go-away: use config.yml and vendor all configuration files

Signed-off-by: Robert Günzler <r@gnzler.io>

15 files changed, 250 insertions, 0 deletions

diff --git a/bundles/go-away/files/config/snippets/bot-betterstack.yml b/bundles/go-away/files/config/snippets/bot-betterstack.yml
new file mode 100644
index 0000000..41bd2a1
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-betterstack.yml
@@ -0,0 +1,8 @@
+networks:
+  betterstack:
+    - url: https://uptime.betterstack.com/ips-by-cluster.json
+      jq-path: '.[] | .[]'
+
+conditions:
+  is-bot-betterstack:
+      - &is-bot-betterstack '((userAgent.startsWith("Better Stack Better Uptime Bot") || userAgent.startsWith("Better Uptime Bot") || userAgent == "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/127.0.6533.17 Safari/537.36")) && remoteAddress.network("betterstack")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-bingbot.yml b/bundles/go-away/files/config/snippets/bot-bingbot.yml
new file mode 100644
index 0000000..98fb3db
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-bingbot.yml
@@ -0,0 +1,8 @@
+networks:
+  bingbot:
+    - url: https://www.bing.com/toolbox/bingbot.json
+      jq-path: '(.prefixes[] | select(has("ipv4Prefix")) | .ipv4Prefix), (.prefixes[] | select(has("ipv6Prefix")) | .ipv6Prefix)'
+
+conditions:
+  is-bot-bingbot:
+      - &is-bot-bingbot 'userAgent.contains("+http://www.bing.com/bingbot.htm") && remoteAddress.network("bingbot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-duckduckbot.yml b/bundles/go-away/files/config/snippets/bot-duckduckbot.yml
new file mode 100644
index 0000000..b5f45e1
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-duckduckbot.yml
@@ -0,0 +1,8 @@
+networks:
+  duckduckbot:
+    - url: https://duckduckgo.com/duckduckgo-help-pages/results/duckduckbot
+      regex: "<li><div>(?P<prefix>[0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+)</div></li>"
+
+conditions:
+  is-bot-duckduckbot:
+      - &is-bot-duckduckbot 'userAgent.contains("+http://duckduckgo.com/duckduckbot.html") && remoteAddress.network("duckduckbot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-googlebot.yml b/bundles/go-away/files/config/snippets/bot-googlebot.yml
new file mode 100644
index 0000000..da26884
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-googlebot.yml
@@ -0,0 +1,8 @@
+networks:
+  googlebot:
+    - url: https://developers.google.com/static/search/apis/ipranges/googlebot.json
+      jq-path: '(.prefixes[] | select(has("ipv4Prefix")) | .ipv4Prefix), (.prefixes[] | select(has("ipv6Prefix")) | .ipv6Prefix)'
+
+conditions:
+  is-bot-googlebot:
+      - &is-bot-googlebot '(userAgent.contains("+http://www.google.com/bot.html") || userAgent.contains("Google-PageRenderer") || userAgent.contains("Google-InspectionTool") || userAgent.contains("Googlebot")) && remoteAddress.network("googlebot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-kagibot.yml b/bundles/go-away/files/config/snippets/bot-kagibot.yml
new file mode 100644
index 0000000..8edb3c3
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-kagibot.yml
@@ -0,0 +1,8 @@
+networks:
+  kagibot:
+    - url: https://kagi.com/bot
+      regex: "\\n(?P<prefix>[0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+) "
+
+conditions:
+  is-bot-kagibot:
+      - &is-bot-kagibot 'userAgent.contains("+https://kagi.com/bot") && remoteAddress.network("kagibot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-qwantbot.yml b/bundles/go-away/files/config/snippets/bot-qwantbot.yml
new file mode 100644
index 0000000..095e1be
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-qwantbot.yml
@@ -0,0 +1,8 @@
+networks:
+  qwantbot:
+    - url: https://help.qwant.com/wp-content/uploads/sites/2/2025/01/qwantbot.json
+      jq-path: '(.prefixes[] | select(has("ipv4Prefix")) | .ipv4Prefix), (.prefixes[] | select(has("ipv6Prefix")) | .ipv6Prefix)'
+
+conditions:
+  is-bot-qwantbot:
