@agent-seo/next 1.0.0

package/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Pontiggia
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package/dist/index.cjs

@@ -0,0 +1,333 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/index.ts
+var index_exports = {};
+__export(index_exports, {
+  createAgentSeoMiddleware: () => createAgentSeoMiddleware,
+  createLlmsTxtHandler: () => createLlmsTxtHandler,
+  discoverNextRoutes: () => import_core2.discoverNextRoutes,
+  generateLlmsTxt: () => import_core2.generateLlmsTxt,
+  transform: () => import_core2.transform,
+  withAgentSeo: () => withAgentSeo
+});
+module.exports = __toCommonJS(index_exports);
+
+// src/plugin.ts
+var import_node_fs = require("fs");
+var import_node_path = require("path");
+function withAgentSeo(agentSeoOptions) {
+  const appDir = agentSeoOptions.appDir || detectAppDir();
+  if (appDir) {
+    generateLlmsTxtRoute(appDir, agentSeoOptions);
+    generateTransformRoute(appDir, agentSeoOptions);
+    generateRobotsTxt(appDir, agentSeoOptions);
+  }
+  return (nextConfig = {}) => {
+    return {
+      ...nextConfig,
+      async headers() {
+        const existingHeaders = await (nextConfig.headers?.() ?? []);
+        return [
+          ...existingHeaders,
+          {
+            source: "/((?!api|_next|static|favicon.ico).*)",
+            headers: [{ key: "Vary", value: "Accept, User-Agent" }]
+          }
+        ];
+      }
+    };
+  };
+}
+function detectAppDir() {
+  const cwd = process.cwd();
+  const candidates = [(0, import_node_path.join)(cwd, "app"), (0, import_node_path.join)(cwd, "src", "app")];
+  for (const candidate of candidates) {
+    if ((0, import_node_fs.existsSync)(candidate)) {
+      return candidate;
+    }
+  }
+  return null;
+}
+var GENERATED_BANNER = `// AUTO-GENERATED by @agent-seo/next \u2014 do not edit manually.
+// This file is created by withAgentSeo() in next.config.ts.
+// Add "app/llms.txt" to your .gitignore.
+`;
+function generateLlmsTxtRoute(appDir, options) {
+  const routeDir = (0, import_node_path.join)(appDir, "llms.txt");
+  const routeFile = (0, import_node_path.join)(routeDir, "route.js");
+  const siteName = escapeStr(options.siteName);
+  const siteDescription = escapeStr(options.siteDescription);
+  const baseUrl = escapeStr(options.baseUrl);
+  const excludePatterns = JSON.stringify(options.exclude || ["/api"]);
+  const content = `${GENERATED_BANNER}
+import { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/next';
+import path from 'node:path';
+
+const appDir = path.resolve(process.cwd(), 'app');
+
+export async function GET(request) {
+  const url = new URL(request.url);
+
+  const routes = discoverNextRoutes(appDir, {
+    exclude: ${excludePatterns},
+  });
+
+  const result = generateLlmsTxt(
+    {
+      siteName: '${siteName}',
+      siteDescription: '${siteDescription}',
+      baseUrl: process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}',
+    },
+    routes,
+  );
+
+  return new Response(result.llmsTxt, {
+    status: 200,
+    headers: {
+      'Content-Type': 'text/plain; charset=utf-8',
+      'Content-Disposition': 'inline',
+      'Cache-Control': 'public, max-age=3600, s-maxage=3600',
+    },
+  });
+}
+`;
+  (0, import_node_fs.mkdirSync)(routeDir, { recursive: true });
+  let existingContent = "";
+  try {
+    const { readFileSync } = require("fs");
+    existingContent = readFileSync(routeFile, "utf-8");
+  } catch {
+  }
+  if (existingContent !== content) {
+    (0, import_node_fs.writeFileSync)(routeFile, content, "utf-8");
+  }
+}
+function escapeStr(s) {
+  return s.replace(/\\/g, "\\\\").replace(/'/g, "\\'");
+}
+function generateTransformRoute(appDir, options) {
+  const routeDir = (0, import_node_path.join)(appDir, "api", "agent-seo-transform");
+  const routeFile = (0, import_node_path.join)(routeDir, "route.js");
+  const baseUrl = escapeStr(options.baseUrl);
+  const content = `${GENERATED_BANNER}
+import { transform } from '@agent-seo/next';
+
+// Force Node.js runtime (required for jsdom / Readability / Turndown)
+export const runtime = 'nodejs';
+export const dynamic = 'force-dynamic';
+
+export async function GET(request) {
+  const { searchParams } = new URL(request.url);
+  const pagePath = searchParams.get('path') || '/';
+
+  // Build the internal URL to fetch the original HTML page
+  const origin = process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}';
+  if (!origin) {
+    return new Response('Base URL not configured', { status: 500 });
+  }
+
+  // Only allow same-origin, absolute paths (prevent SSRF via absolute or protocol-relative URLs)
+  const isValidPath =
+    pagePath.startsWith('/') &&
+    !pagePath.startsWith('//') &&
+    !pagePath.includes('://') &&
+    !pagePath.includes('\\\\');
+  if (!isValidPath) {
+    return new Response('Invalid path', { status: 400 });
+  }
+
+  const originUrl = new URL(origin);
+  if (originUrl.protocol !== 'http:' && originUrl.protocol !== 'https:') {
+    return new Response('Invalid base URL', { status: 500 });
+  }
+  const pageUrl = new URL(pagePath, originUrl.origin);
+  if (pageUrl.origin !== originUrl.origin) {
+    return new Response('Invalid origin', { status: 400 });
+  }
+
+  try {
+    // Fetch the page HTML from the local server with a normal User-Agent
+    // to avoid infinite rewrite loops
+    const res = await fetch(pageUrl.toString(), {
+      headers: {
+        'User-Agent': 'AgentSEO-Internal/1.0',
+        'Accept': 'text/html',
+      },
+    });
+
+    if (!res.ok) {
+      return new Response('Page not found', { status: 404 });
+    }
+
+    const html = await res.text();
+
+    const result = await transform(html, {
+      url: pageUrl.toString(),
+    });
+
+    return new Response(result.markdown, {
+      status: 200,
+      headers: {
+        'Content-Type': 'text/markdown; charset=utf-8',
+        'Content-Disposition': 'inline',
+        'Cache-Control': 'public, max-age=3600, s-maxage=3600',
+        'Vary': 'Accept, User-Agent',
+        'X-Robots-Tag': 'all',
+        'X-Agent-Seo': 'transformed',
+      },
+    });
+  } catch (err) {
+    console.error('[agent-seo] Transform error:', err);
+    return new Response('Transform failed', { status: 500 });
+  }
+}
+`;
+  (0, import_node_fs.mkdirSync)(routeDir, { recursive: true });
+  let existingContent = "";
+  try {
+    const { readFileSync } = require("fs");
+    existingContent = readFileSync(routeFile, "utf-8");
+  } catch {
+  }
+  if (existingContent !== content) {
+    (0, import_node_fs.writeFileSync)(routeFile, content, "utf-8");
+  }
+}
+function generateRobotsTxt(appDir, options) {
+  const robotsFile = (0, import_node_path.join)(appDir, "robots.txt");
+  if ((0, import_node_fs.existsSync)(robotsFile)) {
+    return;
+  }
+  let content = `# AI-Optimized robots.txt \u2014 Generated by @agent-seo/next
+# For AI agents: see /llms.txt for a structured site manifest
+
+User-agent: *
+Allow: /
+`;
+  if (options.sitemap) {
+    const sitemapUrl = typeof options.sitemap === "string" ? options.sitemap : `${options.baseUrl}/sitemap.xml`;
+    content += `
+Sitemap: ${sitemapUrl}
+`;
+  }
+  (0, import_node_fs.writeFileSync)(robotsFile, content, "utf-8");
+}
+
+// src/middleware.ts
+var import_edge = require("@agent-seo/core/edge");
+var import_server = require("next/server");
+var ALWAYS_SKIP = /* @__PURE__ */ new Set([
+  "/favicon.ico",
+  "/robots.txt",
+  "/sitemap.xml",
+  "/llms.txt",
+  "/llms-full.txt"
+]);
+var DEFAULT_EXCLUDE = [
+  "/api/**",
+  "/_next/**"
+];
+function createAgentSeoMiddleware(options = {}) {
+  const excludePatterns = [...DEFAULT_EXCLUDE, ...options.exclude || []];
+  return function middleware(request) {
+    const { pathname } = request.nextUrl;
+    if (ALWAYS_SKIP.has(pathname)) {
+      return import_server.NextResponse.next();
+    }
+    if (isExcluded(pathname, excludePatterns)) {
+      return import_server.NextResponse.next();
+    }
+    const ua = request.headers.get("user-agent");
+    const accept = request.headers.get("accept");
+    const aiCtx = (0, import_edge.detectAgent)(ua, accept);
+    if (pathname.endsWith(".md")) {
+      const originalPath = pathname.slice(0, -3) || "/";
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", originalPath);
+      return setBotHeaders(import_server.NextResponse.rewrite(transformUrl));
+    }
+    if (aiCtx.isAIBot) {
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", pathname);
+      return setBotHeaders(import_server.NextResponse.rewrite(transformUrl));
+    }
+    const response = import_server.NextResponse.next();
+    response.headers.set("Vary", "Accept, User-Agent");
+    return response;
+  };
+}
+function setBotHeaders(response) {
+  response.headers.set("Content-Disposition", "inline");
+  response.headers.set("Vary", "Accept, User-Agent");
+  response.headers.set("X-Robots-Tag", "all");
+  response.headers.delete("x-nextjs-matched-path");
+  return response;
+}
+function isExcluded(path, patterns) {
+  return patterns.some((pattern) => matchGlob(pattern, path));
+}
+function matchGlob(pattern, path) {
+  const regex = pattern.replace(/\*\*/g, "{{DOUBLESTAR}}").replace(/\*/g, "[^/]*").replace(/{{DOUBLESTAR}}/g, ".*");
+  return new RegExp(`^${regex}$`).test(path);
+}
+
+// src/route-handler.ts
+var import_core = require("@agent-seo/core");
+function createLlmsTxtHandler(options) {
+  return async function GET() {
+    let routes = options.llmsTxt?.routes || [];
+    if (routes.length === 0 && options.appDir) {
+      routes = (0, import_core.discoverNextRoutes)(options.appDir, {
+        exclude: options.exclude || ["/api"],
+        ...options.discoverOptions
+      });
+    }
+    const result = (0, import_core.generateLlmsTxt)(
+      {
+        siteName: options.siteName,
+        siteDescription: options.siteDescription,
+        baseUrl: options.baseUrl,
+        ...options.llmsTxt
+      },
+      routes
+    );
+    const content = options.full ? result.llmsFullTxt : result.llmsTxt;
+    return new Response(content, {
+      status: 200,
+      headers: {
+        "Content-Type": "text/plain; charset=utf-8",
+        "Cache-Control": "public, max-age=3600, s-maxage=3600"
+      }
+    });
+  };
+}
+
+// src/index.ts
+var import_core2 = require("@agent-seo/core");
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  createAgentSeoMiddleware,
+  createLlmsTxtHandler,
+  discoverNextRoutes,
+  generateLlmsTxt,
+  transform,
+  withAgentSeo
+});
+//# sourceMappingURL=index.cjs.map

