diff --git a/Dockerfile b/Dockerfile
index b9b60e4..6f3132d 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,10 +1,14 @@
-FROM ruby:3.0-alpine
+FROM alpine:3.17
 
-RUN apk --no-cache add \
-    build-base \
-    curl \
-    ruby-dev \
-  && gem install html-proofer -v 3.19.4
+RUN apk --no-cache add build-base ruby-dev
+RUN apk --no-cache add curl
+RUN gem install html-proofer -v 5.0.2
+
+RUN apk --no-cache add python3 py3-pip
+RUN apk --no-cache add openjdk8
+RUN pip install html5validator
+
+RUN apk --no-cache add bash
 
 COPY entrypoint.sh proof-html.rb /
 
diff --git a/README.md b/README.md
index cbdfa8b..0e1c18f 100644
--- a/README.md
+++ b/README.md
@@ -1,12 +1,16 @@
 # proof-html
 
 proof-html is a [GitHub Action](https://github.com/features/actions) to
-validate HTML using [HTMLProofer](https://github.com/gjtorikian/html-proofer).
+validate HTML using the [Nu HTML
+Validator](https://github.com/validator/validator) through
+[html5validator](https://github.com/svenkreiss/html5validator), and check
+links, images, and more using
+[HTMLProofer](https://github.com/gjtorikian/html-proofer).
 
 ## Usage
 
 ```yaml
-- uses: anishathalye/proof-html@v1
+- uses: anishathalye/proof-html@v2
   with:
     directory: ./site
 ```
@@ -21,14 +25,12 @@ See below for a [full example](#full-example).
 | `check_external_hash` | Check whether external anchors exist | true |
 | `check_favicon` | Check whether favicons are valid | true |
 | `check_html` | Validate HTML | true |
-| `check_img_http` | Enforce that images use HTTPS | true |
 | `check_opengraph` | Check images and URLs in Open Graph metadata | true |
-| `empty_alt_ignore` | Allow images with empty alt tags | false |
+| `ignore_empty_alt` | Allow images with empty alt tags | false |
 | `enforce_https` | Require that links use HTTPS | true |
-| `external_only` | Only check external links | false |
-| `internal_domains` | Newline-separated list of domains to treat as internal URLs | (empty) |
-| `url_ignore` | Newline-separated list of URLs to ignore | (empty) |
-| `url_ignore_re` | Newline-separated list of URL regexes to ignore | (empty) |
+| `swap_urls` | JSON-encoded map of URL rewrite rules | (empty) |
+| `ignore_url` | Newline-separated list of URLs to ignore | (empty) |
+| `ignore_url_re` | Newline-separated list of URL regexes to ignore | (empty) |
 | `connect_timeout` | HTTP connection timeout | 30 |
 | `tokens` | JSON-encoded map of domains to authorization tokens | (empty) |
 | `max_concurrency` | Maximum number of concurrent requests | 50 |
@@ -38,6 +40,8 @@ See below for a [full example](#full-example).
 Most of the options correspond directly to [configuration options for
 HTMLProofer](https://github.com/gjtorikian/html-proofer#configuration).
 
+**tokens**
+
 `tokens` is a _JSON-encoded_ map of domains to authorization tokens. So it's
 "doubly encoded": the workflow file is written in YAML and `tokens` is a string
 (not a map!), a JSON encoding of the data. This option can be used to provide
@@ -55,6 +59,16 @@ tokens: |
 You can also see the full example below for how to pass on the `GITHUB_TOKEN`
 supplied by the workflow runner.
 
+**swap_urls**
+
+`swap_urls` is a _JSON-encoded_ map, mapping regexes to strings. This can be
+useful to strip a base path for an internal domain. For example:
+
+```yaml
+swap_urls: |
+  {"^https://example.com/": "/"}
+```
+
 ## Full Example
 
