Redesign an existing website to make it better. Built on top of impeccable.
88%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Low
Low-risk findings worth noting
#!/usr/bin/env node
/**
* skills/replica/scripts/css-rules.mjs — the rule blocks of a stylesheet whose
* selector matches a regex, each with the @media / @supports condition it
* sits in, instead of line ranges of the whole sheet.
*
* Why: lifting values (recreation-procedure § lifted, not eyeballed) means
* finding the live rule for a component. In one recorded hands-off session
* (2026-09-18) that was done as `sed -n 4370,4773p site.pretty.css` and
* ad-hoc `node -e` scans — 102k + 141k characters of stylesheet in the
* agent's context for perhaps thirty rules that mattered. This prints the
* thirty.
*
* Usage:
* node css-rules.mjs <file.css> "<selector regex>" [--media <regex> | --no-media]
* [--decl <regex>] [--max <n>=40] [--count]
*
* Output, one rule per entry:
* L2231 .cmp-teaser__title, .cmp-teaser__pretitle @media (min-width: 1024px)
* font-size: 32px; line-height: 40px; margin: 0 0 8px;
* --media filters on the condition text (`--no-media`: top-level rules only);
* --decl keeps rules with a matching declaration and shows only those
* declarations (+N more); --count prints the number of matches. Values longer
* than 80 characters (data URIs) are cut. Minified sheets get byte offsets
* (@1234) instead of line numbers. Exit 0 = printed, 2 = no match, 125 = usage
* (bad regex, unknown flag, or a value flag — --media, --decl, --max — followed
* by nothing or by another --flag: the flag is named, never swallowed).
*
* Also importable: parseCss(text) → [{ selector, body, media[], offset }].
*/
/* eslint-disable no-restricted-syntax, brace-style, object-curly-newline, max-len */
import { readFileSync, realpathSync } from 'node:fs';
import { fileURLToPath } from 'node:url';
const HELP = 'Usage: node css-rules.mjs <file.css> "<selector regex>" [--media <regex> | --no-media] [--decl <regex>] [--max <n>] [--count]';
const CONDITIONAL_AT_RULES = new Set(['media', 'supports', 'layer', 'container', 'document', 'scope']);
export function parseCss(text) {
const rules = [];
const n = text.length;
let i = 0;
const skipComment = () => { if (text.startsWith('/*', i)) { const e = text.indexOf('*/', i + 2); i = e < 0 ? n : e + 2; return true; } return false; };
const skipWs = () => { for (;;) { while (i < n && /\s/.test(text[i])) i += 1; if (!skipComment()) return; } };
const skipString = () => { const q = text[i]; i += 1; while (i < n && text[i] !== q) { if (text[i] === '\\') i += 1; i += 1; } i += 1; };
// Read up to a stop character at paren depth 0 (outside strings/comments); returns the text read.
const readUntil = (stops, braces = false) => {
const start = i; let paren = 0; let brace = 0;
while (i < n) {
const c = text[i];
if (c === '"' || c === "'") { skipString(); continue; }
if (text.startsWith('/*', i)) { skipComment(); continue; }
if (c === '(') paren += 1; else if (c === ')') paren = Math.max(0, paren - 1);
else if (braces && c === '{') brace += 1;
else if (braces && c === '}' && brace > 0) brace -= 1;
else if (paren === 0 && brace === 0 && stops.includes(c)) return text.slice(start, i);
i += 1;
}
return text.slice(start);
};
const skipBlock = () => { // i is just after '{'; returns the raw inner text
const start = i; let depth = 1;
while (i < n && depth) { if (text.startsWith('/*', i)) { skipComment(); continue; } const c = text[i]; if (c === '"' || c === "'") { skipString(); continue; } if (c === '{') depth += 1; else if (c === '}') depth -= 1; i += 1; }
return text.slice(start, Math.max(start, i - 1));
};
const parseList = (media) => {
for (;;) {
skipWs();
if (i >= n) return;
if (text[i] === '}') { i += 1; return; }
const start = i;
const prelude = readUntil('{;}').trim();
if (i >= n) return;
if (text[i] === ';') { i += 1; if (prelude) rules.push({ selector: prelude, body: '', media, offset: start }); continue; }
if (text[i] === '}') { i += 1; return; }
i += 1; // '{'
if (prelude.startsWith('@')) {
const name = prelude.slice(1).split(/[\s(]/)[0].toLowerCase();
if (CONDITIONAL_AT_RULES.has(name)) { parseList([...media, prelude]); continue; }
rules.push({ selector: prelude, body: skipBlock().replace(/\s+/g, ' ').trim(), media, offset: start });
continue;
}
const body = readUntil('}', true);
i += 1;
rules.push({ selector: prelude.replace(/\s+/g, ' '), body: body.trim(), media, offset: start });
}
};
parseList([]);
return rules;
}
export function splitDeclarations(body) {
const out = []; let cur = ''; let paren = 0; let q = null;
for (const c of body) {
if (q) { cur += c; if (c === q) q = null; continue; }
if (c === '"' || c === "'") { q = c; cur += c; continue; }
