diff --git a/src/robots.txt.ejs b/src/robots.txt.ejs
index b77ec10a556..64d93747026 100644
--- a/src/robots.txt.ejs
+++ b/src/robots.txt.ejs
@@ -9,8 +9,6 @@ Sitemap: <%= origin %>/sitemap_index.html
 ##########################
 User-agent: *
 # Disable access to Discovery search and filters; admin pages; processes; submission; workspace; workflow & profile page
-Disallow: /browse
-Disallow: /handle
 Disallow: /search
 Disallow: /admin/*
 Disallow: /processes
@@ -18,6 +16,21 @@ Disallow: /submit
 Disallow: /workspaceitems
 Disallow: /profile
 Disallow: /workflowitems
+# Crawlers should be able to access entity pages, but not the facet search links present on entity pages
+Disallow: /entities/*?f
+# Scoped search inside a community or collection, and every facet link.
+# "Disallow: /search" above only matches paths that START with /search, so
+# /collections/<uuid>/search and /communities/<uuid>/search stayed crawlable.
+# Crawlers enumerating those facet combinations saturated SSR on 2026-07-22
+# (HTTP 504 for everyone); see the comment in config/config.yml.
+Disallow: /collections/*/search
+Disallow: /communities/*/search
+# The two rules below cover any URL carrying a Discovery facet filter
+# (f.author, f.subject, f.dateIssued.min, ...), wherever it appears. Two of
+# them because the facet can be the first query parameter (?f.) or a later
+# one (&f.), and robots.txt has no way to express "either".
+Disallow: /*?f.
+Disallow: /*&f.
 
 # Optionally uncomment the following line ONLY if sitemaps are working
 # and you have verified that your site is being indexed correctly.