health

Runs project quality tools and computes a weighted 0-10 health score with trend tracking.

Updated Jul 3, 2026
One-click install
npx skills add https://github.com/alex-jordan547/agent-setup --skill health-alex-jordan547
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: health
Source: https://github.com/alex-jordan547/agent-setup/tree/main/archive/health
Command: npx skills add https://github.com/alex-jordan547/agent-setup --skill health-alex-jordan547

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? It replaces the manual work of running type checkers, linters, test suites, and dead-code detectors one by one by executing them all at once and presenting a single scored dashboard with historical trends. ## Core Features & Use Cases - Automatic tool detection: Detects the project's own type checker, linter, test runner, dead-code analyzer, and shellcheck from config files, or reads a ## Health Stack section from AGENTS.md. - Weighted composite scoring: Scores each category 0-10 and combines them with fixed weights (test 32%, typecheck 26%, lint 21%, deadcode 14%, shell 7%) into one composite score. - Trend tracking and HTML dashboard: Appends each run to a JSONL history file and renders a self-contained HTML dashboard with an SVG trend line and per-category deltas. - Use Case: Ask for a health check before a release to get a composite score, see whether quality improved since the last run, and receive the three highest-impact issues to fix first. ## Quick Start Ask the agent to run a health check on this repository and show the quality dashboard with the score trend.

Frequently Asked Questions about health

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I run all code quality checks at once in a project?

Request a health check and the skill auto-detects your type checker, linter, test runner, dead-code tool, and shellcheck from config files, then runs each sequentially. It captures exit codes, durations, and output, then scores every category 0-10.

How is the composite code health score calculated?

Each category is scored 0-10 using fixed thresholds, then combined with weights: test 32%, typecheck 26%, lint 21%, deadcode 14%, and shell 7%. Weights are renormalized over the categories that actually ran.

Does the health check fix the issues it finds?

No, the skill is report-only by design and never modifies code. It presents a dashboard with the three highest-impact recommendations ranked by score impact per effort, and the user decides what to act on.

Can I pin which tools the health check uses?

Yes, add a `## Health Stack` section to AGENTS.md or CLAUDE.md listing your tools, and the skill uses those instead of auto-detection. After detecting tools it also offers to persist that section for you.

How do I track code quality trends over time?

Every run appends a JSON line to ~/.cache/agent-health/<repo>.jsonl with the score and per-category details. Running the bundled render-dashboard.mjs script generates a self-contained HTML page with an SVG trend line and per-category deltas.

What happens if a tool like shellcheck is not installed?

Tools outside the npm ecosystem that are unavailable are reported as SKIPPED with the reason, never as failures. npm-ecosystem tools are run via npx -y even when not installed locally, so they are never skipped.