+      - &is-bot-qwantbot 'userAgent.contains("+https://help.qwant.com/bot/") && remoteAddress.network("qwantbot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/bot-uptimerobot.yml b/bundles/go-away/files/config/snippets/bot-uptimerobot.yml
new file mode 100644
index 0000000..848424c
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-uptimerobot.yml
@@ -0,0 +1,8 @@
+networks:
+  uptimerobot:
+    - url: https://uptimerobot.com/inc/files/ips/IPv4andIPv6.txt
+      regex: "(?P<prefix>[0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+(/[0-9]+)?|[0-9a-f:]+:.+)"
+
+conditions:
+  is-bot-uptimerobot:
+      - &is-bot-uptimerobot 'userAgent.contains("http://www.uptimerobot.com/") && remoteAddress.network("uptimerobot")'
diff --git a/bundles/go-away/files/config/snippets/bot-yandexbot.yml b/bundles/go-away/files/config/snippets/bot-yandexbot.yml
new file mode 100644
index 0000000..981ee42
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/bot-yandexbot.yml
@@ -0,0 +1,24 @@
+networks:
+  yandexbot:
+    # todo: detected as bot
+    # - url: https://yandex.com/ips
+    #  regex: "<span>(?P<prefix>(([0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+)|([0-9a-f:]+::))/[0-9]+)[ \\\\t]*</span><br/>"
+    - prefixes:
+      - "5.45.192.0/18"
+      - "5.255.192.0/18"
+      - "37.9.64.0/18"
+      - "37.140.128.0/18"
+      - "77.88.0.0/18"
+      - "84.252.160.0/19"
+      - "87.250.224.0/19"
+      - "90.156.176.0/22"
+      - "93.158.128.0/18"
+      - "95.108.128.0/17"
+      - "141.8.128.0/18"
+      - "178.154.128.0/18"
+      - "185.32.187.0/24"
+      - "2a02:6b8::/29"
+
+conditions:
+  is-bot-yandexbot:
+      - &is-bot-yandexbot 'userAgent.contains("+http://yandex.com/bots") && remoteAddress.network("yandexbot")'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/challenge-dnsbl.yml b/bundles/go-away/files/config/snippets/challenge-dnsbl.yml
new file mode 100644
index 0000000..893116c
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/challenge-dnsbl.yml
@@ -0,0 +1,6 @@
+challenges:
+  dnsbl:
+    runtime: dnsbl
+    parameters:
+      dnsbl-decay: 1h
+      dnsbl-timeout: 1s
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/challenge-js-pow-sha256.yml b/bundles/go-away/files/config/snippets/challenge-js-pow-sha256.yml
new file mode 100644
index 0000000..78f57f0
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/challenge-js-pow-sha256.yml
@@ -0,0 +1,15 @@
+challenges:
+  js-pow-sha256:
+    runtime: js
+    parameters:
+      # specifies the folder path that assets are under
+      # can be either embedded or external path
+      # defaults to name of challenge
+      path: "js-pow-sha256"
+      # needs to be under static folder
+      js-loader: load.mjs
+      # needs to be under runtime folder
+      wasm-runtime: runtime.wasm
+      wasm-runtime-settings:
+        difficulty: 20
+      verify-probability: 0.1
diff --git a/bundles/go-away/files/config/snippets/challenge-js-refresh.yml b/bundles/go-away/files/config/snippets/challenge-js-refresh.yml
new file mode 100644
index 0000000..73ba729
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/challenge-js-refresh.yml
@@ -0,0 +1,6 @@
+challenges:
+  js-refresh:
+    # Challenges with a redirect via window.location (requires HTML parsing and JavaScript logic)
+    runtime: "refresh"
+    parameters:
+      refresh-via: "javascript"
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/challenges-non-js.yml b/bundles/go-away/files/config/snippets/challenges-non-js.yml
new file mode 100644
index 0000000..b8aff89
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/challenges-non-js.yml
@@ -0,0 +1,28 @@
+challenges:
+  # Challenges with a cookie, self redirect (non-JS, requires HTTP parsing)
+  cookie:
+    runtime: "cookie"
+
+  # Challenges with a redirect via Link header with rel=preload and early hints (non-JS, requires HTTP parsing, fetching and logic)
+  # Works on HTTP/2 and above!
