28 eval-informed mental models and critical-thinking skills for Claude Code, GitHub Copilot, Codex, Cursor, and other Agent Skills-compatible tools
JavaScript agent-skillsai-toolsanthropicclaudeclaude-codecritical-thinkingdecision-makingfirst-principlesmental-modelsproblem-solvingsystems-thinking
#!/usr/bin/env node
'use strict';
/**
* Tier 0 — Structural lint (free, deterministic, no API cost).
*
* Uses the same pure checks exported by scripts/validate-skills.js
* (required sections, description length, word budgets, catalog count).
* Also reports a legacy "loose" substance score for comparison only.
*/
const path = require('path');
const {
validateAllSkills,
validateSkillContent,
DEFAULT_REQUIRED_SECTIONS,
} = require('../scripts/validate-skills');
const { loadAllSkills } = require('./lib/skills');
const { runDir, writeJson } = require('./lib/io');
function looseScore(content) {
const checks = {
has_frontmatter: /^---\n[\s\S]*?name:[\s\S]*?description:[\s\S]*?---/m.test(content),
has_when_to_use: /^##\s+When to Use\s*$/mi.test(content),
has_when_not_to_use: /^##\s+When NOT to Use\s*$/mi.test(content),
has_procedure: /^##\s+Procedure\s*$/mi.test(content),
has_output: /^##\s+Output\s*$/mi.test(content),
has_verification: /^##\s+Verification\s*$/mi.test(content),
desc_ok: (() => {
const m = content.match(/^---\n[\s\S]*?\ndescription:\s*(.+)\n[\s\S]*?\n---/m) || content.match(/description:\s*(.+)/);
const len = (m && m[1] ? m[1].trim().length : 0);
return len > 0 && len <= 200;
})(),
word_ok: (() => {
const w = content.trim().split(/\s+/).filter(Boolean).length;
return w > 0 && w <= 3000;
})(),
};
const vals = Object.values(checks);
const score = vals.filter(Boolean).length;
return { score, max: vals.length, pct: Math.round((score / vals.length) * 100), checks };
}
function main() {
const report = validateAllSkills();
const skills = loadAllSkills();
const byName = new Map(report.results.map(r => [r.name, r]));
const rows = skills.map(s => {
const structural = byName.get(s.name) || validateSkillContent(s.content, {
name: s.name,
requiredSections: report.required_sections || DEFAULT_REQUIRED_SECTIONS,
});
const loose = looseScore(s.content);
return {
name: s.name,
descLen: structural.description_length ?? s.description.length,
words: structural.words,
max_words: structural.max_words,
structural_pass: structural.pass,
structural_score: structural.score,
structural_max: structural.maxScore,
structural_pct: Math.round((structural.score / Math.max(1, structural.maxScore)) * 100),
missing: (structural.failed || []).map(f => f.name),
loose_pct: loose.pct,
gap: loose.pct - Math.round((structural.score / Math.max(1, structural.maxScore)) * 100),
};
});
rows.sort((a, b) => Number(a.structural_pass) - Number(b.structural_pass) || a.name.localeCompare(b.name));
const overallStructural = rows.length
? Math.round(rows.reduce((a, r) => a + r.structural_pct, 0) / rows.length)
: 0;
const overallLoose = rows.length
? Math.round(rows.reduce((a, r) => a + r.loose_pct, 0) / rows.length)
: 0;
const out = {
tier: 0,
validator: 'scripts/validate-skills.js',
catalog: {
expected_count: report.expected_count,
found_count: report.found_count,
catalog_errors: report.catalog_errors,
ok: report.ok,
},
required_sections: report.required_sections,
overall_structural_pct: overallStructural,
overall_loose_pct: overallLoose,
failed_skills: rows.filter(r => !r.structural_pass).map(r => ({
name: r.name,
missing: r.missing,
words: r.words,
max_words: r.max_words,
descLen: r.descLen,
})),
over_length_descriptions: rows.filter(r => r.descLen > 200).map(r => ({ name: r.name, len: r.descLen })),
over_budget: rows.filter(r => r.max_words != null && r.words > r.max_words).map(r => ({
name: r.name,
words: r.words,
max_words: r.max_words,
})),
rows,
};
// writeJson only under EVAL_RUN results dir (existing runner behavior / untracked results)
const file = path.join(runDir(), 'tier0-structural.json');
writeJson(file, out);
console.log('Tier 0 — structural lint (shared validate-skills checks)');
console.log(` catalog: ${out.catalog.found_count} found / expected ${out.catalog.expected_count} / ok=${out.catalog.ok}`);
console.log(` structural overall: ${overallStructural}%`);
console.log(` loosened overall: ${overallLoose}%`);
console.log(` failed skills: ${out.failed_skills.length}`);
for (const r of out.failed_skills.slice(0, 12)) {
console.log(` ${r.name}: ${r.missing.join(', ')}`);
}
console.log(` over-length descriptions (>200): ${out.over_length_descriptions.map(d => `${d.name}(${d.len})`).join(', ') || 'none'}`);
console.log(` over word budget: ${out.over_budget.length}`);
console.log(` -> ${file}`);
if (!report.ok) process.exitCode = 1;
}
if (require.main === module) {
main();
}
module.exports = { looseScore, main, validateAllSkills };