sa-updates/20_head_tests.cf

   1 # SpamAssassin rules file: header tests
   2 #
   3 # Please don't modify this file as your changes will be overwritten with
   4 # the next update. Use /etc/mail/spamassassin/local.cf instead.
   5 # See 'perldoc Mail::SpamAssassin::Conf' for details.
   6 #
   7 # <@LICENSE>
   8 # Licensed to the Apache Software Foundation (ASF) under one or more
   9 # contributor license agreements.  See the NOTICE file distributed with
  10 # this work for additional information regarding copyright ownership.
  11 # The ASF licenses this file to you under the Apache License, Version 2.0
  12 # (the "License"); you may not use this file except in compliance with
  13 # the License.  You may obtain a copy of the License at:
  14 #
  15 #     http://www.apache.org/licenses/LICENSE-2.0
  16 #
  17 # Unless required by applicable law or agreed to in writing, software
  18 # distributed under the License is distributed on an "AS IS" BASIS,
  19 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
  20 # See the License for the specific language governing permissions and
  21 # limitations under the License.
  22 # </@LICENSE>
  23 #
  24 ###########################################################################
  25
  26 require_version 4.000000
  27
  28 ###########################################################################
  29
  30 # partial messages; currently-theoretical attack
  31 # unsurprisingly this hits 0/0 right now.
  32 header FRAGMENTED_MESSAGE       Content-Type =~ /\bmessage\/partial/i
  33 describe FRAGMENTED_MESSAGE     Partial message
  34 tflags FRAGMENTED_MESSAGE       userconf
  35
  36 ###########################################################################
  37
  38 header FROM_BLANK_NAME          From =~ /(?:\s|^)"" <\S+>/i
  39 describe FROM_BLANK_NAME        From: contains empty name
  40
  41 ###########################################################################
  42 # numeric address rules, these are written to avoid overlap with each other
  43
  44 header __FROM_ENDS_IN_NUMS      From:addr =~ /\D\d{8,}\@/i
  45
  46 header FROM_STARTS_WITH_NUMS    From:addr =~ /^\d{3,50}[^0-9\@]/
  47 describe FROM_STARTS_WITH_NUMS  From: starts with several numbers
  48
  49 # don't match US/Canada phone numbers: 10 digits optionally preceded by a "1"
  50 header __FROM_ALL_NUMS          From:addr =~ /^(?:\d{1,9}|[02-9]\d{10}|\d{12,})@/
  51
  52 ###########################################################################
  53
  54 header FROM_OFFERS              From:addr =~ /\@\S*offers(?![eo]n\b)/i
  55 describe FROM_OFFERS            From address is "at something-offers"
  56
  57 header FROM_NO_USER             From =~ /(?:^\@|<\@| \@[^\)<]*$|<>)/ [if-unset: unset@unset.unset]
  58 describe FROM_NO_USER           From: has no local-part before @ sign
  59
  60 # also 100% valid
  61 # bug 6149: avoid common .jp false positives
  62 header __PLING_QUERY            Subject =~ /\?.*!|!.*\?/