+  preload-link:
+    condition: '"Sec-Fetch-Mode" in headers && headers["Sec-Fetch-Mode"] == "navigate"'
+    runtime: "preload-link"
+    parameters:
+      preload-early-hint-deadline: 2s
+
+  # Challenges with a redirect via Refresh header (non-JS, requires HTTP parsing and logic)
+  header-refresh:
+    runtime: "refresh"
+    parameters:
+      refresh-via: "header"
+
+  # Challenges with a redirect via Refresh meta (non-JS, requires HTML parsing and logic)
+  meta-refresh:
+    runtime: "refresh"
+    parameters:
+      refresh-via: "meta"
+
+  # Challenges with loading a random CSS or image document (non-JS, requires HTML parsing and logic)
+  resource-load:
+    runtime: "resource-load"
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/conditions-generic.yml b/bundles/go-away/files/config/snippets/conditions-generic.yml
new file mode 100644
index 0000000..4e64cca
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/conditions-generic.yml
@@ -0,0 +1,56 @@
+conditions:
+  is-well-known-asset:
+    # general txt files or scraper
+    - 'path == "/robots.txt" || path == "/security.txt"'
+
+    # ads txt files
+    - 'path == "/app-ads.txt" || path == "/ads.txt"'
+
+    # generally requested by browsers
+    - 'path == "/favicon.ico"'
+
+    # used by some applications
+    - 'path == "/crossdomain.xml"'
+
+    # well-known paths
+    - 'path.startsWith("/.well-known/")'
+
+  is-git-ua:
+    - 'userAgent.startsWith("git/") || userAgent.contains("libgit")'
+    - 'userAgent.startsWith("go-git")'
+    - 'userAgent.startsWith("JGit/") || userAgent.startsWith("JGit-")'
+    # Golang proxy and initial fetch
+    - 'userAgent.startsWith("GoModuleMirror/")'
+    - 'userAgent.startsWith("Go-http-client/") && "go-get" in query && query["go-get"] == "1"'
+    - '"Git-Protocol" in headers && headers["Git-Protocol"] == "version=2"'
+
+  is-generic-browser:
+    - 'userAgent.startsWith("Mozilla/") || userAgent.startsWith("Opera/")'
+
+  is-generic-robot-ua:
+    - 'userAgent.matches("compatible[;)]") && !userAgent.contains("Trident/")'
+    - 'userAgent.matches("\\+https?://")'
+    - 'userAgent.contains("@")'
+    - 'userAgent.matches("[bB]ot/[0-9]")'
+
+  is-tool-ua:
+    - 'userAgent.startsWith("python-requests/")'
+    - 'userAgent.startsWith("Python-urllib/")'
+    - 'userAgent.startsWith("python-httpx/")'
+    - 'userAgent.contains("aoihttp/")'
+    - 'userAgent.startsWith("http.rb/")'
+    - 'userAgent.startsWith("curl/")'
+    - 'userAgent.startsWith("Wget/")'
+    - 'userAgent.startsWith("libcurl/")'
+    - 'userAgent.startsWith("okhttp/")'
+    - 'userAgent.startsWith("Java/")'
+    - 'userAgent.startsWith("Apache-HttpClient//")'
+    - 'userAgent.startsWith("Go-http-client/")'
+    - 'userAgent.startsWith("node-fetch/")'
+    - 'userAgent.startsWith("reqwest/")'
+
+  # Checks to detect a headless chromium via headers only
+  is-headless-chromium:
+    - 'userAgent.contains("HeadlessChrome") || userAgent.contains("HeadlessChromium")'
+    - '"Sec-Ch-Ua" in headers && (headers["Sec-Ch-Ua"].contains("HeadlessChrome") || headers["Sec-Ch-Ua"].contains("HeadlessChromium"))'