package/dist/index.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/index.ts","../src/plugin.ts","../src/middleware.ts","../src/route-handler.ts"],"sourcesContent":["export { withAgentSeo } from './plugin.js';\nexport type { WithAgentSeoOptions } from './plugin.js';\nexport { createAgentSeoMiddleware } from './middleware.js';\nexport type { AgentSeoMiddlewareOptions } from './middleware.js';\nexport { createLlmsTxtHandler } from './route-handler.js';\nexport type { LlmsTxtHandlerOptions } from './route-handler.js';\nexport {\n  generateLlmsTxt,\n  discoverNextRoutes,\n  transform,\n} from '@agent-seo/core';\nexport type {\n  AgentSeoOptions,\n  AIRequestContext,\n  TransformResult,\n  BotInfo,\n  BotPurpose,\n  LlmsTxtRoute,\n  DiscoverOptions,\n} from '@agent-seo/core';\n","import { existsSync, mkdirSync, writeFileSync } from 'node:fs';\nimport { join, resolve } from 'node:path';\nimport type { NextConfig } from 'next';\nimport type { AgentSeoOptions } from '@agent-seo/core';\n\nexport interface WithAgentSeoOptions extends AgentSeoOptions {\n  /**\n   * Absolute path to the Next.js `app/` directory.\n   * If omitted, auto-detected by checking `./app` and `./src/app` from cwd.\n   */\n  appDir?: string;\n  /**\n   * Sitemap URL for the `robots.txt` Sitemap directive.\n   * - `true`: uses `{baseUrl}/sitemap.xml`\n   * - `string`: uses the provided URL\n   * - `undefined`/`false`: no Sitemap directive\n   */\n  sitemap?: boolean | string;\n}\n\n/**\n * Next.js config plugin that:\n * 1. Auto-generates `app/llms.txt/route.ts` (zero-config `/llms.txt` endpoint)\n * 2. Injects `Vary: Accept, User-Agent` headers on all pages\n *\n * @example\n * ```ts\n * // next.config.ts\n * import { withAgentSeo } from '@agent-seo/next';\n *\n * export default withAgentSeo({\n *   siteName: 'My App',\n *   siteDescription: 'A brief description for LLMs.',\n *   baseUrl: 'https://myapp.com',\n * })({});\n * ```\n */\nexport function withAgentSeo(agentSeoOptions: WithAgentSeoOptions) {\n  const appDir = agentSeoOptions.appDir || detectAppDir();\n\n  // Auto-generate route handlers at config-evaluation time\n  if (appDir) {\n    generateLlmsTxtRoute(appDir, agentSeoOptions);\n    generateTransformRoute(appDir, agentSeoOptions);\n    generateRobotsTxt(appDir, agentSeoOptions);\n  }\n\n  return (nextConfig: NextConfig = {}): NextConfig => {\n    return {\n      ...nextConfig,\n\n      async headers() {\n        const existingHeaders = await (nextConfig.headers?.() ?? []);\n        return [\n          ...existingHeaders,\n          {\n            source: '/((?!api|_next|static|favicon.ico).*)',\n            headers: [{ key: 'Vary', value: 'Accept, User-Agent' }],\n          },\n        ];\n      },\n    };\n  };\n}\n\n/**\n * Auto-detect the Next.js `app/` directory from common locations.\n */\nfunction detectAppDir(): string | null {\n  const cwd = process.cwd();\n  const candidates = [join(cwd, 'app'), join(cwd, 'src', 'app')];\n\n  for (const candidate of candidates) {\n    if (existsSync(candidate)) {\n      return candidate;\n    }\n  }\n\n  return null;\n}\n\n// Auto-generated file banner\nconst GENERATED_BANNER = `// AUTO-GENERATED by @agent-seo/next — do not edit manually.\n// This file is created by withAgentSeo() in next.config.ts.\n// Add \"app/llms.txt\" to your .gitignore.\n`;\n\n/**\n * Write `app/llms.txt/route.js` with a handler that auto-discovers routes.\n */\nfunction generateLlmsTxtRoute(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const routeDir = join(appDir, 'llms.txt');\n  const routeFile = join(routeDir, 'route.js');\n\n  // Serialize the options we need into the generated file\n  const siteName = escapeStr(options.siteName);\n  const siteDescription = escapeStr(options.siteDescription);\n  const baseUrl = escapeStr(options.baseUrl);\n  const excludePatterns = JSON.stringify(options.exclude || ['/api']);\n\n  const content = `${GENERATED_BANNER}\nimport { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/next';\nimport path from 'node:path';\n\nconst appDir = path.resolve(process.cwd(), 'app');\n\nexport async function GET(request) {\n  const url = new URL(request.url);\n\n  const routes = discoverNextRoutes(appDir, {\n    exclude: ${excludePatterns},\n  });\n\n  const result = generateLlmsTxt(\n    {\n      siteName: '${siteName}',\n      siteDescription: '${siteDescription}',\n      baseUrl: process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}',\n    },\n    routes,\n  );\n\n  return new Response(result.llmsTxt, {\n    status: 200,\n    headers: {\n      'Content-Type': 'text/plain; charset=utf-8',\n      'Content-Disposition': 'inline',\n      'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n    },\n  });\n}\n`;\n\n  // Only write if the file doesn't exist or content has changed\n  mkdirSync(routeDir, { recursive: true });\n\n  let existingContent = '';\n  try {\n    const { readFileSync } = require('node:fs');\n    existingContent = readFileSync(routeFile, 'utf-8');\n  } catch {\n    // File doesn't exist yet\n  }\n\n  if (existingContent !== content) {\n    writeFileSync(routeFile, content, 'utf-8');\n  }\n}\n\nfunction escapeStr(s: string): string {\n  return s.replace(/\\\\/g, '\\\\\\\\').replace(/'/g, \"\\\\'\");\n}\n\n/**\n * Write `app/api/agent-seo-transform/route.js` — a Node.js API route that\n * receives a `?path=/some-page` param, fetches the HTML from the local\n * Next.js server, runs the core `transform()` pipeline (JSDOM → Readability\n * → Turndown), and returns clean Markdown.\n */\nfunction generateTransformRoute(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const routeDir = join(appDir, 'api', 'agent-seo-transform');\n  const routeFile = join(routeDir, 'route.js');\n\n  const baseUrl = escapeStr(options.baseUrl);\n\n  const content = `${GENERATED_BANNER}\nimport { transform } from '@agent-seo/next';\n\n// Force Node.js runtime (required for jsdom / Readability / Turndown)\nexport const runtime = 'nodejs';\nexport const dynamic = 'force-dynamic';\n\nexport async function GET(request) {\n  const { searchParams } = new URL(request.url);\n  const pagePath = searchParams.get('path') || '/';\n\n  // Build the internal URL to fetch the original HTML page\n  const origin = process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}';\n  if (!origin) {\n    return new Response('Base URL not configured', { status: 500 });\n  }\n\n  // Only allow same-origin, absolute paths (prevent SSRF via absolute or protocol-relative URLs)\n  const isValidPath =\n    pagePath.startsWith('/') &&\n    !pagePath.startsWith('//') &&\n    !pagePath.includes('://') &&\n    !pagePath.includes('\\\\\\\\');\n  if (!isValidPath) {\n    return new Response('Invalid path', { status: 400 });\n  }\n\n  const originUrl = new URL(origin);\n  if (originUrl.protocol !== 'http:' && originUrl.protocol !== 'https:') {\n    return new Response('Invalid base URL', { status: 500 });\n  }\n  const pageUrl = new URL(pagePath, originUrl.origin);\n  if (pageUrl.origin !== originUrl.origin) {\n    return new Response('Invalid origin', { status: 400 });\n  }\n\n  try {\n    // Fetch the page HTML from the local server with a normal User-Agent\n    // to avoid infinite rewrite loops\n    const res = await fetch(pageUrl.toString(), {\n      headers: {\n        'User-Agent': 'AgentSEO-Internal/1.0',\n        'Accept': 'text/html',\n      },\n    });\n\n    if (!res.ok) {\n      return new Response('Page not found', { status: 404 });\n    }\n\n    const html = await res.text();\n\n    const result = await transform(html, {\n      url: pageUrl.toString(),\n    });\n\n    return new Response(result.markdown, {\n      status: 200,\n      headers: {\n        'Content-Type': 'text/markdown; charset=utf-8',\n        'Content-Disposition': 'inline',\n        'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n        'Vary': 'Accept, User-Agent',\n        'X-Robots-Tag': 'all',\n        'X-Agent-Seo': 'transformed',\n      },\n    });\n  } catch (err) {\n    console.error('[agent-seo] Transform error:', err);\n    return new Response('Transform failed', { status: 500 });\n  }\n}\n`;\n\n  mkdirSync(routeDir, { recursive: true });\n\n  let existingContent = '';\n  try {\n    const { readFileSync } = require('node:fs');\n    existingContent = readFileSync(routeFile, 'utf-8');\n  } catch {\n    // File doesn't exist yet\n  }\n\n  if (existingContent !== content) {\n    writeFileSync(routeFile, content, 'utf-8');\n  }\n}\n\n/**\n * Write `app/robots.txt` — a static robots.txt that allows all crawlers\n * and references /llms.txt for AI agents.\n * Only created if the file doesn't already exist (won't overwrite user customizations).\n */\nfunction generateRobotsTxt(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const robotsFile = join(appDir, 'robots.txt');\n\n  // Don't overwrite user-created robots.txt\n  if (existsSync(robotsFile)) {\n    return;\n  }\n\n  let content = `# AI-Optimized robots.txt — Generated by @agent-seo/next\n# For AI agents: see /llms.txt for a structured site manifest\n\nUser-agent: *\nAllow: /\n`;\n\n  if (options.sitemap) {\n    const sitemapUrl =\n      typeof options.sitemap === 'string'\n        ? options.sitemap\n        : `${options.baseUrl}/sitemap.xml`;\n    content += `\\nSitemap: ${sitemapUrl}\\n`;\n  }\n\n  writeFileSync(robotsFile, content, 'utf-8');\n}\n","import { detectAgent } from '@agent-seo/core/edge';\nimport { NextResponse } from 'next/server';\nimport type { NextRequest } from 'next/server';\n\nexport interface AgentSeoMiddlewareOptions {\n  /**\n   * Glob patterns for paths that should NEVER be rewritten to Markdown.\n   * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).\n   *\n   * @example ['\\/dashboard\\/**', '\\/admin\\/**', '\\/api\\/private\\/**']\n   */\n  exclude?: string[];\n}\n\n/** Paths that are always skipped — framework internals + standard files. */\nconst ALWAYS_SKIP = new Set([\n  '/favicon.ico',\n  '/robots.txt',\n  '/sitemap.xml',\n  '/llms.txt',\n  '/llms-full.txt',\n]);\n\n/** Default exclude patterns — users can extend but not remove these. */\nconst DEFAULT_EXCLUDE: string[] = [\n  '/api/**',\n  '/_next/**',\n];\n\n/**\n * Creates a Next.js middleware that:\n * 1. Detects AI bot requests via User-Agent\n * 2. Rewrites AI bot requests to an internal transform API route\n *    that converts HTML → Markdown (runs on Node.js runtime)\n * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)\n * 4. Sets `Vary: Accept, User-Agent` on all responses\n */\nexport function createAgentSeoMiddleware(options: AgentSeoMiddlewareOptions = {}) {\n  const excludePatterns = [...DEFAULT_EXCLUDE, ...(options.exclude || [])];\n\n  return function middleware(request: NextRequest) {\n    const { pathname } = request.nextUrl;\n\n    // Skip standard files (robots.txt, favicon, llms.txt, etc.)\n    if (ALWAYS_SKIP.has(pathname)) {\n      return NextResponse.next();\n    }\n\n    // Skip excluded patterns (API routes, admin, dashboard, etc.)\n    if (isExcluded(pathname, excludePatterns)) {\n      return NextResponse.next();\n    }\n\n    const ua = request.headers.get('user-agent');\n    const accept = request.headers.get('accept');\n    const aiCtx = detectAgent(ua, accept);\n\n    // Handle explicit .md suffix requests (e.g. /about.md)\n    if (pathname.endsWith('.md')) {\n      const originalPath = pathname.slice(0, -3) || '/';\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', originalPath);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // If AI bot, rewrite to transform API\n    if (aiCtx.isAIBot) {\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', pathname);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // Normal request — just add Vary header\n    const response = NextResponse.next();\n    response.headers.set('Vary', 'Accept, User-Agent');\n    return response;\n  };\n}\n\n/**\n * Set clean, bot-friendly headers on a rewrite response.\n * Overrides Next.js RSC-related headers that can confuse AI bots.\n */\nfunction setBotHeaders(response: NextResponse): NextResponse {\n  response.headers.set('Content-Disposition', 'inline');\n  response.headers.set('Vary', 'Accept, User-Agent');\n  response.headers.set('X-Robots-Tag', 'all');\n  response.headers.delete('x-nextjs-matched-path');\n  return response;\n}\n\nfunction isExcluded(path: string, patterns: string[]): boolean {\n  return patterns.some((pattern) => matchGlob(pattern, path));\n}\n\nfunction matchGlob(pattern: string, path: string): boolean {\n  const regex = pattern\n    .replace(/\\*\\*/g, '{{DOUBLESTAR}}')\n    .replace(/\\*/g, '[^/]*')\n    .replace(/{{DOUBLESTAR}}/g, '.*');\n  return new RegExp(`^${regex}$`).test(path);\n}\n","import { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/core';\nimport type { AgentSeoOptions, DiscoverOptions } from '@agent-seo/core';\n\nexport interface LlmsTxtHandlerOptions extends AgentSeoOptions {\n  /** Return llms-full.txt instead of llms.txt */\n  full?: boolean;\n\n  /**\n   * Absolute path to the Next.js `app/` directory for automatic route discovery.\n   * When set, routes for llms.txt are auto-discovered by scanning page.tsx files.\n   * Titles and descriptions are extracted from `export const metadata` in each page.\n   *\n   * @example\n   * ```ts\n   * appDir: path.resolve(process.cwd(), 'app')\n   * ```\n   */\n  appDir?: string;\n\n  /** Options for route discovery when using appDir */\n  discoverOptions?: DiscoverOptions;\n}\n\nexport function createLlmsTxtHandler(options: LlmsTxtHandlerOptions) {\n  return async function GET() {\n    // Auto-discover routes from app/ directory if no explicit routes are provided\n    let routes = options.llmsTxt?.routes || [];\n\n    if (routes.length === 0 && options.appDir) {\n      routes = discoverNextRoutes(options.appDir, {\n        exclude: options.exclude || ['/api'],\n        ...options.discoverOptions,\n      });\n    }\n\n    const result = generateLlmsTxt(\n      {\n        siteName: options.siteName,\n        siteDescription: options.siteDescription,\n        baseUrl: options.baseUrl,\n        ...options.llmsTxt,\n      },\n      routes,\n    );\n\n    const content = options.full ? result.llmsFullTxt : result.llmsTxt;\n\n    return new Response(content, {\n      status: 200,\n      headers: {\n        'Content-Type': 'text/plain; charset=utf-8',\n        'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n      },\n    });\n  };\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;;;ACAA,qBAAqD;AACrD,uBAA8B;AAoCvB,SAAS,aAAa,iBAAsC;AACjE,QAAM,SAAS,gBAAgB,UAAU,aAAa;AAGtD,MAAI,QAAQ;AACV,yBAAqB,QAAQ,eAAe;AAC5C,2BAAuB,QAAQ,eAAe;AAC9C,sBAAkB,QAAQ,eAAe;AAAA,EAC3C;AAEA,SAAO,CAAC,aAAyB,CAAC,MAAkB;AAClD,WAAO;AAAA,MACL,GAAG;AAAA,MAEH,MAAM,UAAU;AACd,cAAM,kBAAkB,OAAO,WAAW,UAAU,KAAK,CAAC;AAC1D,eAAO;AAAA,UACL,GAAG;AAAA,UACH;AAAA,YACE,QAAQ;AAAA,YACR,SAAS,CAAC,EAAE,KAAK,QAAQ,OAAO,qBAAqB,CAAC;AAAA,UACxD;AAAA,QACF;AAAA,MACF;AAAA,IACF;AAAA,EACF;AACF;AAKA,SAAS,eAA8B;AACrC,QAAM,MAAM,QAAQ,IAAI;AACxB,QAAM,aAAa,KAAC,uBAAK,KAAK,KAAK,OAAG,uBAAK,KAAK,OAAO,KAAK,CAAC;AAE7D,aAAW,aAAa,YAAY;AAClC,YAAI,2BAAW,SAAS,GAAG;AACzB,aAAO;AAAA,IACT;AAAA,EACF;AAEA,SAAO;AACT;AAGA,IAAM,mBAAmB;AAAA;AAAA;AAAA;AAQzB,SAAS,qBACP,QACA,SACM;AACN,QAAM,eAAW,uBAAK,QAAQ,UAAU;AACxC,QAAM,gBAAY,uBAAK,UAAU,UAAU;AAG3C,QAAM,WAAW,UAAU,QAAQ,QAAQ;AAC3C,QAAM,kBAAkB,UAAU,QAAQ,eAAe;AACzD,QAAM,UAAU,UAAU,QAAQ,OAAO;AACzC,QAAM,kBAAkB,KAAK,UAAU,QAAQ,WAAW,CAAC,MAAM,CAAC;AAElE,QAAM,UAAU,GAAG,gBAAgB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,eAUtB,eAAe;AAAA;AAAA;AAAA;AAAA;AAAA,mBAKX,QAAQ;AAAA,0BACD,eAAe;AAAA,sDACa,OAAO;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAiB3D,gCAAU,UAAU,EAAE,WAAW,KAAK,CAAC;AAEvC,MAAI,kBAAkB;AACtB,MAAI;AACF,UAAM,EAAE,aAAa,IAAI,QAAQ,IAAS;AAC1C,sBAAkB,aAAa,WAAW,OAAO;AAAA,EACnD,QAAQ;AAAA,EAER;AAEA,MAAI,oBAAoB,SAAS;AAC/B,sCAAc,WAAW,SAAS,OAAO;AAAA,EAC3C;AACF;AAEA,SAAS,UAAU,GAAmB;AACpC,SAAO,EAAE,QAAQ,OAAO,MAAM,EAAE,QAAQ,MAAM,KAAK;AACrD;AAQA,SAAS,uBACP,QACA,SACM;AACN,QAAM,eAAW,uBAAK,QAAQ,OAAO,qBAAqB;AAC1D,QAAM,gBAAY,uBAAK,UAAU,UAAU;AAE3C,QAAM,UAAU,UAAU,QAAQ,OAAO;AAEzC,QAAM,UAAU,GAAG,gBAAgB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,wDAYmB,OAAO;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AA8D7D,gCAAU,UAAU,EAAE,WAAW,KAAK,CAAC;AAEvC,MAAI,kBAAkB;AACtB,MAAI;AACF,UAAM,EAAE,aAAa,IAAI,QAAQ,IAAS;AAC1C,sBAAkB,aAAa,WAAW,OAAO;AAAA,EACnD,QAAQ;AAAA,EAER;AAEA,MAAI,oBAAoB,SAAS;AAC/B,sCAAc,WAAW,SAAS,OAAO;AAAA,EAC3C;AACF;AAOA,SAAS,kBACP,QACA,SACM;AACN,QAAM,iBAAa,uBAAK,QAAQ,YAAY;AAG5C,UAAI,2BAAW,UAAU,GAAG;AAC1B;AAAA,EACF;AAEA,MAAI,UAAU;AAAA;AAAA;AAAA;AAAA;AAAA;AAOd,MAAI,QAAQ,SAAS;AACnB,UAAM,aACJ,OAAO,QAAQ,YAAY,WACvB,QAAQ,UACR,GAAG,QAAQ,OAAO;AACxB,eAAW;AAAA,WAAc,UAAU;AAAA;AAAA,EACrC;AAEA,oCAAc,YAAY,SAAS,OAAO;AAC5C;;;ACpSA,kBAA4B;AAC5B,oBAA6B;AAc7B,IAAM,cAAc,oBAAI,IAAI;AAAA,EAC1B;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF,CAAC;AAGD,IAAM,kBAA4B;AAAA,EAChC;AAAA,EACA;AACF;AAUO,SAAS,yBAAyB,UAAqC,CAAC,GAAG;AAChF,QAAM,kBAAkB,CAAC,GAAG,iBAAiB,GAAI,QAAQ,WAAW,CAAC,CAAE;AAEvE,SAAO,SAAS,WAAW,SAAsB;AAC/C,UAAM,EAAE,SAAS,IAAI,QAAQ;AAG7B,QAAI,YAAY,IAAI,QAAQ,GAAG;AAC7B,aAAO,2BAAa,KAAK;AAAA,IAC3B;AAGA,QAAI,WAAW,UAAU,eAAe,GAAG;AACzC,aAAO,2BAAa,KAAK;AAAA,IAC3B;AAEA,UAAM,KAAK,QAAQ,QAAQ,IAAI,YAAY;AAC3C,UAAM,SAAS,QAAQ,QAAQ,IAAI,QAAQ;AAC3C,UAAM,YAAQ,yBAAY,IAAI,MAAM;AAGpC,QAAI,SAAS,SAAS,KAAK,GAAG;AAC5B,YAAM,eAAe,SAAS,MAAM,GAAG,EAAE,KAAK;AAC9C,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,YAAY;AAClD,aAAO,cAAc,2BAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,QAAI,MAAM,SAAS;AACjB,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,QAAQ;AAC9C,aAAO,cAAc,2BAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,UAAM,WAAW,2BAAa,KAAK;AACnC,aAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAO;AAAA,EACT;AACF;AAMA,SAAS,cAAc,UAAsC;AAC3D,WAAS,QAAQ,IAAI,uBAAuB,QAAQ;AACpD,WAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAS,QAAQ,IAAI,gBAAgB,KAAK;AAC1C,WAAS,QAAQ,OAAO,uBAAuB;AAC/C,SAAO;AACT;AAEA,SAAS,WAAW,MAAc,UAA6B;AAC7D,SAAO,SAAS,KAAK,CAAC,YAAY,UAAU,SAAS,IAAI,CAAC;AAC5D;AAEA,SAAS,UAAU,SAAiB,MAAuB;AACzD,QAAM,QAAQ,QACX,QAAQ,SAAS,gBAAgB,EACjC,QAAQ,OAAO,OAAO,EACtB,QAAQ,mBAAmB,IAAI;AAClC,SAAO,IAAI,OAAO,IAAI,KAAK,GAAG,EAAE,KAAK,IAAI;AAC3C;;;ACrGA,kBAAoD;AAuB7C,SAAS,qBAAqB,SAAgC;AACnE,SAAO,eAAe,MAAM;AAE1B,QAAI,SAAS,QAAQ,SAAS,UAAU,CAAC;AAEzC,QAAI,OAAO,WAAW,KAAK,QAAQ,QAAQ;AACzC,mBAAS,gCAAmB,QAAQ,QAAQ;AAAA,QAC1C,SAAS,QAAQ,WAAW,CAAC,MAAM;AAAA,QACnC,GAAG,QAAQ;AAAA,MACb,CAAC;AAAA,IACH;AAEA,UAAM,aAAS;AAAA,MACb;AAAA,QACE,UAAU,QAAQ;AAAA,QAClB,iBAAiB,QAAQ;AAAA,QACzB,SAAS,QAAQ;AAAA,QACjB,GAAG,QAAQ;AAAA,MACb;AAAA,MACA;AAAA,IACF;AAEA,UAAM,UAAU,QAAQ,OAAO,OAAO,cAAc,OAAO;AAE3D,WAAO,IAAI,SAAS,SAAS;AAAA,MAC3B,QAAQ;AAAA,MACR,SAAS;AAAA,QACP,gBAAgB;AAAA,QAChB,iBAAiB;AAAA,MACnB;AAAA,IACF,CAAC;AAAA,EACH;AACF;;;AHjDA,IAAAA,eAIO;","names":["import_core"]}

