> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cognerd.in/llms.txt
> Use this file to discover all available pages before exploring further.

# Generate GEO Files for AI Crawlers and Agents

> Use CogNerd to auto-generate llms.txt, schema.org JSON-LD, sitemap.xml, and five other structured files that help AI platforms understand your brand.

Generative Engine Optimization (GEO) files are structured assets that tell AI crawlers, agents, and language models how to read and represent your brand. Without them, AI platforms have to guess your brand's context from unstructured web content — and they often get it wrong or miss you entirely. CogNerd's GEO pipeline crawls your site and generates seven ready-to-deploy files in one automated job.

<Steps>
  <Step title="Go to GEO Assets">
    In the left sidebar, click **GEO Assets**, or navigate to `/dashboard/geo-assets`.
  </Step>

  <Step title="Enter your website URL">
    Type in your primary domain (e.g., `https://yourdomain.com`) and click **Generate**. Make sure this is the publicly accessible root of your site — CogNerd needs to crawl it.

    <Warning>
      Generating GEO files costs **20 credits** per run. You'll be shown a confirmation before the job starts. Check your credit balance under **Credits** in the sidebar if you're unsure.
    </Warning>
  </Step>

  <Step title="Wait for the pipeline to complete">
    CogNerd runs a four-stage pipeline in the background:

    1. **Page Discovery** — maps all crawlable pages on your site
    2. **Content Crawl** — reads and extracts content from each discovered page
    3. **Enhanced Asset Generation** — builds structured representations of your brand, entities, and content
    4. **Final Report** — compiles and packages all seven output files

    The full job typically takes several minutes. You'll see live status updates on the dashboard — no need to stay on the page.

    <Note>
      You receive a job ID when the run starts. If you navigate away and come back, the dashboard will show the current status of your job automatically.
    </Note>
  </Step>

  <Step title="Download your GEO files">
    When the pipeline finishes, click **Download ZIP** to get all seven files in a single archive:

    | File                  | Purpose                                                           |
    | --------------------- | ----------------------------------------------------------------- |
    | `llms.txt`            | Machine-readable brand summary for AI agents and language models  |
    | `robots.txt`          | AI-aware crawl rules that guide what AI bots can and can't access |
    | `sitemap.xml`         | Page index for efficient crawling and indexing                    |
    | `site.jsonld`         | Site-level entity schema (schema.org)                             |
    | `organization.jsonld` | Brand identity schema including name, logo, and social profiles   |
    | `website.jsonld`      | Website and search action schema                                  |
    | `entities.json`       | Entity relationship graph describing your brand's key concepts    |
  </Step>

  <Step title="Deploy the files to your site">
    Each file has a specific deployment target:

    <Tabs>
      <Tab title="llms.txt">
        Place `llms.txt` at the root of your domain so it's accessible at `yourdomain.com/llms.txt`. This is the most important file — AI agents check for it the same way browsers check for `robots.txt`.
      </Tab>

      <Tab title="robots.txt">
        Replace or merge with your existing `robots.txt` at the root of your domain (`yourdomain.com/robots.txt`). Review the AI-specific directives before deploying to make sure they match your preferences.
      </Tab>

      <Tab title="sitemap.xml">
        Deploy to the root of your domain (`yourdomain.com/sitemap.xml`) and reference it in your `robots.txt` file. If you already have a sitemap, compare both files before replacing.
      </Tab>

      <Tab title="JSON-LD schemas">
        Embed `site.jsonld`, `organization.jsonld`, and `website.jsonld` as `<script type="application/ld+json">` tags in the `<head>` of your relevant pages. The organization schema typically goes on every page; the website schema goes on your homepage.
      </Tab>

      <Tab title="entities.json">
        This file is intended for advanced use — it can be referenced by AI agents consuming your llms.txt, or used to power internal tooling. It doesn't need to be publicly served unless your llms.txt links to it.
      </Tab>
    </Tabs>
  </Step>
</Steps>

<Tip>
  After deploying your GEO files, wait a week and then re-run your Brand Analysis. Most brands see a measurable improvement in their AI Visibility Score within 7–14 days of deployment as AI crawlers pick up the new structured data.
</Tip>

<Info>
  GEO files reflect your site at the time of the crawl. If you make significant changes to your site, products, or brand, re-run the pipeline to keep your files current. You can regenerate as often as needed.
</Info>