chore: update vendored skills and record provenance

humanizer 2.2.0 -> 2.9.1 (blader/humanizer, MIT): adds a Voice Calibration section and a passive-voice pattern, reworks negative parallelisms and em dashes. Version moved to metadata.version upstream. impeccable 3.6.0 -> 4.0.4 (pbakaus/impeccable, Apache-2.0): the repo tags the skill and the npm CLI separately, so 3.6.0 was a real release and npm's 3.5.0 was never the comparison. Adds native-platform reference briefs. Each now carries an UPSTREAM file; bin/check-vendored.sh reports drift.
2026-08-01 14:25:26 +00:00
parent 2f7f512cec
commit 9023ed3ef9
144 changed files with 28302 additions and 5511 deletions
@@ -68,6 +68,15 @@ const GENERIC_FONTS = new Set([
 const WCAG_LARGE_TEXT_PX = 18 * (96 / 72);
 const WCAG_LARGE_BOLD_TEXT_PX = 14 * (96 / 72);

+// Em-dash overuse (advisory) thresholds, shared by the regex/static-HTML
+// analyzer and the browser DOM check so both fire on the same saturation
+// pattern. Two gates must hold: an absolute floor of EM_DASH_FLOOR dashes, and
+// a density of at least one dash per EM_DASH_CHARS_PER_DASH characters of body
+// text. A long article that uses a few em-dashes is left alone; a short,
+// dash-per-clause page is not.
+const EM_DASH_FLOOR = 8;
+const EM_DASH_CHARS_PER_DASH = 500;
+
 // Serif faces that show up in italic-display heroes. The rule also fires when
 // the primary face is unknown but the stack ends in the generic `serif` token,
 // which catches custom/private faces with a serif fallback.
@@ -97,5 +106,7 @@ export {
  GENERIC_FONTS,
  WCAG_LARGE_TEXT_PX,
  WCAG_LARGE_BOLD_TEXT_PX,
+  EM_DASH_FLOOR,
+  EM_DASH_CHARS_PER_DASH,
  KNOWN_SERIF_FONTS,
 };
@@ -0,0 +1,30 @@
+const GOOGLE_FONTS_URL_RE = /fonts\.googleapis\.com\/css2?\?[^"'\s)<>]*/gi;
+
+function normalizeGoogleFontFamilyParam(value) {
+  return String(value || '')
+    .split('|')
+    .map(part => part.split(':')[0].trim().toLowerCase())
+    .filter(Boolean);
+}
+
+function extractGoogleFontFamilies(text) {
+  const families = [];
+  if (!text) return families;
+
+  GOOGLE_FONTS_URL_RE.lastIndex = 0;
+  let urlMatch;
+  while ((urlMatch = GOOGLE_FONTS_URL_RE.exec(text)) !== null) {
+    const url = urlMatch[0];
+    const queryStart = url.indexOf('?');
+    if (queryStart === -1) continue;
+
+    const params = new URLSearchParams(url.slice(queryStart + 1).replace(/&amp;/g, '&'));
+    for (const value of params.getAll('family')) {
+      families.push(...normalizeGoogleFontFamilyParam(value));
+    }
+  }
+
+  return families;
+}
+
+export { extractGoogleFontFamilies };
@@ -0,0 +1,148 @@
+/**
+ * Inline, in-file ignore directives — eslint-disable-style waivers that live at
+ * the point they apply and travel with the artifact instead of (or alongside)
+ * an ignore in `.impeccable/config.json`.
+ *
+ * A config ignore is the right default for repo-wide policy. This complements it
+ * for the one case config can't cover: a waiver that belongs to a single file and
+ * needs to follow that file when it leaves the repo — a generated/exported
+ * standalone document, an emailed HTML file, a snippet scanned out of context.
+ *
+ * Comment-syntax-agnostic: the directive is a raw token matched anywhere on a
+ * line, so the same marker works across every comment style impeccable scans —
+ * `//`, `/* *\/`, `<!-- -->`, `#`, `{/* *\/}`, `{# #}`. Trailing comment closers
+ * are stripped before the rule list is parsed.
+ *
+ * Syntax (reason optional; eslint `--` or biome `:` separator):
+ *
+ *   impeccable-disable <rule>[, <rule>...] [-- reason]   whole file
+ *   impeccable-disable-line <rule>...      [-- reason]   the same line
+ *   impeccable-disable-next-line <rule>... [-- reason]   the following line
+ *   impeccable-disable                                   bare / `*` = every rule
+ *
+ * Examples:
+ *
+ *   <!-- impeccable-disable overused-font -- exported brand doc, font is first-party -->
+ *   .brand { font-family: Inter; } /* impeccable-disable-line overused-font *\/
+ *   // impeccable-disable-next-line bounce-easing: intentional playful affordance
+ *
+ * Behavior is suppression, for parity with config ignores: a matched directive
+ * drops the finding. The inline reason is self-documenting in the diff; it is not
+ * required and is discarded at scan time (only used here to keep reason words out
+ * of the parsed rule list).