  63 meta PLING_QUERY                (__PLING_QUERY && !__ISO_2022_JP_DELIM)
  64 describe PLING_QUERY            Subject has exclamation mark and question mark
  65
  66 # A common spam idiosyncrasy
  67 describe        FROMSPACE       Idiosyncratic "From" header format
  68 header          FROMSPACE       From:raw =~ /^\s?\"\s/
  69
  70
  71
  72 header MSGID_SPAM_CAPS          Message-ID =~ /^\s*<?[A-Z]+\@(?!(?:mailcity|whowhere)\.com)/
  73 describe MSGID_SPAM_CAPS        Spam tool Message-Id: (caps variant)
  74
  75 header MSGID_SPAM_LETTERS       Message-Id =~ /<[a-z]{5,}\@(?:\S+\.)+\S+>/
  76 describe MSGID_SPAM_LETTERS     Spam tool Message-Id: (letters variant)
  77
  78
  79
  80 # negative lookahead exempts this MUA from circa 1997-2000
  81 # X-Mailer: Microsoft Outlook Express 4.71.1712.3
  82 # Message-ID: <01bd45da$2649cdc0$LocalHost@andrew>
  83 header __MSGID_DOLLARS_OK       MESSAGEID =~ /<[0-9a-f]{4,}\$[0-9a-f]{4,}\$[0-9a-f]{4,}\@\S+>/
  84 header __MSGID_DOLLARS_MAYBE    MESSAGEID =~ /<\w{4,}\$\w{4,}\$(?!localhost)\w{4,}\@\S+>/i
  85 meta MSGID_DOLLARS_RANDOM       __MSGID_DOLLARS_MAYBE && !__MSGID_DOLLARS_OK
  86
  87 # bit of a ratware rule, but catches a bit more than just the one ratware
  88 header __MSGID_RANDY            Message-ID =~ /<[a-z\d][a-z\d\$-]{10,29}[a-z\d]\@[a-z\d][a-z\d.]{3,12}[a-z\d]>/
  89 # heuristic to eliminate most good Message-ID formats
  90 header __MSGID_OK_HEX           Message-ID =~ /\b[a-f\d]{8}\b/
  91 header __MSGID_OK_DIGITS        Message-ID =~ /\d{10}/
  92 header __MSGID_OK_HOST          Message-ID =~ /\@(?:\D{2,}|(?:\d{1,3}\.){3}\d{1,3})>/
  93 meta MSGID_RANDY        (__MSGID_RANDY && !(__MSGID_OK_HEX || __MSGID_OK_DIGITS || __MSGID_OK_HOST))
  94 describe MSGID_RANDY            Message-Id has pattern used in spam
  95
  96 # bug 3395
  97 header MSGID_YAHOO_CAPS         Message-ID =~ /<[A-Z]+\@yahoo.com>/
  98 describe MSGID_YAHOO_CAPS       Message-ID has ALLCAPS@yahoo.com
  99
 100 ###########################################################################
 101
 102 header   __AT_AOL_MSGID         MESSAGEID =~ /\@aol\.com\b/i
 103 header   __FROM_AOL_COM         From =~ /\@aol\.com\b/i
 104 meta     FORGED_MSGID_AOL       (__AT_AOL_MSGID && !__FROM_AOL_COM)
 105 describe FORGED_MSGID_AOL       Message-ID is forged, (aol.com)
 106
 107 header   __AT_EXCITE_MSGID      MESSAGEID =~ /\@excite\.com\b/i
 108 header   __MY_RCVD_EXCITE       Received =~ /\.excite\.com\b/i
 109 meta     FORGED_MSGID_EXCITE    (__AT_EXCITE_MSGID && !__MY_RCVD_EXCITE)
 110 describe FORGED_MSGID_EXCITE    Message-ID is forged, (excite.com)
 111
 112 header   __AT_HOTMAIL_MSGID     MESSAGEID =~ /\@hotmail\.com\b/i
 113 header   __FROM_HOTMAIL_COM     From =~ /\@hotmail\.com\b/i
 114 meta     FORGED_MSGID_HOTMAIL   (__AT_HOTMAIL_MSGID && (!__FROM_HOTMAIL_COM && !__FROM_MSN_COM && !__FROM_YAHOO_COM))
 115 describe FORGED_MSGID_HOTMAIL   Message-ID is forged, (hotmail.com)
 116
 117 header   __AT_MSN_MSGID         MESSAGEID =~ /\@msn\.com\b/i
