diff --git a/Makefile b/Makefile
index 58df4ac5f..854e8669c 100644
--- a/Makefile
+++ b/Makefile
@@ -75,6 +75,7 @@ docs-site: docs-commands
 	@node scripts/build-docs-site.mjs
 
 docs-check: docs-site
+	@node --test scripts/check-docs-coverage.test.mjs
 	@node scripts/check-docs-coverage.mjs
 
 tools:
diff --git a/scripts/check-docs-coverage.mjs b/scripts/check-docs-coverage.mjs
index 65de52d92..b1ea80962 100644
--- a/scripts/check-docs-coverage.mjs
+++ b/scripts/check-docs-coverage.mjs
@@ -2,6 +2,7 @@
 import { execFileSync } from "node:child_process";
 import fs from "node:fs";
 import path from "node:path";
+import { fileURLToPath } from "node:url";
 
 const root = process.cwd();
 const bin = process.env.GOG_BIN || path.join(root, "bin", "gog");
@@ -34,53 +35,66 @@ const requiredFeatureDocs = [
   "dates.md",
 ];
 
-const schema = JSON.parse(execFileSync(bin, ["schema", "--json"], { encoding: "utf8", maxBuffer: 16 * 1024 * 1024 }));
-const commands = Array.from(walk(schema.command || {}));
-const seenSlugs = new Set();
-const missingCommandPages = [];
-
-for (const command of commands) {
-  const base = commandSlug(command);
-  let slug = base;
-  let suffix = 2;
-  while (seenSlugs.has(slug)) {
-    slug = `${base}-${suffix}`;
-    suffix += 1;
-  }
-  seenSlugs.add(slug);
+function main() {
+  const schema = JSON.parse(
+    execFileSync(bin, ["schema", "--json"], { encoding: "utf8", maxBuffer: 16 * 1024 * 1024 }),
+  );
+  const commands = Array.from(walk(schema.command || {}));
+  const seenSlugs = new Set();
+  const missingCommandPages = [];
+
+  for (const command of commands) {
+    const base = commandSlug(command);
+    let slug = base;
+    let suffix = 2;
+    while (seenSlugs.has(slug)) {
+      slug = `${base}-${suffix}`;
+      suffix += 1;
+    }
+    seenSlugs.add(slug);
 
-  const page = path.join(commandsDir, `${slug}.md`);
-  if (!fs.existsSync(page)) {
-    missingCommandPages.push(path.relative(root, page));
+    const page = path.join(commandsDir, `${slug}.md`);
+    if (!fs.existsSync(page)) {
+      missingCommandPages.push(path.relative(root, page));
+    }
   }
-}
 
-const navSourcePath = path.join(root, "scripts", "build-docs-site.mjs");
-const navSource = fs.readFileSync(navSourcePath, "utf8");
-const missingFeaturePages = [];
-const unlinkedFeaturePages = [];
-const brokenLinks = checkMarkdownLinks(docsDir);
-
-for (const rel of requiredFeatureDocs) {
-  const page = path.join(docsDir, rel);
-  if (!fs.existsSync(page)) {
-    missingFeaturePages.push(`docs/${rel}`);
-    continue;
+  const navSourcePath = path.join(root, "scripts", "build-docs-site.mjs");
+  const navSource = fs.readFileSync(navSourcePath, "utf8");
+  const missingFeaturePages = [];
+  const unlinkedFeaturePages = [];
+  const brokenLinks = checkMarkdownLinks(docsDir);
+
+  for (const rel of requiredFeatureDocs) {
+    const page = path.join(docsDir, rel);
+    if (!fs.existsSync(page)) {
+      missingFeaturePages.push(`docs/${rel}`);
+      continue;
+    }
+    if (!navSource.includes(`"${rel}"`)) {
+      unlinkedFeaturePages.push(`docs/${rel}`);
+    }
   }
-  if (!navSource.includes(`"${rel}"`)) {
-    unlinkedFeaturePages.push(`docs/${rel}`);
+
+  if (
+    missingCommandPages.length ||
+    missingFeaturePages.length ||
+    unlinkedFeaturePages.length ||
+    brokenLinks.length
+  ) {
+    for (const name of missingCommandPages) console.error(`missing command doc: ${name}`);
+    for (const name of missingFeaturePages) console.error(`missing feature doc: ${name}`);
+    for (const name of unlinkedFeaturePages) console.error(`feature doc not in scripts/build-docs-site.mjs sidebar: ${name}`);
+    for (const item of brokenLinks) console.error(`broken docs link: ${item}`);
+    process.exit(1);
   }
-}
 