 This is the entire `.github/workflows/build.yml` file for a GitHub Pages /
@@ -84,19 +98,19 @@ jobs:
           bundle config path vendor/bundle
           bundle install --jobs 4 --retry 3
       - run: bundle exec jekyll build
-      - uses: anishathalye/proof-html@v1
+      - uses: anishathalye/proof-html@v2
         with:
           directory: ./_site
           enforce_https: false
           tokens: |
             {"https://github.com": "${{ secrets.GITHUB_TOKEN }}"}
-          url_ignore: |
+          ignore_url: |
             http://www.example.com/
             https://en.wikipedia.org/wiki/Main_Page
-          url_ignore_re: |
+          ignore_url_re: |
             ^https://twitter.com/
-          internal_domains: |
-            www.anishathalye.com
+          swap_urls: |
+            {"^https://www.anishathalye.com/": "/"}
 ```
 
 ### Real-world examples
diff --git a/action.yml b/action.yml
index 992fef5..110d520 100644
--- a/action.yml
+++ b/action.yml
@@ -12,30 +12,24 @@ inputs:
   check_html:
     description: Validate HTML
     required: false
-  check_img_http:
-    description: Enforce that images use HTTPS
-    required: false
   check_opengraph:
     description: Check images and URLs in Open Graph metadata
     required: false
   check_favicon:
     description: Check whether favicons are valid
     required: false
-  empty_alt_ignore:
+  ignore_empty_alt:
     description: Allow images with empty alt tags
     required: false
   enforce_https:
     description: Require that links use HTTPS
     required: false
-  external_only:
-    description: Only check external links
-    required: false
-  internal_domains:
-    description: Newline-separated list of domains to treat as internal URLs
-    required: false
   tokens:
     description: JSON-encoded map of domains to authorization tokens
     required: false
+  swap_urls:
+    description: JSON-encoded map of URL rewrite rules
+    required: false
   max_concurrency:
     description: Maximum number of concurrent requests
     required: false
@@ -45,10 +39,10 @@ inputs:
   timeout:
     description: HTTP request timeout
     required: false
-  url_ignore:
+  ignore_url:
     description: Newline-separated list of URLs to ignore
     required: false
-  url_ignore_re:
+  ignore_url_re:
     description: Newline-separated list of URL regexes to ignore
     required: false
   retries:
@@ -57,7 +51,7 @@ inputs:
 
 runs:
   using: docker
-  image: docker://anishathalye/proof-html:1.4.1
+  image: docker://anishathalye/proof-html:2.0.0
 
 branding:
   icon: check-square
diff --git a/entrypoint.sh b/entrypoint.sh
index 309ac94..641182e 100755
--- a/entrypoint.sh
+++ b/entrypoint.sh
@@ -1,18 +1,27 @@
-#!/bin/sh
+#!/usr/bin/env bash
+
+failed=0
+
+check_html="${INPUT_CHECK_HTML:-true}"
+if [[ "$check_html" =~ ^t.*|^T.*|^y.*|^Y.*|^1.* ]]; then
+  if ! html5validator --also-check-css --log INFO --root "${INPUT_DIRECTORY}"; then
+    failed=1
+  fi
+fi
 
 tries="${INPUT_RETRIES:-3}"
 
-code=1
 while [ "$tries" -ge 1 ]; do
   tries=$((tries-1))
-  ruby /proof-html.rb
-  code="$?"
-  if [ "$code" -eq 0 ]; then
+  if ruby /proof-html.rb; then
     break
   fi
   if [ "$tries" -ge 1 ]; then
     sleep 5
   fi
+  if [ "$tries" -eq 0 ]; then
+    failed=1
+  fi
 done
 
-exit "$code"
+exit $failed
diff --git a/proof-html.rb b/proof-html.rb
index 72f0e49..66c8951 100644
--- a/proof-html.rb
+++ b/proof-html.rb
@@ -30,26 +30,35 @@ def get_str(name)
   s.nil? ? "" : s
 end
 
-url_ignore_re = get_str("URL_IGNORE_RE").split("\n").map { |s| Regexp.new s }
-url_ignore = get_str("URL_IGNORE").split("\n").concat url_ignore_re
+ignore_url_re = get_str("IGNORE_URL_RE").split("\n").map { |s| Regexp.new s }
+ignore_url = get_str("IGNORE_URL").split("\n").concat ignore_url_re
 tokens_str = get_str("TOKENS")
 tokens = JSON.parse (tokens_str == "" ? "{}" : tokens_str)
-internal_domains = get_str("INTERNAL_DOMAINS").split("\n")
+
+swap_urls_str = get_str("SWAP_URLS")
+swap_urls = JSON.parse (swap_urls_str == "" ? "{}" : swap_urls_str)
+swap_urls.transform_keys! { |k| Regexp.new k }
+
+checks = ["Links", "Scripts", "Images"]
+if get_bool("CHECK_FAVICON", true)
+  checks.push("Favicon")
+end
+if get_bool("CHECK_OPENGRAPH", true)
+  checks.push("OpenGraph")
+end
 
 options = {
-  :cache => { :timeframe => "1d" },
+  :checks => checks,
+  :cache => { :timeframe => {
+    :internal => "1d",
+    :external => "1d",
+  } },
   :check_external_hash => get_bool("CHECK_EXTERNAL_HASH", true),
-  :check_html => get_bool("CHECK_HTML", true),
-  :check_img_http => get_bool("CHECK_IMG_HTTP", true),
-  :check_opengraph => get_bool("CHECK_OPENGRAPH", true),
-  :check_favicon => get_bool("CHECK_FAVICON", true),
-  :empty_alt_ignore => get_bool("EMPTY_ALT_IGNORE", false),
+  :ignore_empty_alt => get_bool("IGNORE_EMPTY_ALT", false),
   :enforce_https => get_bool("ENFORCE_HTTPS", true),
-  :external_only => get_bool("EXTERNAL_ONLY", false),
   :hydra => {
     :max_concurrency => get_int("MAX_CONCURRENCY", 50),
   },
-  :internal_domains => internal_domains,
   :typhoeus => {
     :connecttimeout => get_int("CONNECT_TIMEOUT", 30),
     :followlocation => true,
@@ -58,7 +67,8 @@ def get_str(name)
     },
     :timeout => get_int("TIMEOUT", 120),
   },
-  :url_ignore => url_ignore,
+  :ignore_urls => ignore_url,
+  :swap_urls => swap_urls,
 }
 
 begin