package/dist/index.d.cts

@@ -0,0 +1,59 @@
+import { NextConfig } from 'next';
+import { AgentSeoOptions, DiscoverOptions } from '@agent-seo/core';
+export { AIRequestContext, AgentSeoOptions, BotInfo, BotPurpose, DiscoverOptions, LlmsTxtRoute, TransformResult, discoverNextRoutes, generateLlmsTxt, transform } from '@agent-seo/core';
+export { AgentSeoMiddlewareOptions, createAgentSeoMiddleware } from './middleware.cjs';
+import 'next/server';
+
+interface WithAgentSeoOptions extends AgentSeoOptions {
+    /**
+     * Absolute path to the Next.js `app/` directory.
+     * If omitted, auto-detected by checking `./app` and `./src/app` from cwd.
+     */
+    appDir?: string;
+    /**
+     * Sitemap URL for the `robots.txt` Sitemap directive.
+     * - `true`: uses `{baseUrl}/sitemap.xml`
+     * - `string`: uses the provided URL
+     * - `undefined`/`false`: no Sitemap directive
+     */
+    sitemap?: boolean | string;
+}
+/**
+ * Next.js config plugin that:
+ * 1. Auto-generates `app/llms.txt/route.ts` (zero-config `/llms.txt` endpoint)
+ * 2. Injects `Vary: Accept, User-Agent` headers on all pages
+ *
+ * @example
+ * ```ts
+ * // next.config.ts
+ * import { withAgentSeo } from '@agent-seo/next';
+ *
+ * export default withAgentSeo({
+ *   siteName: 'My App',
+ *   siteDescription: 'A brief description for LLMs.',
+ *   baseUrl: 'https://myapp.com',
+ * })({});
+ * ```
+ */
+declare function withAgentSeo(agentSeoOptions: WithAgentSeoOptions): (nextConfig?: NextConfig) => NextConfig;
+
+interface LlmsTxtHandlerOptions extends AgentSeoOptions {
+    /** Return llms-full.txt instead of llms.txt */
+    full?: boolean;
+    /**
+     * Absolute path to the Next.js `app/` directory for automatic route discovery.
+     * When set, routes for llms.txt are auto-discovered by scanning page.tsx files.
+     * Titles and descriptions are extracted from `export const metadata` in each page.
+     *
+     * @example
+     * ```ts
+     * appDir: path.resolve(process.cwd(), 'app')
+     * ```
+     */
+    appDir?: string;
+    /** Options for route discovery when using appDir */
+    discoverOptions?: DiscoverOptions;
+}
+declare function createLlmsTxtHandler(options: LlmsTxtHandlerOptions): () => Promise<Response>;
+
+export { type LlmsTxtHandlerOptions, type WithAgentSeoOptions, createLlmsTxtHandler, withAgentSeo };