-if (missingCommandPages.length || missingFeaturePages.length || unlinkedFeaturePages.length || brokenLinks.length) {
-  for (const name of missingCommandPages) console.error(`missing command doc: ${name}`);
-  for (const name of missingFeaturePages) console.error(`missing feature doc: ${name}`);
-  for (const name of unlinkedFeaturePages) console.error(`feature doc not in scripts/build-docs-site.mjs sidebar: ${name}`);
-  for (const item of brokenLinks) console.error(`broken docs link: ${item}`);
-  process.exit(1);
+  console.log(`docs coverage ok: ${commands.length} command pages, ${requiredFeatureDocs.length} feature pages`);
 }
 
-console.log(`docs coverage ok: ${commands.length} command pages, ${requiredFeatureDocs.length} feature pages`);
+if (process.argv[1] && path.resolve(process.argv[1]) === fileURLToPath(import.meta.url)) {
+  main();
+}
 
 function* walk(command) {
   yield command;
@@ -107,31 +121,104 @@ function commandSlug(command) {
   return slug || "gog";
 }
 
-function checkMarkdownLinks(dir) {
+export function checkMarkdownLinks(dir) {
   const broken = [];
   for (const file of allMarkdown(dir)) {
     const markdown = fs.readFileSync(file, "utf8");
+    const headings = headingAnchors(markdown);
     const linkPattern = /!?\[[^\]]*\]\(([^)]+)\)/g;
     let match;
     while ((match = linkPattern.exec(markdown)) !== null) {
-      const rawTarget = match[1].trim().replace(/^<|>$/g, "");
-      if (!rawTarget || rawTarget.startsWith("#")) continue;
+      const rawTarget = splitMarkdownTarget(match[1].trim());
+      if (!rawTarget) continue;
       if (/^[a-z][a-z0-9+.-]*:/i.test(rawTarget)) continue;
 
-      const targetWithoutTitle = rawTarget.split(/\s+["'][^"']*["']\s*$/)[0];
-      const targetPath = targetWithoutTitle.split("#")[0];
-      if (!targetPath) continue;
+      const [rawPath, rawAnchor] = rawTarget.split("#", 2);
+      const targetPath = decodeMarkdownTarget(rawPath);
       if (/^(url|path|file)$/i.test(targetPath)) continue;
 
-      const resolved = path.resolve(path.dirname(file), targetPath);
+      const resolved = targetPath ? path.resolve(path.dirname(file), targetPath) : file;
       if (!fs.existsSync(resolved)) {
         broken.push(`${path.relative(root, file)} -> ${targetPath}`);
+        continue;
+      }
+
+      if (rawAnchor && resolved.toLowerCase().endsWith(".md")) {
+        const anchor = decodeMarkdownTarget(rawAnchor);
+        const targetHeadings = resolved === file ? headings : headingAnchors(fs.readFileSync(resolved, "utf8"));
+        if (!targetHeadings.has(anchor)) {
+          broken.push(`${path.relative(root, file)} -> ${rawTarget}`);
+        }
       }
     }
   }
   return broken;
 }
 
