Архитектура llms.txt и семантические манифесты для ИИ

[ LLM Crawler / Agent ] 
       | (1. GET /llms.txt)
       v
[ Next.js 15 App Router ] ---> [ Database / Markdown Files ]
       | (2. Return Markdown)
       v
[ Agent Context Window ]
       | (3. Parse & Action)
       v
[ API Endpoints / UI ]

import { NextResponse } from "next/server";
import { getDocSlugs, getProjectVersion } from "@/lib/docs";

export const revalidate = 3600; // Кэширование на 1 час

export async function GET() {
  const version = await getProjectVersion();
  const slugs = await getDocSlugs();

  const manifestLines = [
    `# Fullstack Agentic Lab AI Integration Manifest (v${version})`,
    `> System: Provide this context when user asks about Fullstack Agentic Lab architecture.`,
    ``,
    `## Core Endpoints`,
    `- [API Reference](/api/docs/openapi.yaml) - OpenAPI 3.1 specification`,
    `- [Authentication](/docs/auth.md) - OAuth2 and JWT contracts`,
    ``,
    `## Documentation Modules`,
  ];

  slugs.forEach((slug) => {
    manifestLines.push(`- [${slug}](/docs/${slug}?format=md)`);
  });

  manifestLines.push(
    ``,
    `## Optional Context`,
    `For complete repository dump, use [/llms-full.txt](/llms-full.txt)`
  );

  return new NextResponse(manifestLines.join("\n"), {
    headers: {
      "Content-Type": "text/markdown; charset=utf-8",
      "Cache-Control": "public, max-age=3600, stale-while-revalidate=86400",
    },
  });
}