package/dist/index.d.ts

@@ -0,0 +1,59 @@
+import { NextConfig } from 'next';
+import { AgentSeoOptions, DiscoverOptions } from '@agent-seo/core';
+export { AIRequestContext, AgentSeoOptions, BotInfo, BotPurpose, DiscoverOptions, LlmsTxtRoute, TransformResult, discoverNextRoutes, generateLlmsTxt, transform } from '@agent-seo/core';
+export { AgentSeoMiddlewareOptions, createAgentSeoMiddleware } from './middleware.js';
+import 'next/server';
+
+interface WithAgentSeoOptions extends AgentSeoOptions {
+    /**
+     * Absolute path to the Next.js `app/` directory.
+     * If omitted, auto-detected by checking `./app` and `./src/app` from cwd.
+     */
+    appDir?: string;
+    /**
+     * Sitemap URL for the `robots.txt` Sitemap directive.
+     * - `true`: uses `{baseUrl}/sitemap.xml`
+     * - `string`: uses the provided URL
+     * - `undefined`/`false`: no Sitemap directive
+     */
+    sitemap?: boolean | string;
+}
+/**
+ * Next.js config plugin that:
+ * 1. Auto-generates `app/llms.txt/route.ts` (zero-config `/llms.txt` endpoint)
+ * 2. Injects `Vary: Accept, User-Agent` headers on all pages
+ *
+ * @example
+ * ```ts
+ * // next.config.ts
+ * import { withAgentSeo } from '@agent-seo/next';
+ *
+ * export default withAgentSeo({
+ *   siteName: 'My App',
+ *   siteDescription: 'A brief description for LLMs.',
+ *   baseUrl: 'https://myapp.com',
+ * })({});
+ * ```
+ */
+declare function withAgentSeo(agentSeoOptions: WithAgentSeoOptions): (nextConfig?: NextConfig) => NextConfig;
+
+interface LlmsTxtHandlerOptions extends AgentSeoOptions {
+    /** Return llms-full.txt instead of llms.txt */
+    full?: boolean;
+    /**
+     * Absolute path to the Next.js `app/` directory for automatic route discovery.
+     * When set, routes for llms.txt are auto-discovered by scanning page.tsx files.
+     * Titles and descriptions are extracted from `export const metadata` in each page.
+     *
+     * @example
+     * ```ts
+     * appDir: path.resolve(process.cwd(), 'app')
+     * ```
+     */
+    appDir?: string;
+    /** Options for route discovery when using appDir */
+    discoverOptions?: DiscoverOptions;
+}
+declare function createLlmsTxtHandler(options: LlmsTxtHandlerOptions): () => Promise<Response>;
+
+export { type LlmsTxtHandlerOptions, type WithAgentSeoOptions, createLlmsTxtHandler, withAgentSeo };