 118 header   __FROM_MSN_COM         From =~ /\@msn\.com\b/i
 119 meta     FORGED_MSGID_MSN       (__AT_MSN_MSGID && (!__FROM_MSN_COM && !__FROM_HOTMAIL_COM && !__FROM_YAHOO_COM))
 120 describe FORGED_MSGID_MSN       Message-ID is forged, (msn.com)
 121
 122 header   __AT_YAHOO_MSGID       MESSAGEID =~ /\@yahoo\.com\b/i
 123 header   __FROM_YAHOO_COM       From =~ /\@yahoo\.com\b/i
 124 meta     FORGED_MSGID_YAHOO     (__AT_YAHOO_MSGID && !__FROM_YAHOO_COM)
 125 describe FORGED_MSGID_YAHOO     Message-ID is forged, (yahoo.com)
 126
 127 ###########################################################################
 128
 129 header __MSGID_BEFORE_RECEIVED  ALL =~ /^Message-Id:.*?^Received:/msi
 130 header __MSGID_BEFORE_OKAY      Message-Id =~ /\@[a-z0-9.-]+\.(?:yahoo|wanadoo)(?:\.[a-z]{2,3}){1,2}>/
 131
 132 meta MSGID_FROM_MTA_HEADER      (__MSGID_BEFORE_RECEIVED && !__MSGID_BEFORE_OKAY && !__FROM_HOTMAIL_COM)
 133 describe MSGID_FROM_MTA_HEADER  Message-Id was added by a relay
 134
 135
 136
 137 header MSGID_SHORT              MESSAGEID =~ /^.{1,15}$|<.{0,4}\@/
 138 describe MSGID_SHORT            Message-ID is unusually short
 139
 140 #DEMOTED TO SANDBOX - 2012-03-21
 141 #header MSGID_MULTIPLE_AT       MESSAGEID =~ /<[^>]*\@[^>]*\@/
 142 #describe MSGID_MULTIPLE_AT     Message-ID contains multiple '@' characters
 143
 144 ###########################################################################
 145
 146 header DATE_SPAMWARE_Y2K        Date =~ /^[A-Z][a-z]{2}, \d\d [A-Z][a-z]{2} [0-6]\d \d\d:\d\d:\d\d [A-Z]{3}$/
 147 describe DATE_SPAMWARE_Y2K      Date header uses unusual Y2K formatting
 148
 149 # as noted on the dev@ list, ":60" is valid for seconds when there's a leap
 150 # second (12/31/2005 for instance), so let's accept that as valid.  ISO 8601
 151 # apparently allows for it.
 152 # there were a few whitespace issues in the original RE, and I wanted to avoid my
 153 # two common, but yes invalid, date headers.  specifically / \(GMT\)$/ and
 154 # / 0000 GMT$/.  dos has / "GMT"$/ - tvd
 155 # 2.229   2.7267   0.0517    0.981   0.86    0.00  INVALID_DATE
 156 # 2.263   2.7486   0.1368    0.953   0.78    0.00  INVALID_DATE_OLD
 157 #
 158 # WRT the tests, remember that ok and fail are reversed -- so valid dates
 159 # should be "fail" and invalid dates should be "ok".
 160 header INVALID_DATE             Date !~ /^\s*(?:(?i:Mon|Tue|Wed|Thu|Fri|Sat|Sun),\s)?\s*(?:[12]\d|3[01]|0?[1-9])\s+(?i:Jan|Feb|Ma[ry]|Apr|Ju[nl]|Aug|Sep|Oct|Nov|Dec)\s+(?:19[7-9]\d|2\d{3})\s+(?:[01]?\d|2[0-3])\:[0-5]\d(?::(?:[0-5]\d|60))?(?:\s+[AP]M)?(?:\s+(?:[+-][0-9]{4}|UT|[A-Z]{2,3}T|0000 GMT|"GMT"))?(?:\s*\(.*\))?\s*$/ [if-unset: Wed, 31 Jul 2002 16:41:57 +0200]
 161 describe INVALID_DATE           Invalid Date: header (not RFC 2822)
 162 test INVALID_DATE fail    Sat, 31 Dec 2005 23:59:60 -0500
 163 test INVALID_DATE fail    Wed, 31 Jul 2002 16:41:57 +0200
 164 test INVALID_DATE fail    Sat, 31 Dec 2005 23:00:00
 165 test INVALID_DATE ok      Sat, 31 Dec 2005 24:00:00 -0500
