refactor: remove implicit content sniffing and table conversion

- Split loadCSV into parseCSVString (content) and loadCSVFromPath
  (path); drop isCSV/looksLikeCsv heuristics
- Delete coerceSparkTables and markedTable label-header magic;
  plain markdown tables now render as plain tables
- Add explicit markdown role=spark-table fence syntax that converts
  pipe tables to CSV at scan time, with dice-header validation
- Map ESM-only github-slugger and csv-parse browser build to CJS in
  jest config; add content-registry tests

__mocks__/github-slugger.js

@@ -0,0 +1,26 @@
+/**
+ * CJS mock of `github-slugger` (v2 is ESM-only, which jest's CJS runtime
+ * cannot require). Auto-applied to all test files via the root `__mocks__`
+ * directory — no `jest.mock()` call needed.
+ */
+class Slugger {
+  constructor() {
+    this.seen = new Map();
+  }
+
+  slug(value, maintainCase) {
+    let slug = String(value).trim();
+    if (!maintainCase) slug = slug.toLowerCase();
+    slug = slug
+      .replace(/[^\p{L}\p{N}\s_-]/gu, "")
+      .replace(/\s/g, "-");
+
+    // github-slugger dedups repeated slugs with a -1, -2, ... suffix
+    const count = this.seen.get(slug) || 0;
+    this.seen.set(slug, count + 1);
+    if (count > 0) return `${slug}-${count}`;
+    return slug;
+  }
+}
+
+module.exports = Slugger;