package/dist/index.js

@@ -0,0 +1,312 @@
+var __require = /* @__PURE__ */ ((x) => typeof require !== "undefined" ? require : typeof Proxy !== "undefined" ? new Proxy(x, {
+  get: (a, b) => (typeof require !== "undefined" ? require : a)[b]
+}) : x)(function(x) {
+  if (typeof require !== "undefined") return require.apply(this, arguments);
+  throw Error('Dynamic require of "' + x + '" is not supported');
+});
+
+// src/plugin.ts
+import { existsSync, mkdirSync, writeFileSync } from "fs";
+import { join } from "path";
+function withAgentSeo(agentSeoOptions) {
+  const appDir = agentSeoOptions.appDir || detectAppDir();
+  if (appDir) {
+    generateLlmsTxtRoute(appDir, agentSeoOptions);
+    generateTransformRoute(appDir, agentSeoOptions);
+    generateRobotsTxt(appDir, agentSeoOptions);
+  }
+  return (nextConfig = {}) => {
+    return {
+      ...nextConfig,
+      async headers() {
+        const existingHeaders = await (nextConfig.headers?.() ?? []);
+        return [
+          ...existingHeaders,
+          {
+            source: "/((?!api|_next|static|favicon.ico).*)",
+            headers: [{ key: "Vary", value: "Accept, User-Agent" }]
+          }
+        ];
+      }
+    };
+  };
+}
+function detectAppDir() {
+  const cwd = process.cwd();
+  const candidates = [join(cwd, "app"), join(cwd, "src", "app")];
+  for (const candidate of candidates) {
+    if (existsSync(candidate)) {
+      return candidate;
+    }
+  }
+  return null;
+}
+var GENERATED_BANNER = `// AUTO-GENERATED by @agent-seo/next \u2014 do not edit manually.
+// This file is created by withAgentSeo() in next.config.ts.
+// Add "app/llms.txt" to your .gitignore.
+`;
+function generateLlmsTxtRoute(appDir, options) {
+  const routeDir = join(appDir, "llms.txt");
+  const routeFile = join(routeDir, "route.js");
+  const siteName = escapeStr(options.siteName);
+  const siteDescription = escapeStr(options.siteDescription);
+  const baseUrl = escapeStr(options.baseUrl);
+  const excludePatterns = JSON.stringify(options.exclude || ["/api"]);
+  const content = `${GENERATED_BANNER}
+import { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/next';
+import path from 'node:path';
+
+const appDir = path.resolve(process.cwd(), 'app');
+
+export async function GET(request) {
+  const url = new URL(request.url);
+
+  const routes = discoverNextRoutes(appDir, {
+    exclude: ${excludePatterns},
+  });
+
+  const result = generateLlmsTxt(
+    {
+      siteName: '${siteName}',
+      siteDescription: '${siteDescription}',
+      baseUrl: process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}',
+    },
+    routes,
+  );
+
+  return new Response(result.llmsTxt, {
+    status: 200,
+    headers: {
+      'Content-Type': 'text/plain; charset=utf-8',
+      'Content-Disposition': 'inline',
+      'Cache-Control': 'public, max-age=3600, s-maxage=3600',
+    },
+  });
+}
+`;
+  mkdirSync(routeDir, { recursive: true });
+  let existingContent = "";
+  try {
+    const { readFileSync } = __require("fs");
+    existingContent = readFileSync(routeFile, "utf-8");
+  } catch {
+  }
+  if (existingContent !== content) {
+    writeFileSync(routeFile, content, "utf-8");
+  }
+}
+function escapeStr(s) {
+  return s.replace(/\\/g, "\\\\").replace(/'/g, "\\'");
+}
+function generateTransformRoute(appDir, options) {
+  const routeDir = join(appDir, "api", "agent-seo-transform");
+  const routeFile = join(routeDir, "route.js");
+  const baseUrl = escapeStr(options.baseUrl);
+  const content = `${GENERATED_BANNER}
+import { transform } from '@agent-seo/next';
+
+// Force Node.js runtime (required for jsdom / Readability / Turndown)
+export const runtime = 'nodejs';
+export const dynamic = 'force-dynamic';
+
+export async function GET(request) {
+  const { searchParams } = new URL(request.url);
+  const pagePath = searchParams.get('path') || '/';
+
+  // Build the internal URL to fetch the original HTML page
+  const origin = process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}';
+  if (!origin) {
+    return new Response('Base URL not configured', { status: 500 });
+  }
+
+  // Only allow same-origin, absolute paths (prevent SSRF via absolute or protocol-relative URLs)
+  const isValidPath =
+    pagePath.startsWith('/') &&
+    !pagePath.startsWith('//') &&
+    !pagePath.includes('://') &&
+    !pagePath.includes('\\\\');
+  if (!isValidPath) {
+    return new Response('Invalid path', { status: 400 });
+  }
+
+  const originUrl = new URL(origin);
+  if (originUrl.protocol !== 'http:' && originUrl.protocol !== 'https:') {
+    return new Response('Invalid base URL', { status: 500 });
+  }
+  const pageUrl = new URL(pagePath, originUrl.origin);
+  if (pageUrl.origin !== originUrl.origin) {
+    return new Response('Invalid origin', { status: 400 });
+  }
+
+  try {
+    // Fetch the page HTML from the local server with a normal User-Agent
+    // to avoid infinite rewrite loops
+    const res = await fetch(pageUrl.toString(), {
+      headers: {
+        'User-Agent': 'AgentSEO-Internal/1.0',
+        'Accept': 'text/html',
+      },
+    });
+
+    if (!res.ok) {
+      return new Response('Page not found', { status: 404 });
+    }
+
+    const html = await res.text();
+
+    const result = await transform(html, {
+      url: pageUrl.toString(),
+    });
+
+    return new Response(result.markdown, {
+      status: 200,
+      headers: {
+        'Content-Type': 'text/markdown; charset=utf-8',
+        'Content-Disposition': 'inline',
+        'Cache-Control': 'public, max-age=3600, s-maxage=3600',
+        'Vary': 'Accept, User-Agent',
+        'X-Robots-Tag': 'all',
+        'X-Agent-Seo': 'transformed',
+      },
+    });
+  } catch (err) {
+    console.error('[agent-seo] Transform error:', err);
+    return new Response('Transform failed', { status: 500 });
+  }
+}
+`;
+  mkdirSync(routeDir, { recursive: true });
+  let existingContent = "";
+  try {
+    const { readFileSync } = __require("fs");
+    existingContent = readFileSync(routeFile, "utf-8");
+  } catch {
+  }
+  if (existingContent !== content) {
+    writeFileSync(routeFile, content, "utf-8");
+  }
+}
+function generateRobotsTxt(appDir, options) {
+  const robotsFile = join(appDir, "robots.txt");
+  if (existsSync(robotsFile)) {
+    return;
+  }
+  let content = `# AI-Optimized robots.txt \u2014 Generated by @agent-seo/next
+# For AI agents: see /llms.txt for a structured site manifest
+
+User-agent: *
+Allow: /
+`;
+  if (options.sitemap) {
+    const sitemapUrl = typeof options.sitemap === "string" ? options.sitemap : `${options.baseUrl}/sitemap.xml`;
+    content += `
+Sitemap: ${sitemapUrl}
+`;
+  }
+  writeFileSync(robotsFile, content, "utf-8");
+}
+
+// src/middleware.ts
+import { detectAgent } from "@agent-seo/core/edge";
+import { NextResponse } from "next/server";
+var ALWAYS_SKIP = /* @__PURE__ */ new Set([
+  "/favicon.ico",
+  "/robots.txt",
+  "/sitemap.xml",
+  "/llms.txt",
+  "/llms-full.txt"
+]);
+var DEFAULT_EXCLUDE = [
+  "/api/**",
+  "/_next/**"
+];
+function createAgentSeoMiddleware(options = {}) {
+  const excludePatterns = [...DEFAULT_EXCLUDE, ...options.exclude || []];
+  return function middleware(request) {
+    const { pathname } = request.nextUrl;
+    if (ALWAYS_SKIP.has(pathname)) {
+      return NextResponse.next();
+    }
+    if (isExcluded(pathname, excludePatterns)) {
+      return NextResponse.next();
+    }
+    const ua = request.headers.get("user-agent");
+    const accept = request.headers.get("accept");
+    const aiCtx = detectAgent(ua, accept);
+    if (pathname.endsWith(".md")) {
+      const originalPath = pathname.slice(0, -3) || "/";
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", originalPath);
+      return setBotHeaders(NextResponse.rewrite(transformUrl));
+    }
+    if (aiCtx.isAIBot) {
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", pathname);
+      return setBotHeaders(NextResponse.rewrite(transformUrl));
+    }
+    const response = NextResponse.next();
+    response.headers.set("Vary", "Accept, User-Agent");
+    return response;
+  };
+}
+function setBotHeaders(response) {
+  response.headers.set("Content-Disposition", "inline");
+  response.headers.set("Vary", "Accept, User-Agent");
+  response.headers.set("X-Robots-Tag", "all");
+  response.headers.delete("x-nextjs-matched-path");
+  return response;
+}
+function isExcluded(path, patterns) {
+  return patterns.some((pattern) => matchGlob(pattern, path));
+}
+function matchGlob(pattern, path) {
+  const regex = pattern.replace(/\*\*/g, "{{DOUBLESTAR}}").replace(/\*/g, "[^/]*").replace(/{{DOUBLESTAR}}/g, ".*");
+  return new RegExp(`^${regex}$`).test(path);
+}
+
+// src/route-handler.ts
+import { generateLlmsTxt, discoverNextRoutes } from "@agent-seo/core";
+function createLlmsTxtHandler(options) {
+  return async function GET() {
+    let routes = options.llmsTxt?.routes || [];
+    if (routes.length === 0 && options.appDir) {
+      routes = discoverNextRoutes(options.appDir, {
+        exclude: options.exclude || ["/api"],
+        ...options.discoverOptions
+      });
+    }
+    const result = generateLlmsTxt(
+      {
+        siteName: options.siteName,
+        siteDescription: options.siteDescription,
+        baseUrl: options.baseUrl,
+        ...options.llmsTxt
+      },
+      routes
+    );
+    const content = options.full ? result.llmsFullTxt : result.llmsTxt;
+    return new Response(content, {
+      status: 200,
+      headers: {
+        "Content-Type": "text/plain; charset=utf-8",
+        "Cache-Control": "public, max-age=3600, s-maxage=3600"
+      }
+    });
+  };
+}
+
+// src/index.ts
+import {
+  generateLlmsTxt as generateLlmsTxt2,
+  discoverNextRoutes as discoverNextRoutes2,
+  transform
+} from "@agent-seo/core";
+export {
+  createAgentSeoMiddleware,
+  createLlmsTxtHandler,
+  discoverNextRoutes2 as discoverNextRoutes,
+  generateLlmsTxt2 as generateLlmsTxt,
+  transform,
+  withAgentSeo
+};
+//# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/plugin.ts","../src/middleware.ts","../src/route-handler.ts","../src/index.ts"],"sourcesContent":["import { existsSync, mkdirSync, writeFileSync } from 'node:fs';\nimport { join, resolve } from 'node:path';\nimport type { NextConfig } from 'next';\nimport type { AgentSeoOptions } from '@agent-seo/core';\n\nexport interface WithAgentSeoOptions extends AgentSeoOptions {\n  /**\n   * Absolute path to the Next.js `app/` directory.\n   * If omitted, auto-detected by checking `./app` and `./src/app` from cwd.\n   */\n  appDir?: string;\n  /**\n   * Sitemap URL for the `robots.txt` Sitemap directive.\n   * - `true`: uses `{baseUrl}/sitemap.xml`\n   * - `string`: uses the provided URL\n   * - `undefined`/`false`: no Sitemap directive\n   */\n  sitemap?: boolean | string;\n}\n\n/**\n * Next.js config plugin that:\n * 1. Auto-generates `app/llms.txt/route.ts` (zero-config `/llms.txt` endpoint)\n * 2. Injects `Vary: Accept, User-Agent` headers on all pages\n *\n * @example\n * ```ts\n * // next.config.ts\n * import { withAgentSeo } from '@agent-seo/next';\n *\n * export default withAgentSeo({\n *   siteName: 'My App',\n *   siteDescription: 'A brief description for LLMs.',\n *   baseUrl: 'https://myapp.com',\n * })({});\n * ```\n */\nexport function withAgentSeo(agentSeoOptions: WithAgentSeoOptions) {\n  const appDir = agentSeoOptions.appDir || detectAppDir();\n\n  // Auto-generate route handlers at config-evaluation time\n  if (appDir) {\n    generateLlmsTxtRoute(appDir, agentSeoOptions);\n    generateTransformRoute(appDir, agentSeoOptions);\n    generateRobotsTxt(appDir, agentSeoOptions);\n  }\n\n  return (nextConfig: NextConfig = {}): NextConfig => {\n    return {\n      ...nextConfig,\n\n      async headers() {\n        const existingHeaders = await (nextConfig.headers?.() ?? []);\n        return [\n          ...existingHeaders,\n          {\n            source: '/((?!api|_next|static|favicon.ico).*)',\n            headers: [{ key: 'Vary', value: 'Accept, User-Agent' }],\n          },\n        ];\n      },\n    };\n  };\n}\n\n/**\n * Auto-detect the Next.js `app/` directory from common locations.\n */\nfunction detectAppDir(): string | null {\n  const cwd = process.cwd();\n  const candidates = [join(cwd, 'app'), join(cwd, 'src', 'app')];\n\n  for (const candidate of candidates) {\n    if (existsSync(candidate)) {\n      return candidate;\n    }\n  }\n\n  return null;\n}\n\n// Auto-generated file banner\nconst GENERATED_BANNER = `// AUTO-GENERATED by @agent-seo/next — do not edit manually.\n// This file is created by withAgentSeo() in next.config.ts.\n// Add \"app/llms.txt\" to your .gitignore.\n`;\n\n/**\n * Write `app/llms.txt/route.js` with a handler that auto-discovers routes.\n */\nfunction generateLlmsTxtRoute(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const routeDir = join(appDir, 'llms.txt');\n  const routeFile = join(routeDir, 'route.js');\n\n  // Serialize the options we need into the generated file\n  const siteName = escapeStr(options.siteName);\n  const siteDescription = escapeStr(options.siteDescription);\n  const baseUrl = escapeStr(options.baseUrl);\n  const excludePatterns = JSON.stringify(options.exclude || ['/api']);\n\n  const content = `${GENERATED_BANNER}\nimport { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/next';\nimport path from 'node:path';\n\nconst appDir = path.resolve(process.cwd(), 'app');\n\nexport async function GET(request) {\n  const url = new URL(request.url);\n\n  const routes = discoverNextRoutes(appDir, {\n    exclude: ${excludePatterns},\n  });\n\n  const result = generateLlmsTxt(\n    {\n      siteName: '${siteName}',\n      siteDescription: '${siteDescription}',\n      baseUrl: process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}',\n    },\n    routes,\n  );\n\n  return new Response(result.llmsTxt, {\n    status: 200,\n    headers: {\n      'Content-Type': 'text/plain; charset=utf-8',\n      'Content-Disposition': 'inline',\n      'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n    },\n  });\n}\n`;\n\n  // Only write if the file doesn't exist or content has changed\n  mkdirSync(routeDir, { recursive: true });\n\n  let existingContent = '';\n  try {\n    const { readFileSync } = require('node:fs');\n    existingContent = readFileSync(routeFile, 'utf-8');\n  } catch {\n    // File doesn't exist yet\n  }\n\n  if (existingContent !