+function splitMarkdownTarget(rawTarget) {
+  const targetWithoutTitle = rawTarget.replace(/\s+["'][^"']*["']\s*$/, "");
+  return targetWithoutTitle.replace(/^<|>$/g, "");
+}
+
+function decodeMarkdownTarget(value) {
+  try {
+    return decodeURIComponent(value);
+  } catch {
+    return value;
+  }
+}
+
+export function headingAnchors(markdown) {
+  const anchors = new Set();
+  const occurrences = new Map();
+  let fence = null;
+  for (const rawLine of markdown.split("\n")) {
+    const line = rawLine.endsWith("\r") ? rawLine.slice(0, -1) : rawLine;
+    const fenceMatch = line.match(/^(?: {0,3})(`{3,}|~{3,})(.*)$/);
+    if (fence) {
+      if (
+        fenceMatch &&
+        fenceMatch[1][0] === fence.char &&
+        fenceMatch[1].length >= fence.length &&
+        fenceMatch[2].trim() === ""
+      ) {
+        fence = null;
+      }
+      continue;
+    }
+    if (fenceMatch) {
+      fence = { char: fenceMatch[1][0], length: fenceMatch[1].length };
+      continue;
+    }
+
+    const match = line.match(/^(#{1,6})\s+(.*)$/);
+    if (!match) continue;
+
+    const heading = match[2].replace(/\s+#+\s*$/, "").trim();
+    const base = slugifyHeading(heading);
+    if (!base) continue;
+
+    let anchor = base;
+    while (occurrences.has(anchor)) {
+      const count = (occurrences.get(base) || 0) + 1;
+      occurrences.set(base, count);
+      anchor = `${base}-${count}`;
+    }
+    occurrences.set(anchor, 0);
+    anchors.add(anchor);
+  }
+  return anchors;
+}
+
+function slugifyHeading(text) {
+  return text
+    .replace(/<[^>]*>/g, "")
+    .replace(/`/g, "")
+    .toLowerCase()
+    .replace(/[^\p{L}\p{M}\p{N}\p{Pc}\- ]/gu, "")
+    .replace(/ /g, "-");
+}
+
 function allMarkdown(dir) {
   return fs
     .readdirSync(dir, { withFileTypes: true })
diff --git a/scripts/check-docs-coverage.test.mjs b/scripts/check-docs-coverage.test.mjs
new file mode 100644
index 000000000..c858ad7fb
--- /dev/null
+++ b/scripts/check-docs-coverage.test.mjs
@@ -0,0 +1,68 @@
+import assert from "node:assert/strict";
+import fs from "node:fs";
+import os from "node:os";
+import path from "node:path";
+import test from "node:test";
+
+import { checkMarkdownLinks, headingAnchors } from "./check-docs-coverage.mjs";
+
+test("headingAnchors ignores headings inside fenced code blocks", () => {
+  const anchors = headingAnchors(`# Real Heading
+
+\`\`\`md
+# Not A Heading
+## Duplicate
+\`\`\`
+
+## Duplicate
+## Duplicate
+
+~~~text
+# Also Not A Heading
+~~~
+`);
+
+  assert.equal(anchors.has("not-a-heading"), false);
+  assert.equal(anchors.has("also-not-a-heading"), false);
+  assert.deepEqual([...anchors], ["real-heading", "duplicate", "duplicate-1"]);
+});
+
+test("headingAnchors follows GitHub-style heading slugs", () => {
+  const anchors = headingAnchors(`# What's new?
+## Привет non-latin 你好
+## A  B
+## foo
+## foo
+## foo-1
+## Heading ##
+`);
+
+  assert.deepEqual([...anchors], [
+    "whats-new",
+    "привет-non-latin-你好",
+    "a--b",
+    "foo",
+    "foo-1",
+    "foo-1-1",
+    "heading",
+  ]);
+});
+
+test("checkMarkdownLinks accepts encoded Unicode anchors", (t) => {
+  const dir = fs.mkdtempSync(path.join(os.tmpdir(), "gog-doc-links-"));
+  t.after(() => fs.rmSync(dir, { recursive: true, force: true }));
+
+  fs.writeFileSync(path.join(dir, "target.md"), "# Привет мир\n");
+  fs.writeFileSync(
+    path.join(dir, "index.md"),
+    [
+      "[valid](target.md#%D0%BF%D1%80%D0%B8%D0%B2%D0%B5%D1%82-%D0%BC%D0%B8%D1%80)",
+      "[broken](target.md#missing)",
+      "",
+    ].join("\n"),
+  );
+
+  const broken = checkMarkdownLinks(dir);
+  assert.equal(broken.length, 1);
+  assert.match(broken[0], /target\.md#missing$/);
+});