Feed integration / Set up the proxy

Feed integration

Set up the proxy

Connection details and the five steps to route /feed to the Branly feed service, verify it and launch.

Everything your engineering team needs to route /feed on your website to the Branly feed service, from the values to configure to the checks before launch.

Connection details

SettingValue
Paths to route/feed and /feed/* (exactly these; not /feedback and so on)
Upstream (origin)https://feeds.branly.ai, HTTPS only, publicly trusted certificate
Upstream pathUnchanged: /feed/<anything> → https://feeds.branly.ai/feed/<anything>
Host headerfeeds.branly.ai (also the TLS SNI name)
Required headerX-Forwarded-Host: <your public host>, exactly the host you gave us (with or without www). Set it in your proxy and overwrite any value sent by the visitor.
MethodsGET and HEAD
Query stringsForward all (used for pagination and page navigation)
Request headersForward the visitor's headers, except Cookie and Authorization
Origin credentialNone required
Your public URLsFeed /feed · post /feed/<post-slug> · sitemap /feed/sitemap.xml

Engineering setup

Confirm the host

Send us the exact public host for the feed (for example www.example.com) and the names and emails of your editors. We register the host and send the invitations.

A request whose X-Forwarded-Host doesn't match a registered host gets a 404.

Configure the proxy

Use the values in Connection details, before your default website rule. Keep responses as they come: status codes, content types, cache headers, and redirects. Our redirects are relative (for example /feed/ → /feed), so they stay on your host. Platform examples are in Edge routing examples.

Caching: don't cache feed pages at your edge; they are sent with Cache-Control: no-store so new posts show at once. Files under /feed/_next/static/ are versioned and may be cached for a year. Images under /feed/api/media/file/ may be cached briefly.

Connect the sitemap

Add this line to your root robots.txt, or add the URL to your sitemap index, then submit it in Google Search Console:

robots.txt
Sitemap: https://www.example.com/feed/sitemap.xml

The sitemap and each page's canonical URL use your public host. Make sure bot challenges and rate limits don't block approved crawlers on /feed.

Verify and launch

Before changing your site, check our side (after we register your host):

Check our side
curl -s -H 'X-Forwarded-Host: www.example.com' https://feeds.branly.ai/feed \
  | grep -o '<title>[^<]*'          # shows your feed title

After the rule is live, check from outside your network:

Check your site
curl -sI https://www.example.com/feed                    # 200, text/html
curl -s  https://www.example.com/feed | grep canonical   # https://www.example.com/feed
curl -sI https://www.example.com/feed/sitemap.xml        # 200, application/xml
curl -sI https://www.example.com/feed/no-such-post       # 404
curl -sI https://www.example.com/feed/admin              # 404 (admin is not public here)

Also open a post, click through to another post and to page 2 of the feed, and use Search Console's live test on a sample post.

Roll back if needed

Remove the /feed rule, purge any cached responses, and remove the sitemap line. Contact us to coordinate redirects for URLs that were already published.

Last updated 19 November 2026

Questions about this page? Write to connect@branly.ai