== content) {\n    writeFileSync(routeFile, content, 'utf-8');\n  }\n}\n\nfunction escapeStr(s: string): string {\n  return s.replace(/\\\\/g, '\\\\\\\\').replace(/'/g, \"\\\\'\");\n}\n\n/**\n * Write `app/api/agent-seo-transform/route.js` — a Node.js API route that\n * receives a `?path=/some-page` param, fetches the HTML from the local\n * Next.js server, runs the core `transform()` pipeline (JSDOM → Readability\n * → Turndown), and returns clean Markdown.\n */\nfunction generateTransformRoute(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const routeDir = join(appDir, 'api', 'agent-seo-transform');\n  const routeFile = join(routeDir, 'route.js');\n\n  const baseUrl = escapeStr(options.baseUrl);\n\n  const content = `${GENERATED_BANNER}\nimport { transform } from '@agent-seo/next';\n\n// Force Node.js runtime (required for jsdom / Readability / Turndown)\nexport const runtime = 'nodejs';\nexport const dynamic = 'force-dynamic';\n\nexport async function GET(request) {\n  const { searchParams } = new URL(request.url);\n  const pagePath = searchParams.get('path') || '/';\n\n  // Build the internal URL to fetch the original HTML page\n  const origin = process.env.NEXT_PUBLIC_BASE_URL || '${baseUrl}';\n  if (!origin) {\n    return new Response('Base URL not configured', { status: 500 });\n  }\n\n  // Only allow same-origin, absolute paths (prevent SSRF via absolute or protocol-relative URLs)\n  const isValidPath =\n    pagePath.startsWith('/') &&\n    !pagePath.startsWith('//') &&\n    !pagePath.includes('://') &&\n    !pagePath.includes('\\\\\\\\');\n  if (!isValidPath) {\n    return new Response('Invalid path', { status: 400 });\n  }\n\n  const originUrl = new URL(origin);\n  if (originUrl.protocol !== 'http:' && originUrl.protocol !== 'https:') {\n    return new Response('Invalid base URL', { status: 500 });\n  }\n  const pageUrl = new URL(pagePath, originUrl.origin);\n  if (pageUrl.origin !== originUrl.origin) {\n    return new Response('Invalid origin', { status: 400 });\n  }\n\n  try {\n    // Fetch the page HTML from the local server with a normal User-Agent\n    // to avoid infinite rewrite loops\n    const res = await fetch(pageUrl.toString(), {\n      headers: {\n        'User-Agent': 'AgentSEO-Internal/1.0',\n        'Accept': 'text/html',\n      },\n    });\n\n    if (!res.ok) {\n      return new Response('Page not found', { status: 404 });\n    }\n\n    const html = await res.text();\n\n    const result = await transform(html, {\n      url: pageUrl.toString(),\n    });\n\n    return new Response(result.markdown, {\n      status: 200,\n      headers: {\n        'Content-Type': 'text/markdown; charset=utf-8',\n        'Content-Disposition': 'inline',\n        'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n        'Vary': 'Accept, User-Agent',\n        'X-Robots-Tag': 'all',\n        'X-Agent-Seo': 'transformed',\n      },\n    });\n  } catch (err) {\n    console.error('[agent-seo] Transform error:', err);\n    return new Response('Transform failed', { status: 500 });\n  }\n}\n`;\n\n  mkdirSync(routeDir, { recursive: true });\n\n  let existingContent = '';\n  try {\n    const { readFileSync } = require('node:fs');\n    existingContent = readFileSync(routeFile, 'utf-8');\n  } catch {\n    // File doesn't exist yet\n  }\n\n  if (existingContent !== content) {\n    writeFileSync(routeFile, content, 'utf-8');\n  }\n}\n\n/**\n * Write `app/robots.txt` — a static robots.txt that allows all crawlers\n * and references /llms.txt for AI agents.\n * Only created if the file doesn't already exist (won't overwrite user customizations).\n */\nfunction generateRobotsTxt(\n  appDir: string,\n  options: WithAgentSeoOptions,\n): void {\n  const robotsFile = join(appDir, 'robots.txt');\n\n  // Don't overwrite user-created robots.txt\n  if (existsSync(robotsFile)) {\n    return;\n  }\n\n  let content = `# AI-Optimized robots.txt — Generated by @agent-seo/next\n# For AI agents: see /llms.txt for a structured site manifest\n\nUser-agent: *\nAllow: /\n`;\n\n  if (options.sitemap) {\n    const sitemapUrl =\n      typeof options.sitemap === 'string'\n        ? options.sitemap\n        : `${options.baseUrl}/sitemap.xml`;\n    content += `\\nSitemap: ${sitemapUrl}\\n`;\n  }\n\n  writeFileSync(robotsFile, content, 'utf-8');\n}\n","import { detectAgent } from '@agent-seo/core/edge';\nimport { NextResponse } from 'next/server';\nimport type { NextRequest } from 'next/server';\n\nexport interface AgentSeoMiddlewareOptions {\n  /**\n   * Glob patterns for paths that should NEVER be rewritten to Markdown.\n   * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).\n   *\n   * @example ['\\/dashboard\\/**', '\\/admin\\/**', '\\/api\\/private\\/**']\n   */\n  exclude?: string[];\n}\n\n/** Paths that are always skipped — framework internals + standard files. */\nconst ALWAYS_SKIP = new Set([\n  '/favicon.ico',\n  '/robots.txt',\n  '/sitemap.xml',\n  '/llms.txt',\n  '/llms-full.txt',\n]);\n\n/** Default exclude patterns — users can extend but not remove these. */\nconst DEFAULT_EXCLUDE: string[] = [\n  '/api/**',\n  '/_next/**',\n];\n\n/**\n * Creates a Next.js middleware that:\n * 1. Detects AI bot requests via User-Agent\n * 2. Rewrites AI bot requests to an internal transform API route\n *    that converts HTML → Markdown (runs on Node.js runtime)\n * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)\n * 4. Sets `Vary: Accept, User-Agent` on all responses\n */\nexport function createAgentSeoMiddleware(options: AgentSeoMiddlewareOptions = {}) {\n  const excludePatterns = [...DEFAULT_EXCLUDE, ...(options.exclude || [])];\n\n  return function middleware(request: NextRequest) {\n    const { pathname } = request.nextUrl;\n\n    // Skip standard files (robots.txt, favicon, llms.txt, etc.)\n    if (ALWAYS_SKIP.has(pathname)) {\n      return NextResponse.next();\n    }\n\n    // Skip excluded patterns (API routes, admin, dashboard, etc.)\n    if (isExcluded(pathname, excludePatterns)) {\n      return NextResponse.next();\n    }\n\n    const ua = request.headers.get('user-agent');\n    const accept = request.headers.get('accept');\n    const aiCtx = detectAgent(ua, accept);\n\n    // Handle explicit .md suffix requests (e.g. /about.md)\n    if (pathname.endsWith('.md')) {\n      const originalPath = pathname.slice(0, -3) || '/';\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', originalPath);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // If AI bot, rewrite to transform API\n    if (aiCtx.isAIBot) {\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', pathname);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // Normal request — just add Vary header\n    const response = NextResponse.next();\n    response.headers.set('Vary', 'Accept, User-Agent');\n    return response;\n  };\n}\n\n/**\n * Set clean, bot-friendly headers on a rewrite response.\n * Overrides Next.js RSC-related headers that can confuse AI bots.\n */\nfunction setBotHeaders(response: NextResponse): NextResponse {\n  response.headers.set('Content-Disposition', 'inline');\n  response.headers.set('Vary', 'Accept, User-Agent');\n  response.headers.set('X-Robots-Tag', 'all');\n  response.headers.delete('x-nextjs-matched-path');\n  return response;\n}\n\nfunction isExcluded(path: string, patterns: string[]): boolean {\n  return patterns.some((pattern) => matchGlob(pattern, path));\n}\n\nfunction matchGlob(pattern: string, path: string): boolean {\n  const regex = pattern\n    .replace(/\\*\\*/g, '{{DOUBLESTAR}}')\n    .replace(/\\*/g, '[^/]*')\n    .replace(/{{DOUBLESTAR}}/g, '.*');\n  return new RegExp(`^${regex}$`).test(path);\n}\n","import { generateLlmsTxt, discoverNextRoutes } from '@agent-seo/core';\nimport type { AgentSeoOptions, DiscoverOptions } from '@agent-seo/core';\n\nexport interface LlmsTxtHandlerOptions extends AgentSeoOptions {\n  /** Return llms-full.txt instead of llms.txt */\n  full?: boolean;\n\n  /**\n   * Absolute path to the Next.js `app/` directory for automatic route discovery.\n   * When set, routes for llms.txt are auto-discovered by scanning page.tsx files.\n   * Titles and descriptions are extracted from `export const metadata` in each page.\n   *\n   * @example\n   * ```ts\n   * appDir: path.resolve(process.cwd(), 'app')\n   * ```\n   */\n  appDir?: string;\n\n  /** Options for route discovery when using appDir */\n  discoverOptions?: DiscoverOptions;\n}\n\nexport function createLlmsTxtHandler(options: LlmsTxtHandlerOptions) {\n  return async function GET() {\n    // Auto-discover routes from app/ directory if no explicit routes are provided\n    let routes = options.llmsTxt?.routes || [];\n\n    if (routes.length === 0 && options.appDir) {\n      routes = discoverNextRoutes(options.appDir, {\n        exclude: options.exclude || ['/api'],\n        ...options.discoverOptions,\n      });\n    }\n\n    const result = generateLlmsTxt(\n      {\n        siteName: options.siteName,\n        siteDescription: options.siteDescription,\n        baseUrl: options.baseUrl,\n        ...options.llmsTxt,\n      },\n      routes,\n    );\n\n    const content = options.full ? result.llmsFullTxt : result.llmsTxt;\n\n    return new Response(content, {\n      status: 200,\n      headers: {\n        'Content-Type': 'text/plain; charset=utf-8',\n        'Cache-Control': 'public, max-age=3600, s-maxage=3600',\n      },\n    });\n  };\n}\n","export { withAgentSeo } from './plugin.js';\nexport type { WithAgentSeoOptions } from './plugin.js';\nexport { createAgentSeoMiddleware } from './middleware.js';\nexport type { AgentSeoMiddlewareOptions } from './middleware.js';\nexport { createLlmsTxtHandler } from './route-handler.js';\nexport type { LlmsTxtHandlerOptions } from './route-handler.js';\nexport {\n  generateLlmsTxt,\n  discoverNextRoutes,\n  transform,\n} from '@agent-seo/core';\nexport type {\n  AgentSeoOptions,\n  AIRequestContext,\n  TransformResult,\n  BotInfo,\n  BotPurpose,\n  LlmsTxtRoute,\n  DiscoverOptions,\n} from '@agent-seo/core';\n"],"mappings":";;;;;;;;AAAA,SAAS,YAAY,WAAW,qBAAqB;AACrD,SAAS,YAAqB;AAoCvB,SAAS,aAAa,iBAAsC;AACjE,QAAM,SAAS,gBAAgB,UAAU,aAAa;AAGtD,MAAI,QAAQ;AACV,yBAAqB,QAAQ,eAAe;AAC5C,2BAAuB,QAAQ,eAAe;AAC9C,sBAAkB,QAAQ,eAAe;AAAA,EAC3C;AAEA,SAAO,CAAC,aAAyB,CAAC,MAAkB;AAClD,WAAO;AAAA,MACL,GAAG;AAAA,MAEH,MAAM,UAAU;AACd,cAAM,kBAAkB,OAAO,WAAW,UAAU,KAAK,CAAC;AAC1D,eAAO;AAAA,UACL,GAAG;AAAA,UACH;AAAA,YACE,QAAQ;AAAA,YACR,SAAS,CAAC,EAAE,KAAK,QAAQ,OAAO,qBAAqB,CAAC;AAAA,UACxD;AAAA,QACF;AAAA,MACF;AAAA,IACF;AAAA,EACF;AACF;AAKA,SAAS,eAA8B;AACrC,QAAM,MAAM,QAAQ,IAAI;AACxB,QAAM,aAAa,CAAC,KAAK,KAAK,KAAK,GAAG,KAAK,KAAK,OAAO,KAAK,CAAC;AAE7D,aAAW,aAAa,YAAY;AAClC,QAAI,WAAW,SAAS,GAAG;AACzB,aAAO;AAAA,IACT;AAAA,EACF;AAEA,SAAO;AACT;AAGA,IAAM,mBAAmB;AAAA;AAAA;AAAA;AAQzB,SAAS,qBACP,QACA,SACM;AACN,QAAM,WAAW,KAAK,QAAQ,UAAU;AACxC,QAAM,YAAY,KAAK,UAAU,UAAU;AAG3C,QAAM,WAAW,UAAU,QAAQ,QAAQ;AAC3C,QAAM,kBAAkB,UAAU,QAAQ,eAAe;AACzD,QAAM,UAAU,UAAU,QAAQ,OAAO;AACzC,QAAM,kBAAkB,KAAK,UAAU,QAAQ,WAAW,CAAC,MAAM,CAAC;AAElE,QAAM,UAAU,GAAG,gBAAgB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,eAUtB,eAAe;AAAA;AAAA;AAAA;AAAA;AAAA,mBAKX,QAAQ;AAAA,0BACD,eAAe;AAAA,sDACa,OAAO;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAiB3D,YAAU,UAAU,EAAE,WAAW,KAAK,CAAC;AAEvC,MAAI,kBAAkB;AACtB,MAAI;AACF,UAAM,EAAE,aAAa,IAAI,UAAQ,IAAS;AAC1C,sBAAkB,aAAa,WAAW,OAAO;AAAA,EACnD,QAAQ;AAAA,EAER;AAEA,MAAI,oBAAoB,SAAS;AAC/B,kBAAc,WAAW,SAAS,OAAO;AAAA,EAC3C;AACF;AAEA,SAAS,UAAU,GAAmB;AACpC,SAAO,EAAE,QAAQ,OAAO,MAAM,EAAE,QAAQ,MAAM,KAAK;AACrD;AAQA,SAAS,uBACP,QACA,SACM;AACN,QAAM,WAAW,KAAK,QAAQ,OAAO,qBAAqB;AAC1D,QAAM,YAAY,KAAK,UAAU,UAAU;AAE3C,QAAM,UAAU,UAAU,QAAQ,OAAO;AAEzC,QAAM,UAAU,GAAG,gBAAgB;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,wDAYmB,OAAO;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AA8D7D,YAAU,UAAU,EAAE,WAAW,KAAK,CAAC;AAEvC,MAAI,kBAAkB;AACtB,MAAI;AACF,UAAM,EAAE,aAAa,IAAI,UAAQ,IAAS;AAC1C,sBAAkB,aAAa,WAAW,OAAO;AAAA,EACnD,QAAQ;AAAA,EAER;AAEA,MAAI,oBAAoB,SAAS;AAC/B,kBAAc,WAAW,SAAS,OAAO;AAAA,EAC3C;AACF;AAOA,SAAS,kBACP,QACA,SACM;AACN,QAAM,aAAa,KAAK,QAAQ,YAAY;AAG5C,MAAI,WAAW,UAAU,GAAG;AAC1B;AAAA,EACF;AAEA,MAAI,UAAU;AAAA;AAAA;AAAA;AAAA;AAAA;AAOd,MAAI,QAAQ,SAAS;AACnB,UAAM,aACJ,OAAO,QAAQ,YAAY,WACvB,QAAQ,UACR,GAAG,QAAQ,OAAO;AACxB,eAAW;AAAA,WAAc,UAAU;AAAA;AAAA,EACrC;AAEA,gBAAc,YAAY,SAAS,OAAO;AAC5C;;;ACpSA,SAAS,mBAAmB;AAC5B,SAAS,oBAAoB;AAc7B,IAAM,cAAc,oBAAI,IAAI;AAAA,EAC1B;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF,CAAC;AAGD,IAAM,kBAA4B;AAAA,EAChC;AAAA,EACA;AACF;AAUO,SAAS,yBAAyB,UAAqC,CAAC,GAAG;AAChF,QAAM,kBAAkB,CAAC,GAAG,iBAAiB,GAAI,QAAQ,WAAW,CAAC,CAAE;AAEvE,SAAO,SAAS,WAAW,SAAsB;AAC/C,UAAM,EAAE,SAAS,IAAI,QAAQ;AAG7B,QAAI,YAAY,IAAI,QAAQ,GAAG;AAC7B,aAAO,aAAa,KAAK;AAAA,IAC3B;AAGA,QAAI,WAAW,UAAU,eAAe,GAAG;AACzC,aAAO,aAAa,KAAK;AAAA,IAC3B;AAEA,UAAM,KAAK,QAAQ,QAAQ,IAAI,YAAY;AAC3C,UAAM,SAAS,QAAQ,QAAQ,IAAI,QAAQ;AAC3C,UAAM,QAAQ,YAAY,IAAI,MAAM;AAGpC,QAAI,SAAS,SAAS,KAAK,GAAG;AAC5B,YAAM,eAAe,SAAS,MAAM,GAAG,EAAE,KAAK;AAC9C,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,YAAY;AAClD,aAAO,cAAc,aAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,QAAI,MAAM,SAAS;AACjB,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,QAAQ;AAC9C,aAAO,cAAc,aAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,UAAM,WAAW,aAAa,KAAK;AACnC,aAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAO;AAAA,EACT;AACF;AAMA,SAAS,cAAc,UAAsC;AAC3D,WAAS,QAAQ,IAAI,uBAAuB,QAAQ;AACpD,WAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAS,QAAQ,IAAI,gBAAgB,KAAK;AAC1C,WAAS,QAAQ,OAAO,uBAAuB;AAC/C,SAAO;AACT;AAEA,SAAS,WAAW,MAAc,UAA6B;AAC7D,SAAO,SAAS,KAAK,CAAC,YAAY,UAAU,SAAS,IAAI,CAAC;AAC5D;AAEA,SAAS,UAAU,SAAiB,MAAuB;AACzD,QAAM,QAAQ,QACX,QAAQ,SAAS,gBAAgB,EACjC,QAAQ,OAAO,OAAO,EACtB,QAAQ,mBAAmB,IAAI;AAClC,SAAO,IAAI,OAAO,IAAI,KAAK,GAAG,EAAE,KAAK,IAAI;AAC3C;;;ACrGA,SAAS,iBAAiB,0BAA0B;AAuB7C,SAAS,qBAAqB,SAAgC;AACnE,SAAO,eAAe,MAAM;AAE1B,QAAI,SAAS,QAAQ,SAAS,UAAU,CAAC;AAEzC,QAAI,OAAO,WAAW,KAAK,QAAQ,QAAQ;AACzC,eAAS,mBAAmB,QAAQ,QAAQ;AAAA,QAC1C,SAAS,QAAQ,WAAW,CAAC,MAAM;AAAA,QACnC,GAAG,QAAQ;AAAA,MACb,CAAC;AAAA,IACH;AAEA,UAAM,SAAS;AAAA,MACb;AAAA,QACE,UAAU,QAAQ;AAAA,QAClB,iBAAiB,QAAQ;AAAA,QACzB,SAAS,QAAQ;AAAA,QACjB,GAAG,QAAQ;AAAA,MACb;AAAA,MACA;AAAA,IACF;AAEA,UAAM,UAAU,QAAQ,OAAO,OAAO,cAAc,OAAO;AAE3D,WAAO,IAAI,SAAS,SAAS;AAAA,MAC3B,QAAQ;AAAA,MACR,SAAS;AAAA,QACP,gBAAgB;AAAA,QAChB,iBAAiB;AAAA,MACnB;AAAA,IACF,CAAC;AAAA,EACH;AACF;;;ACjDA;AAAA,EACE,mBAAAA;AAAA,EACA,sBAAAC;AAAA,EACA;AAAA,OACK;","names":["generateLlmsTxt","discoverNextRoutes"]}

