Add dedicated pages for every library#22

@@ -7,7 +7,8 @@
   "scripts": {
     "dev": "next dev",
     "build": "next build",
-    "start": "next start"
+    "start": "next start",
+    "test": "node --test \"tests/*.test.mjs\" \"lib/*.test.mjs\""

@@ -0,0 +1,122 @@
+import assert from "node:assert/strict";
+import { existsSync, readFileSync, readdirSync, statSync } from "node:fs";
+import { test } from "node:test";
+import ts from "typescript";
+
+const transpileUrl = (source) =>
+  `data:text/javascript,${encodeURIComponent(ts.transpileModule(source, { compilerOptions: { module: ts.ModuleKind.ESNext } }).outputText)}`;
+
+const registrySource = readFileSync(new URL("../data/libraries.ts", import.meta.url), "utf8");
+const { libraries } = await import(transpileUrl(registrySource));
+
+const readBuilt = (relativePath) => {
+  const file = new URL(`../.next/server/app/${relativePath}`, import.meta.url);
+  return existsSync(file) && statSync(file).isFile() ? readFileSync(file, "utf8") : null;
+};
+
+const loadDetails = async (slug) => {
+  const file = new URL(`../data/library-details/${slug}.ts`, import.meta.url);
+  if (!existsSync(file)) return null;
+  const module = await import(transpileUrl(readFileSync(file, "utf8")));
+  return module.default ?? null;
+};
+
+// Read the real aggregation map from `index.ts` so a slug present on disk but
+// missing from the map (which would 404 at the detail route) fails here.
+// Only the keys are needed, so they are read with the TypeScript AST rather
+// than by executing the module (whose extensionless relative imports are not
+// resolvable outside a bundler).
+const loadIndexKeys = () => {
+  const source = readFileSync(new URL("../data/library-details/index.ts", import.meta.url), "utf8");
+  const file = ts.createSourceFile("index.ts", source, ts.ScriptTarget.Latest, true);
+  const keys = [];
+  for (const statement of file.statements) {
+    if (!ts.isVariableStatement(statement)) continue;
+    for (const declaration of statement.declarationList.declarations) {
+      if (declaration.name.getText(file) !== "libraryDetails") continue;
+      if (!declaration.initializer || !ts.isObjectLiteralExpression(declaration.initializer)) continue;
+      for (const property of declaration.initializer.properties) {
+        if (!ts.isPropertyAssignment(property)) continue;
+        const key = property.name;
+        keys.push(ts.isStringLiteral(key) ? key.text : key.getText(file).replace(/^["']|["']$/g, ""));
+      }
+    }
+  }
+  return keys;
+};
+
+// React escapes text children and attribute values into HTML entities; decode
+// them so assertions can compare against plain source strings.
+const decodeHtml = (html) => html
+  .replaceAll(""", '"')
+  .replaceAll("'", "'")
+  .replaceAll("'", "'")
+  .replaceAll("&lt;", "<")
+  .replaceAll("&gt;", ">")
+  .replaceAll("&amp;", "&");
+
+test("registry slugs are unique", () => {
+  const slugs = libraries.map(({ slug }) => slug);
+  assert.equal(new Set(slugs).size, slugs.length);
+});
+
+test("every library has a built detail page with name, website link, docs link, and agent prompt", async () => {
+  for (const library of libraries) {
+    const builtPage = readBuilt(`libraries/${library.slug}.html`);
+    assert.ok(builtPage, `No built detail page for ${library.slug}`);
+    const html = decodeHtml(builtPage);
+    const details = await loadDetails(library.slug);
+    assert.ok(details, `Missing data/library-details/${library.slug}.ts`);
+
+    assert.ok(html.includes(library.name), `${library.slug}: page is missing the library name`);
+    assert.ok(html.includes(library.url), `${library.slug}: page is missing the website link`);
+    assert.ok(html.includes(details.docsUrl), `${library.slug}: page is missing the docs link`);
+    assert.ok(html.includes("Copy prompt"), `${library.slug}: page is missing the agent prompt block`);
+    assert.ok(html.includes(details.agentPrompt), `${library.slug}: page is missing the agent prompt text`);
+  }
+});
+
+// The discovery suite already asserts sitemap coverage for every slug.
+test("unknown slugs do not get a generated page", () => {
+  for (const extension of ["html", "body"]) {
+    assert.equal(readBuilt(`libraries/definitely-not-a-library.${extension}`), null);
+  }
+});