+ */
+
+const DIRECTIVE_RE = /impeccable-(disable-next-line|disable-line|disable)\b[ \t]*([^\n\r]*)/gi;
+
+// Trailing comment closers, so `*/`, `*/}`, `-->`, `*}`, `#}`, `%>`, `}}` don't
+// leak into the rule list. Anchored to end-of-line; the leading `\s*` mops up the
+// space before the closer. `--+>` covers `-->` and any longer dash run.
+const TRAILING_CLOSER_RE = /\s*(?:\*\/\}?|--+>|\*\}|#\}|%>|\}\})\s*$/;
+
+function normalizeRule(token) {
+  return String(token || '').trim().toLowerCase();
+}
+
+// Split the directive remainder into rule tokens, dropping any human reason that
+// follows an eslint-style `--` or biome-style `:` separator. Rule ids only ever
+// contain single hyphens (`overused-font`, `bounce-easing`), so `--` and `:`
+// are unambiguous separators.
+function parseRuleList(remainder) {
+  let text = String(remainder || '').replace(TRAILING_CLOSER_RE, '').trim();
+  // Cut off a human reason at the first `--` (eslint) or `:` (biome) separator.
+  const reasonSep = text.match(/\s*(?:--+|:)\s*/);
+  if (reasonSep) text = text.slice(0, reasonSep.index);
+  const tokens = text.split(/[\s,]+/).map(normalizeRule).filter(Boolean);
+  if (tokens.length === 0 || tokens.includes('*')) return ['*'];
+  return tokens;
+}
+
+function addRules(set, rules) {
+  for (const rule of rules) set.add(rule);
+}
+
+function getSet(map, key) {
+  let set = map.get(key);
+  if (!set) {
+    set = new Set();
+    map.set(key, set);
+  }
+  return set;
+}
+
+/**
+ * Parse every inline ignore directive in a file's raw text.
+ *
+ * Returns sets keyed by the 1-based line the directive *targets* so matching is a
+ * direct lookup:
+ *   - file:     rules disabled for the whole file
+ *   - line:     line -> rules disabled on that exact line (disable-line)
+ *   - nextLine: line -> rules disabled on that line (disable-next-line on line-1)
+ *
+ * `*` in any set means "every rule".
+ */
+function parseInlineIgnores(content) {
+  const result = { file: new Set(), line: new Map(), nextLine: new Map() };
+  const text = typeof content === 'string' ? content : '';
+  // Cheap bail-out: the substring must be present for any directive to exist.
+  // Case-insensitive to match DIRECTIVE_RE's `i` flag (e.g. `Impeccable-Disable`).
+  if (!/impeccable-disable/i.test(text)) return result;
+
+  // Split on `\n` only, exactly as detectText numbers lines, so directive line
+  // keys line up with finding `line` values (incl. on `\r`-only line endings).
+  // The directive regex excludes `\r`, so a trailing `\r` on `\r\n` files is
+  // never captured into the rule list.
+  const lines = text.split('\n');
+  for (let i = 0; i < lines.length; i++) {
+    DIRECTIVE_RE.lastIndex = 0;
+    let m;
+    while ((m = DIRECTIVE_RE.exec(lines[i])) !== null) {
+      const variant = m[1].toLowerCase();
+      const rules = parseRuleList(m[2]);
+      if (variant === 'disable') {
+        addRules(result.file, rules);
+      } else if (variant === 'disable-line') {
+        addRules(getSet(result.line, i + 1), rules);
+      } else {
+        // disable-next-line on line i+1 targets line i+2.
+        addRules(getSet(result.nextLine, i + 2), rules);
+      }
+    }
+  }
+  return result;
+}
+
+function setMatches(set, rule) {
+  return Boolean(set) && (set.has('*') || set.has(rule));
+}
+
+function isInlineIgnored(finding, directives) {
+  const rule = normalizeRule(finding && finding.antipattern);
+  if (!rule) return false;
+  if (setMatches(directives.file, rule)) return true;
+  const line = Number(finding && finding.line) || 0;
+  if (line > 0) {
+    if (setMatches(directives.line.get(line), rule)) return true;
+    if (setMatches(directives.nextLine.get(line), rule)) return true;
+  }
+  return false;
+}
+
+function hasDirectives(directives) {
+  return directives.file.size > 0 || directives.line.size > 0 || directives.nextLine.size > 0;
+}
+
+/**
+ * Drop findings waived by an inline directive in the same file's source text.
+ * Findings without a usable line number (e.g. static-HTML page-level findings)
+ * are only matched by whole-file directives — which is the standalone-document
+ * case this primitive exists for.
+ */
+function applyInlineIgnores(findings, content) {
+  if (!Array.isArray(findings) || findings.length === 0) return findings;
+  const directives = parseInlineIgnores(content);
+  if (!hasDirectives(directives)) return findings;
+  return findings.filter((finding) => !isInlineIgnored(finding, directives));
+}
+
+export { parseInlineIgnores, applyInlineIgnores, isInlineIgnored };