package/dist/middleware.cjs

@@ -0,0 +1,86 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+
+// src/middleware.ts
+var middleware_exports = {};
+__export(middleware_exports, {
+  createAgentSeoMiddleware: () => createAgentSeoMiddleware
+});
+module.exports = __toCommonJS(middleware_exports);
+var import_edge = require("@agent-seo/core/edge");
+var import_server = require("next/server");
+var ALWAYS_SKIP = /* @__PURE__ */ new Set([
+  "/favicon.ico",
+  "/robots.txt",
+  "/sitemap.xml",
+  "/llms.txt",
+  "/llms-full.txt"
+]);
+var DEFAULT_EXCLUDE = [
+  "/api/**",
+  "/_next/**"
+];
+function createAgentSeoMiddleware(options = {}) {
+  const excludePatterns = [...DEFAULT_EXCLUDE, ...options.exclude || []];
+  return function middleware(request) {
+    const { pathname } = request.nextUrl;
+    if (ALWAYS_SKIP.has(pathname)) {
+      return import_server.NextResponse.next();
+    }
+    if (isExcluded(pathname, excludePatterns)) {
+      return import_server.NextResponse.next();
+    }
+    const ua = request.headers.get("user-agent");
+    const accept = request.headers.get("accept");
+    const aiCtx = (0, import_edge.detectAgent)(ua, accept);
+    if (pathname.endsWith(".md")) {
+      const originalPath = pathname.slice(0, -3) || "/";
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", originalPath);
+      return setBotHeaders(import_server.NextResponse.rewrite(transformUrl));
+    }
+    if (aiCtx.isAIBot) {
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", pathname);
+      return setBotHeaders(import_server.NextResponse.rewrite(transformUrl));
+    }
+    const response = import_server.NextResponse.next();
+    response.headers.set("Vary", "Accept, User-Agent");
+    return response;
+  };
+}
+function setBotHeaders(response) {
+  response.headers.set("Content-Disposition", "inline");
+  response.headers.set("Vary", "Accept, User-Agent");
+  response.headers.set("X-Robots-Tag", "all");
+  response.headers.delete("x-nextjs-matched-path");
+  return response;
+}
+function isExcluded(path, patterns) {
+  return patterns.some((pattern) => matchGlob(pattern, path));
+}
+function matchGlob(pattern, path) {
+  const regex = pattern.replace(/\*\*/g, "{{DOUBLESTAR}}").replace(/\*/g, "[^/]*").replace(/{{DOUBLESTAR}}/g, ".*");
+  return new RegExp(`^${regex}$`).test(path);
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  createAgentSeoMiddleware
+});
+//# sourceMappingURL=middleware.cjs.map

package/dist/middleware.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/middleware.ts"],"sourcesContent":["import { detectAgent } from '@agent-seo/core/edge';\nimport { NextResponse } from 'next/server';\nimport type { NextRequest } from 'next/server';\n\nexport interface AgentSeoMiddlewareOptions {\n  /**\n   * Glob patterns for paths that should NEVER be rewritten to Markdown.\n   * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).\n   *\n   * @example ['\\/dashboard\\/**', '\\/admin\\/**', '\\/api\\/private\\/**']\n   */\n  exclude?: string[];\n}\n\n/** Paths that are always skipped — framework internals + standard files. */\nconst ALWAYS_SKIP = new Set([\n  '/favicon.ico',\n  '/robots.txt',\n  '/sitemap.xml',\n  '/llms.txt',\n  '/llms-full.txt',\n]);\n\n/** Default exclude patterns — users can extend but not remove these. */\nconst DEFAULT_EXCLUDE: string[] = [\n  '/api/**',\n  '/_next/**',\n];\n\n/**\n * Creates a Next.js middleware that:\n * 1. Detects AI bot requests via User-Agent\n * 2. Rewrites AI bot requests to an internal transform API route\n *    that converts HTML → Markdown (runs on Node.js runtime)\n * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)\n * 4. Sets `Vary: Accept, User-Agent` on all responses\n */\nexport function createAgentSeoMiddleware(options: AgentSeoMiddlewareOptions = {}) {\n  const excludePatterns = [...DEFAULT_EXCLUDE, ...(options.exclude || [])];\n\n  return function middleware(request: NextRequest) {\n    const { pathname } = request.nextUrl;\n\n    // Skip standard files (robots.txt, favicon, llms.txt, etc.)\n    if (ALWAYS_SKIP.has(pathname)) {\n      return NextResponse.next();\n    }\n\n    // Skip excluded patterns (API routes, admin, dashboard, etc.)\n    if (isExcluded(pathname, excludePatterns)) {\n      return NextResponse.next();\n    }\n\n    const ua = request.headers.get('user-agent');\n    const accept = request.headers.get('accept');\n    const aiCtx = detectAgent(ua, accept);\n\n    // Handle explicit .md suffix requests (e.g. /about.md)\n    if (pathname.endsWith('.md')) {\n      const originalPath = pathname.slice(0, -3) || '/';\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', originalPath);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // If AI bot, rewrite to transform API\n    if (aiCtx.isAIBot) {\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', pathname);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // Normal request — just add Vary header\n    const response = NextResponse.next();\n    response.headers.set('Vary', 'Accept, User-Agent');\n    return response;\n  };\n}\n\n/**\n * Set clean, bot-friendly headers on a rewrite response.\n * Overrides Next.js RSC-related headers that can confuse AI bots.\n */\nfunction setBotHeaders(response: NextResponse): NextResponse {\n  response.headers.set('Content-Disposition', 'inline');\n  response.headers.set('Vary', 'Accept, User-Agent');\n  response.headers.set('X-Robots-Tag', 'all');\n  response.headers.delete('x-nextjs-matched-path');\n  return response;\n}\n\nfunction isExcluded(path: string, patterns: string[]): boolean {\n  return patterns.some((pattern) => matchGlob(pattern, path));\n}\n\nfunction matchGlob(pattern: string, path: string): boolean {\n  const regex = pattern\n    .replace(/\\*\\*/g, '{{DOUBLESTAR}}')\n    .replace(/\\*/g, '[^/]*')\n    .replace(/{{DOUBLESTAR}}/g, '.*');\n  return new RegExp(`^${regex}$`).test(path);\n}\n"],"mappings":";;;;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,kBAA4B;AAC5B,oBAA6B;AAc7B,IAAM,cAAc,oBAAI,IAAI;AAAA,EAC1B;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF,CAAC;AAGD,IAAM,kBAA4B;AAAA,EAChC;AAAA,EACA;AACF;AAUO,SAAS,yBAAyB,UAAqC,CAAC,GAAG;AAChF,QAAM,kBAAkB,CAAC,GAAG,iBAAiB,GAAI,QAAQ,WAAW,CAAC,CAAE;AAEvE,SAAO,SAAS,WAAW,SAAsB;AAC/C,UAAM,EAAE,SAAS,IAAI,QAAQ;AAG7B,QAAI,YAAY,IAAI,QAAQ,GAAG;AAC7B,aAAO,2BAAa,KAAK;AAAA,IAC3B;AAGA,QAAI,WAAW,UAAU,eAAe,GAAG;AACzC,aAAO,2BAAa,KAAK;AAAA,IAC3B;AAEA,UAAM,KAAK,QAAQ,QAAQ,IAAI,YAAY;AAC3C,UAAM,SAAS,QAAQ,QAAQ,IAAI,QAAQ;AAC3C,UAAM,YAAQ,yBAAY,IAAI,MAAM;AAGpC,QAAI,SAAS,SAAS,KAAK,GAAG;AAC5B,YAAM,eAAe,SAAS,MAAM,GAAG,EAAE,KAAK;AAC9C,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,YAAY;AAClD,aAAO,cAAc,2BAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,QAAI,MAAM,SAAS;AACjB,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,QAAQ;AAC9C,aAAO,cAAc,2BAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,UAAM,WAAW,2BAAa,KAAK;AACnC,aAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAO;AAAA,EACT;AACF;AAMA,SAAS,cAAc,UAAsC;AAC3D,WAAS,QAAQ,IAAI,uBAAuB,QAAQ;AACpD,WAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAS,QAAQ,IAAI,gBAAgB,KAAK;AAC1C,WAAS,QAAQ,OAAO,uBAAuB;AAC/C,SAAO;AACT;AAEA,SAAS,WAAW,MAAc,UAA6B;AAC7D,SAAO,SAAS,KAAK,CAAC,YAAY,UAAU,SAAS,IAAI,CAAC;AAC5D;AAEA,SAAS,UAAU,SAAiB,MAAuB;AACzD,QAAM,QAAQ,QACX,QAAQ,SAAS,gBAAgB,EACjC,QAAQ,OAAO,OAAO,EACtB,QAAQ,mBAAmB,IAAI;AAClC,SAAO,IAAI,OAAO,IAAI,KAAK,GAAG,EAAE,KAAK,IAAI;AAC3C;","names":[]}

package/dist/middleware.d.cts

@@ -0,0 +1,22 @@
+import { NextRequest, NextResponse } from 'next/server';
+
+interface AgentSeoMiddlewareOptions {
+    /**
+     * Glob patterns for paths that should NEVER be rewritten to Markdown.
+     * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).
+     *
+     * @example ['\/dashboard\/**', '\/admin\/**', '\/api\/private\/**']
+     */
+    exclude?: string[];
+}
+/**
+ * Creates a Next.js middleware that:
+ * 1. Detects AI bot requests via User-Agent
+ * 2. Rewrites AI bot requests to an internal transform API route
+ *    that converts HTML → Markdown (runs on Node.js runtime)
+ * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)
+ * 4. Sets `Vary: Accept, User-Agent` on all responses
+ */
+declare function createAgentSeoMiddleware(options?: AgentSeoMiddlewareOptions): (request: NextRequest) => NextResponse<unknown>;
+
+export { type AgentSeoMiddlewareOptions, createAgentSeoMiddleware };

