and combines them into one composite score. - Trend tracking and regression analysis: Appends each run to a JSONL history file, compares against the last 10 runs, and highlights which categories declined with the exact errors responsible. - Use Case: After a week of feature work, run the health check to see the composite score dropped from 9.4 to 8.2 because 12 new lint warnings and 2 failing tests were introduced, then get a prioritized fix list ranked by category weight. ## Quick Start Ask the assistant to run a health check on this codebase and show the quality dashboard with trend data.
