diff --git a/.github/scripts/validate-dataset.mjs b/.github/scripts/validate-dataset.mjs
new file mode 100644
index 0000000..fe4ad7a
--- /dev/null
+++ b/.github/scripts/validate-dataset.mjs
@@ -0,0 +1,138 @@
+// Validates data/exercises.json: JSON Schema, media file existence,
+// duplicate ids, and sync with the copy of the data embedded in index.html.
+import Ajv2020 from "ajv/dist/2020.js";
+import addFormats from "ajv-formats";
+import { existsSync, readFileSync } from "node:fs";
+import assert from "node:assert/strict";
+
+const errors = [];
+const section = (title, fn) => {
+  console.log(`\n— ${title}`);
+  const before = errors.length;
+  fn();
+  if (errors.length === before) console.log("  ok");
+};
+
+function readJson(path) {
+  const text = readFileSync(path, "utf8");
+  try {
+    return JSON.parse(text);
+  } catch (e) {
+    errors.push(`Failed to parse ${path}: ${e.message}`);
+    return null;
+  }
+}
+
+const schema = readJson("data/exercises.schema.json");
+const exercises = readJson("data/exercises.json");
+
+section("JSON Schema validation", () => {
+  if (!schema || !exercises) {
+    errors.push("skipped: schema or dataset failed to parse (see above)");
+    return;
+  }
+  let validate;
+  try {
+    const ajv = new Ajv2020({ allErrors: true, strict: true });
+    addFormats(ajv);
+    validate = ajv.compile(schema);
+  } catch (e) {
+    errors.push(`Failed to compile data/exercises.schema.json: ${e.message}`);
+    return;
+  }
+  if (!validate(exercises)) {
+    for (const err of validate.errors) {
+      errors.push(`schema: ${err.instancePath || "/"} ${err.message}`);
+    }
+  }
+});
+
+section("Duplicate ids", () => {
+  if (!exercises) return errors.push("skipped: data/exercises.json failed to parse (see above)");
+  const seen = new Map();
+  for (const ex of exercises) {
+    seen.set(ex.id, (seen.get(ex.id) ?? 0) + 1);
+  }
+  for (const [id, count] of seen) {
+    if (count > 1) errors.push(`duplicate id "${id}" appears ${count} times`);
+  }
+});
+
+section("Image files exist", () => {
+  if (!exercises) return errors.push("skipped: data/exercises.json failed to parse (see above)");
+  for (const ex of exercises) {
+    if (typeof ex.image !== "string") errors.push(`id ${ex.id}: "image" is missing or not a string`);
+    else if (!existsSync(ex.image)) errors.push(`id ${ex.id}: missing image file "${ex.image}"`);
+  }
+});
+
+section("GIF files exist", () => {
+  if (!exercises) return errors.push("skipped: data/exercises.json failed to parse (see above)");
+  for (const ex of exercises) {
+    if (typeof ex.gif_url !== "string") errors.push(`id ${ex.id}: "gif_url" is missing or not a string`);
+    else if (!existsSync(ex.gif_url)) errors.push(`id ${ex.id}: missing gif file "${ex.gif_url}"`);
+  }
+});
+
+section("index.html data is in sync with data/exercises.json", () => {
+  if (!exercises) return errors.push("skipped: data/exercises.json failed to parse (see above)");
+  const html = readFileSync("index.html", "utf8");
+  const arrayText = extractJsonArray(html, "const EXERCISES = ");
+  if (!arrayText) {
+    errors.push("could not find \"const EXERCISES = [...]\" in index.html");
+    return;
+  }
+  let embedded;
+  try {
+    embedded = JSON.parse(arrayText);
+  } catch (e) {
+    errors.push(`embedded EXERCISES array in index.html is not valid JSON: ${e.message}`);
+    return;
+  }
+  try {
+    assert.deepStrictEqual(embedded, exercises);
+  } catch {
+    errors.push(
+      "index.html's embedded EXERCISES array does not match data/exercises.json — " +
+      "regenerate index.html from the current dataset."
+    );
+  }
+});
+
+if (errors.length) {
+  console.log(`\n${errors.length} problem(s) found:\n`);
+  for (const e of errors) console.log(`  ✗ ${e}`);
+  process.exitCode = 1;
+} else {
+  console.log("\nAll checks passed.");
+}
+
+// Returns the substring of `text` holding the JSON array that starts right
+// after `marker`, found by bracket-depth counting (so it doesn't care
+// whether the array is minified onto one line or pretty-printed).
+function extractJsonArray(text, marker) {
+  const markerIdx = text.indexOf(marker);
+  if (markerIdx === -1) return null;
+  const start = text.indexOf("[", markerIdx);
+  if (start === -1) return null;
+
+  let depth = 0;
+  let inString = false;
+  let escaped = false;
+  for (let i = start; i < text.length; i++) {
+    const ch = text[i];
+    if (inString) {
+      if (escaped) escaped = false;
+      else if (ch === "\\") escaped = true;
+      else if (ch === '"') inString = false;
+      continue;
+    }
+    if (ch === '"') inString = true;
+    else if (ch === "[" || ch === "{") depth++;
+    else if (ch === "]" || ch === "}") {
+      depth--;
+      if (depth === 0) return text.slice(start, i + 1);
+    }
+  }
+  return null;
+}
diff --git a/.github/workflows/validate-dataset.yml b/.github/workflows/validate-dataset.yml
new file mode 100644
index 0000000..1a9274e
--- /dev/null
+++ b/.github/workflows/validate-dataset.yml
@@ -0,0 +1,31 @@
+name: Validate dataset
+
+on:
+  pull_request:
+    paths: &dataset_paths
+      - "data/exercises.json"
+      - "data/exercises.schema.json"
+      - "index.html"
+      - ".github/workflows/validate-dataset.yml"
+      - ".github/scripts/validate-dataset.mjs"
+  push:
+    branches: [main]
+    paths: *dataset_paths
+
+permissions:
+  contents: read
+
+jobs:
+  validate:
+    runs-on: ubuntu-latest
+    timeout-minutes: 5
+    steps:
+      - uses: actions/checkout@v4
+
+      - uses: actions/setup-node@v4
+        with:
+          node-version: 22
+
+      - run: npm install --no-save ajv@8 ajv-formats@3
+
+      - run: node .github/scripts/validate-dataset.mjs