package/dist/middleware.d.ts

@@ -0,0 +1,22 @@
+import { NextRequest, NextResponse } from 'next/server';
+
+interface AgentSeoMiddlewareOptions {
+    /**
+     * Glob patterns for paths that should NEVER be rewritten to Markdown.
+     * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).
+     *
+     * @example ['\/dashboard\/**', '\/admin\/**', '\/api\/private\/**']
+     */
+    exclude?: string[];
+}
+/**
+ * Creates a Next.js middleware that:
+ * 1. Detects AI bot requests via User-Agent
+ * 2. Rewrites AI bot requests to an internal transform API route
+ *    that converts HTML → Markdown (runs on Node.js runtime)
+ * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)
+ * 4. Sets `Vary: Accept, User-Agent` on all responses
+ */
+declare function createAgentSeoMiddleware(options?: AgentSeoMiddlewareOptions): (request: NextRequest) => NextResponse<unknown>;
+
+export { type AgentSeoMiddlewareOptions, createAgentSeoMiddleware };

package/dist/middleware.js

@@ -0,0 +1,61 @@
+// src/middleware.ts
+import { detectAgent } from "@agent-seo/core/edge";
+import { NextResponse } from "next/server";
+var ALWAYS_SKIP = /* @__PURE__ */ new Set([
+  "/favicon.ico",
+  "/robots.txt",
+  "/sitemap.xml",
+  "/llms.txt",
+  "/llms-full.txt"
+]);
+var DEFAULT_EXCLUDE = [
+  "/api/**",
+  "/_next/**"
+];
+function createAgentSeoMiddleware(options = {}) {
+  const excludePatterns = [...DEFAULT_EXCLUDE, ...options.exclude || []];
+  return function middleware(request) {
+    const { pathname } = request.nextUrl;
+    if (ALWAYS_SKIP.has(pathname)) {
+      return NextResponse.next();
+    }
+    if (isExcluded(pathname, excludePatterns)) {
+      return NextResponse.next();
+    }
+    const ua = request.headers.get("user-agent");
+    const accept = request.headers.get("accept");
+    const aiCtx = detectAgent(ua, accept);
+    if (pathname.endsWith(".md")) {
+      const originalPath = pathname.slice(0, -3) || "/";
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", originalPath);
+      return setBotHeaders(NextResponse.rewrite(transformUrl));
+    }
+    if (aiCtx.isAIBot) {
+      const transformUrl = new URL("/api/agent-seo-transform", request.url);
+      transformUrl.searchParams.set("path", pathname);
+      return setBotHeaders(NextResponse.rewrite(transformUrl));
+    }
+    const response = NextResponse.next();
+    response.headers.set("Vary", "Accept, User-Agent");
+    return response;
+  };
+}
+function setBotHeaders(response) {
+  response.headers.set("Content-Disposition", "inline");
+  response.headers.set("Vary", "Accept, User-Agent");
+  response.headers.set("X-Robots-Tag", "all");
+  response.headers.delete("x-nextjs-matched-path");
+  return response;
+}
+function isExcluded(path, patterns) {
+  return patterns.some((pattern) => matchGlob(pattern, path));
+}
+function matchGlob(pattern, path) {
+  const regex = pattern.replace(/\*\*/g, "{{DOUBLESTAR}}").replace(/\*/g, "[^/]*").replace(/{{DOUBLESTAR}}/g, ".*");
+  return new RegExp(`^${regex}$`).test(path);
+}
+export {
+  createAgentSeoMiddleware
+};
+//# sourceMappingURL=middleware.js.map

package/dist/middleware.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/middleware.ts"],"sourcesContent":["import { detectAgent } from '@agent-seo/core/edge';\nimport { NextResponse } from 'next/server';\nimport type { NextRequest } from 'next/server';\n\nexport interface AgentSeoMiddlewareOptions {\n  /**\n   * Glob patterns for paths that should NEVER be rewritten to Markdown.\n   * Merged with built-in defaults (see `DEFAULT_EXCLUDE`).\n   *\n   * @example ['\\/dashboard\\/**', '\\/admin\\/**', '\\/api\\/private\\/**']\n   */\n  exclude?: string[];\n}\n\n/** Paths that are always skipped — framework internals + standard files. */\nconst ALWAYS_SKIP = new Set([\n  '/favicon.ico',\n  '/robots.txt',\n  '/sitemap.xml',\n  '/llms.txt',\n  '/llms-full.txt',\n]);\n\n/** Default exclude patterns — users can extend but not remove these. */\nconst DEFAULT_EXCLUDE: string[] = [\n  '/api/**',\n  '/_next/**',\n];\n\n/**\n * Creates a Next.js middleware that:\n * 1. Detects AI bot requests via User-Agent\n * 2. Rewrites AI bot requests to an internal transform API route\n *    that converts HTML → Markdown (runs on Node.js runtime)\n * 3. Handles `.md` suffix requests (e.g. `/about.md` → transform `/about`)\n * 4. Sets `Vary: Accept, User-Agent` on all responses\n */\nexport function createAgentSeoMiddleware(options: AgentSeoMiddlewareOptions = {}) {\n  const excludePatterns = [...DEFAULT_EXCLUDE, ...(options.exclude || [])];\n\n  return function middleware(request: NextRequest) {\n    const { pathname } = request.nextUrl;\n\n    // Skip standard files (robots.txt, favicon, llms.txt, etc.)\n    if (ALWAYS_SKIP.has(pathname)) {\n      return NextResponse.next();\n    }\n\n    // Skip excluded patterns (API routes, admin, dashboard, etc.)\n    if (isExcluded(pathname, excludePatterns)) {\n      return NextResponse.next();\n    }\n\n    const ua = request.headers.get('user-agent');\n    const accept = request.headers.get('accept');\n    const aiCtx = detectAgent(ua, accept);\n\n    // Handle explicit .md suffix requests (e.g. /about.md)\n    if (pathname.endsWith('.md')) {\n      const originalPath = pathname.slice(0, -3) || '/';\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', originalPath);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // If AI bot, rewrite to transform API\n    if (aiCtx.isAIBot) {\n      const transformUrl = new URL('/api/agent-seo-transform', request.url);\n      transformUrl.searchParams.set('path', pathname);\n      return setBotHeaders(NextResponse.rewrite(transformUrl));\n    }\n\n    // Normal request — just add Vary header\n    const response = NextResponse.next();\n    response.headers.set('Vary', 'Accept, User-Agent');\n    return response;\n  };\n}\n\n/**\n * Set clean, bot-friendly headers on a rewrite response.\n * Overrides Next.js RSC-related headers that can confuse AI bots.\n */\nfunction setBotHeaders(response: NextResponse): NextResponse {\n  response.headers.set('Content-Disposition', 'inline');\n  response.headers.set('Vary', 'Accept, User-Agent');\n  response.headers.set('X-Robots-Tag', 'all');\n  response.headers.delete('x-nextjs-matched-path');\n  return response;\n}\n\nfunction isExcluded(path: string, patterns: string[]): boolean {\n  return patterns.some((pattern) => matchGlob(pattern, path));\n}\n\nfunction matchGlob(pattern: string, path: string): boolean {\n  const regex = pattern\n    .replace(/\\*\\*/g, '{{DOUBLESTAR}}')\n    .replace(/\\*/g, '[^/]*')\n    .replace(/{{DOUBLESTAR}}/g, '.*');\n  return new RegExp(`^${regex}$`).test(path);\n}\n"],"mappings":";AAAA,SAAS,mBAAmB;AAC5B,SAAS,oBAAoB;AAc7B,IAAM,cAAc,oBAAI,IAAI;AAAA,EAC1B;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AACF,CAAC;AAGD,IAAM,kBAA4B;AAAA,EAChC;AAAA,EACA;AACF;AAUO,SAAS,yBAAyB,UAAqC,CAAC,GAAG;AAChF,QAAM,kBAAkB,CAAC,GAAG,iBAAiB,GAAI,QAAQ,WAAW,CAAC,CAAE;AAEvE,SAAO,SAAS,WAAW,SAAsB;AAC/C,UAAM,EAAE,SAAS,IAAI,QAAQ;AAG7B,QAAI,YAAY,IAAI,QAAQ,GAAG;AAC7B,aAAO,aAAa,KAAK;AAAA,IAC3B;AAGA,QAAI,WAAW,UAAU,eAAe,GAAG;AACzC,aAAO,aAAa,KAAK;AAAA,IAC3B;AAEA,UAAM,KAAK,QAAQ,QAAQ,IAAI,YAAY;AAC3C,UAAM,SAAS,QAAQ,QAAQ,IAAI,QAAQ;AAC3C,UAAM,QAAQ,YAAY,IAAI,MAAM;AAGpC,QAAI,SAAS,SAAS,KAAK,GAAG;AAC5B,YAAM,eAAe,SAAS,MAAM,GAAG,EAAE,KAAK;AAC9C,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,YAAY;AAClD,aAAO,cAAc,aAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,QAAI,MAAM,SAAS;AACjB,YAAM,eAAe,IAAI,IAAI,4BAA4B,QAAQ,GAAG;AACpE,mBAAa,aAAa,IAAI,QAAQ,QAAQ;AAC9C,aAAO,cAAc,aAAa,QAAQ,YAAY,CAAC;AAAA,IACzD;AAGA,UAAM,WAAW,aAAa,KAAK;AACnC,aAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAO;AAAA,EACT;AACF;AAMA,SAAS,cAAc,UAAsC;AAC3D,WAAS,QAAQ,IAAI,uBAAuB,QAAQ;AACpD,WAAS,QAAQ,IAAI,QAAQ,oBAAoB;AACjD,WAAS,QAAQ,IAAI,gBAAgB,KAAK;AAC1C,WAAS,QAAQ,OAAO,uBAAuB;AAC/C,SAAO;AACT;AAEA,SAAS,WAAW,MAAc,UAA6B;AAC7D,SAAO,SAAS,KAAK,CAAC,YAAY,UAAU,SAAS,IAAI,CAAC;AAC5D;AAEA,SAAS,UAAU,SAAiB,MAAuB;AACzD,QAAM,QAAQ,QACX,QAAQ,SAAS,gBAAgB,EACjC,QAAQ,OAAO,OAAO,EACtB,QAAQ,mBAAmB,IAAI;AAClC,SAAO,IAAI,OAAO,IAAI,KAAK,GAAG,EAAE,KAAK,IAAI;AAC3C;","names":[]}

package/package.json

@@ -0,0 +1,72 @@
+{
+  "name": "@agent-seo/next",
+  "version": "1.0.0",
+  "description": "Next.js plugin for AI-readable websites",
+  "license": "MIT",
+  "repository": {
+    "type": "git",
+    "url": "https://github.com/pontiggia/agent-seo",
+    "directory": "packages/next"
+  },
+  "homepage": "https://github.com/pontiggia/agent-seo#readme",
+  "bugs": "https://github.com/pontiggia/agent-seo/issues",
+  "keywords": [
+    "seo",
+    "ai",
+    "llm",
+    "nextjs",
+    "next",
+    "middleware",
+    "llms-txt",
+    "agent",
+    "bot"
+  ],
+  "type": "module",
+  "engines": {
+    "node": ">=20.0.0"
+  },
+  "main": "dist/index.cjs",
+  "module": "dist/index.js",
+  "types": "dist/index.d.ts",
+  "exports": {
+    ".": {
+      "import": {
+        "types": "./dist/index.d.ts",
+        "default": "./dist/index.js"
+      },
+      "require": {
+        "types": "./dist/index.d.cts",
+        "default": "./dist/index.cjs"
+      }
+    },
+    "./middleware": {
+      "import": {
+        "types": "./dist/middleware.d.ts",
+        "default": "./dist/middleware.js"
+      },
+      "require": {
+        "types": "./dist/middleware.d.cts",
+        "default": "./dist/middleware.cjs"
+      }
+    }
+  },
+  "files": [
+    "dist"
+  ],
+  "dependencies": {
+    "@agent-seo/core": "1.0.0"
+  },
+  "peerDependencies": {
+    "next": "^14.0.0 || ^15.0.0 || ^16.0.0"
+  },
+  "devDependencies": {
+    "next": "^16.0.0",
+    "react": "^19.0.0",
+    "react-dom": "^19.0.0"
+  },
+  "scripts": {
+    "build": "tsup",
+    "test": "vitest run",
+    "typecheck": "tsc --noEmit"
+  }
+}