if (c === '(') paren += 1; else if (c === ')') paren = Math.max(0, paren - 1);
if (c === ';' && paren === 0) { if (cur.trim()) out.push(cur.trim()); cur = ''; continue; }
cur += c;
}
if (cur.trim()) out.push(cur.trim());
return out.map((d) => d.replace(/\s+/g, ' ').replace(/^([^:]+):\s*/, '$1: '));
}
const cut = (s, max = 80) => (s.length > max ? `${s.slice(0, max - 1)}…` : s);
export function formatRule(rule, lineOf, declRe) {
const where = lineOf(rule.offset);
const media = rule.media.length ? ` ${rule.media.join(' ∧ ')}` : '';
let decls = splitDeclarations(rule.body);
let hidden = 0;
if (declRe) { const kept = decls.filter((d) => declRe.test(d)); hidden = decls.length - kept.length; decls = kept; }
const body = decls.map((d) => `${cut(d)};`).join(' ') + (hidden ? ` (+${hidden} more)` : '');
return `${where} ${cut(rule.selector, 160)}${media}\n ${body || '(no declarations)'}`;
}
export function lineLocator(text) {
const lineCount = text.split('\n').length;
const minified = text.length / lineCount > 400; // a pretty-printed sheet averages well under 100 chars a line
if (minified) return (offset) => `@${offset}`;
const starts = [0];
for (let k = 0; k < text.length; k += 1) if (text[k] === '\n') starts.push(k + 1);
return (offset) => { let lo = 0; let hi = starts.length - 1; while (lo < hi) { const mid = (lo + hi + 1) >> 1; if (starts[mid] <= offset) lo = mid; else hi = mid - 1; } return `L${lo + 1}`; };
}
function cli(argv) {
const rest = argv.slice(2);
if (!rest.length || rest.includes('--help') || rest.includes('-h')) { console.log(HELP); return rest.length ? 0 : 125; }
let file = null; let pattern = null; let mediaRe = null; let noMedia = false; let declRe = null; let max = 40; let count = false;
const re = (v, flag) => { try { return new RegExp(v, 'i'); } catch (e) { throw new Error(`bad ${flag} regex ${v}: ${e.message}`); } };
// A value flag followed by nothing or by another --flag is a usage error naming the flag (125).
const need = (i, flag) => { if (i + 1 >= rest.length || rest[i + 1].startsWith('--')) throw new Error(`${flag} needs a value\n${HELP}`); return rest[i + 1]; };
try {
for (let i = 0; i < rest.length; i += 1) {
const a = rest[i];
if (a === '--media') { mediaRe = re(need(i, a), a); i += 1; }
else if (a === '--no-media') noMedia = true;
else if (a === '--decl') { declRe = re(need(i, a), a); i += 1; }
else if (a === '--max') { max = Number(need(i, a)); i += 1; }
else if (a === '--count') count = true;
else if (a.startsWith('--')) throw new Error(`unknown flag ${a}\n${HELP}`);
else if (!file) file = a;
else if (!pattern) pattern = a;
else throw new Error(`unexpected argument ${a}\n${HELP}`);
}
if (!file || !pattern) throw new Error(`need <file.css> and a selector regex\n${HELP}`);
const selRe = re(pattern, 'selector');
let text; try { text = readFileSync(file, 'utf8'); } catch (e) { console.error(`css-rules: cannot read ${file}: ${e.message}`); return 1; }
const all = parseCss(text);
const matches = all.filter((r) => selRe.test(r.selector)
&& (!noMedia || r.media.length === 0)
&& (!mediaRe || r.media.some((m) => mediaRe.test(m)))
&& (!declRe || splitDeclarations(r.body).some((d) => declRe.test(d))));
if (count) { console.log(`${matches.length} of ${all.length} rules match /${pattern}/i in ${file}`); return matches.length ? 0 : 2; }
if (!matches.length) { console.error(`css-rules: no rule of ${all.length} matches /${pattern}/i${mediaRe ? ` under media /${mediaRe.source}/` : ''}${declRe ? ` with a declaration /${declRe.source}/` : ''} in ${file}`); return 2; }
const lineOf = lineLocator(text);
console.log(`${file}: ${matches.length} of ${all.length} rules match /${pattern}/i${matches.length > max ? `, showing ${max}` : ''}`);
for (const r of matches.slice(0, max)) console.log(formatRule(r, lineOf, declRe));
if (matches.length > max) console.log(`… ${matches.length - max} more — narrow the selector regex, add --media/--decl, or raise --max`);
return 0;
} catch (e) { console.error(`css-rules: ${e.message}`); return 125; }
}
// Main-module guard by REAL path: node resolves the entry's symlinks for import.meta.url but leaves process.argv[1]
// as typed, so a symlinked checkout or temp dir (e.g. /var → /private/var) would otherwise make the CLI a silent no-op.
function isMainModule(metaUrl) {
const argv1 = process.argv[1];
if (!argv1) return false;
try { return realpathSync(argv1) === fileURLToPath(metaUrl); } catch { return false; }
}
if (isMainModule(import.meta.url)) process.exitCode = cli(process.argv);.tessl-plugin
skills
audit
reference
deploy
reference
scripts
test
diff
direct
dynamics
extract
migrate
prepare-migration
prototype
assets
motion
fixtures
reference
qa
replica
reference
scripts
test
reskin
rollout
spec
stardust
uplift
reference