+    #- '(userAgent.contains("Chrome/") || userAgent.contains("Chromium/")) && (!("Accept-Language" in headers) || !("Accept-Encoding" in headers))'
\ No newline at end of file
diff --git a/bundles/go-away/files/config/snippets/networks-other.yml b/bundles/go-away/files/config/snippets/networks-other.yml
new file mode 100644
index 0000000..c5ffdda
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/networks-other.yml
@@ -0,0 +1,37 @@
+networks:
+  aws-cloud:
+    - url: https://ip-ranges.amazonaws.com/ip-ranges.json
+      jq-path: '(.prefixes[] | select(has("ip_prefix")) | .ip_prefix), (.prefixes[] | select(has("ipv6_prefix")) | .ipv6_prefix)'
+  google-cloud:
+    - url: https://www.gstatic.com/ipranges/cloud.json
+      jq-path: '(.prefixes[] | select(has("ipv4Prefix")) | .ipv4Prefix), (.prefixes[] | select(has("ipv6Prefix")) | .ipv6Prefix)'
+  oracle-cloud:
+    - url: https://docs.oracle.com/en-us/iaas/tools/public_ip_ranges.json
+      jq-path: '.regions[] | .cidrs[] | .cidr'
+  azure-cloud:
+    # todo: https://www.microsoft.com/en-us/download/details.aspx?id=56519 does not provide direct JSON
+    - url: https://raw.githubusercontent.com/femueller/cloud-ip-ranges/refs/heads/master/microsoft-azure-ip-ranges.json
+      jq-path: '.values[] | .properties.addressPrefixes[]'
+
+  digitalocean:
+    - url: https://www.digitalocean.com/geo/google.csv
+      regex: "(?P<prefix>(([0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+)|([0-9a-f:]+::))/[0-9]+),"
+  linode:
+    - url: https://geoip.linode.com/
+      regex: "(?P<prefix>(([0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+)|([0-9a-f:]+::))/[0-9]+),"
+  vultr:
+    - url: "https://geofeed.constant.com/?json"
+      jq-path: '.subnets[] | .ip_prefix'
+  cloudflare:
+    - url: https://www.cloudflare.com/ips-v4
+      regex: "(?P<prefix>[0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+/[0-9]+)"
+    - url: https://www.cloudflare.com/ips-v6
+      regex: "(?P<prefix>[0-9a-f:]+::/[0-9]+)"
+
+  icloud-private-relay:
+    - url: https://mask-api.icloud.com/egress-ip-ranges.csv
+      regex: "(?P<prefix>(([0-9]+\\.[0-9]+\\.[0-9]+\\.[0-9]+)|([0-9a-f:]+::))/[0-9]+),"
+  tunnelbroker-relay:
+    # HE Tunnelbroker
+    - url: https://tunnelbroker.net/export/google
+      regex: "(?P<prefix>([0-9a-f:]+::)/[0-9]+),"
diff --git a/bundles/go-away/files/config/snippets/networks-private.yml b/bundles/go-away/files/config/snippets/networks-private.yml
new file mode 100644
index 0000000..4d81b6a
--- /dev/null
+++ b/bundles/go-away/files/config/snippets/networks-private.yml
@@ -0,0 +1,22 @@
+networks:
+  localhost:
+    # localhost and loopback addresses
+    - prefixes:
+      - "127.0.0.0/8"
+      - "::1/128"
+  private:
+    # Private network CIDR blocks
+    - prefixes:
+        # private networks
+        - "10.0.0.0/8"
+        - "172.16.0.0/12"
+        - "192.168.0.0/16"
+        - "fc00::/7"
+        # CGNAT
+        - "100.64.0.0/10"
+
+conditions:
+  is-network-localhost:
+    - &is-network-localhost 'remoteAddress.network("localhost")'
+  is-network-private:
+    - &is-network-private 'remoteAddress.network("private")'
\ No newline at end of file