 166 test INVALID_DATE ok      Thurs, 31 Jul 2002 16:41:57 +0200
 167
 168 # allow +1300, NZ timezone
 169 header INVALID_DATE_TZ_ABSURD   Date =~ /[-+](?!(?:0\d|1[0-4])(?:[03]0|[14]5))\d{4}$/
 170 describe INVALID_DATE_TZ_ABSURD Invalid Date: header (timezone does not exist)
 171
 172 header INVALID_TZ_CST           ALL =~ /[+-]\d\d[30]0(?<!-0600|-0500|\+0800|\+0930|\+1030)\s+(?:\bCST\b|\(CST\))/
 173 describe INVALID_TZ_CST         Invalid date in header (wrong CST timezone)
 174
 175 header INVALID_TZ_EST           ALL =~ /[+-]\d\d[30]0(?<!-0500|-0300|\+1000|\+1100)\s+(?:\bEST\b|\(EST\))/
 176 describe INVALID_TZ_EST         Invalid date in header (wrong EST timezone)
 177
 178
 179 ###########################################################################
 180 # MIME encoding with spam characteristics
 181
 182 ifplugin Mail::SpamAssassin::Plugin::HeaderEval
 183 meta __SUBJECT_NEEDS_MIME       __SUBJ_ILLEGAL_CHARS
 184 endif
 185
 186 header __SUBJECT_ENCODED_QP     Subject:raw =~ /=\?\S+\?Q\?/i
 187 header __SUBJECT_ENCODED_B64    Subject:raw =~ /=\?\S+\?B\?/i
 188
 189
 190
 191 header __FROM_NEEDS_MIME        From:name:raw =~ /[\x00-\x08\x0b\x0c\x0e-\x1f\x7f-\xff]/
 192 header __FROM_NEEDS_MIME2       From:name =~ /[\x00-\x08\x0b\x0c\x0e-\x1f\x7f-\xff]/
 193 header __FROM_ENCODED_QP        From:raw =~ /=\?\S+\?Q\?/i
 194 header __FROM_ENCODED_B64       From:raw =~ /=\?\S+\?B\?/i
 195
 196
 197 meta FROM_EXCESS_BASE64         __FROM_ENCODED_B64 && !__FROM_NEEDS_MIME2
 198 describe FROM_EXCESS_BASE64     From: base64 encoded unnecessarily
 199
 200
 201 ###########################################################################
 202 # ADV tags in various languages
 203
 204 header ENGLISH_UCE_SUBJECT      Subject =~ /^[^0-9a-z]*adv(?:ert)?\b/i
 205 describe ENGLISH_UCE_SUBJECT    Subject contains an English UCE tag
 206
 207 # alan premselaar <alien@12inch.com>, see SpamAssassin-talk list 2003-03
 208 # quinlan: 2003-03-23 here are more generic Japanese iso-2022-jp codes
 209 # ("not yet acceptance" or "email") + "announcement"
 210 # FWIW, according to Peter Evans, this should be sufficient to catch the
 211 # UCE tag and a common attempt at evasion (using the "sue" instead of
 212 # "mi" Chinese character).  2006-10-12: updated by bug 4021.
 213 header JAPANESE_UCE_SUBJECT     Subject =~ /\e\$B.*(?:L\$>5Bz|EE;R%a!<%k)(?:8x|9-)9p/
 214 describe JAPANESE_UCE_SUBJECT   Subject contains a Japanese UCE tag
 215
 216 # check body for "shou nin daku kou koku" UCE tag (bug 4021)
 217 body __JAPANESE_UCE_BODY        /(?:L\$>5Bz|EE;R%a!<%k)(?:8x|9-)9p/
 218
 219 meta JAPANESE_UCE_BODY (__ISO_2022_JP_DELIM && __JAPANESE_UCE_BODY)
 220 describe JAPANESE_UCE_BODY      Body contains Japanese UCE tag
 221
 222 # quinlan: "advertisement" in Russian KOI8-R
 223 # (no longer common, but worth noting in future)
 224 #header RUSSIAN_UCE_SUBJECT     Subject =~ /\xf0\xe5\xea\xeb\xe0\xec\xf3/
 225 #describe RUSSIAN_UCE_SUBJECT   Subject contains a Russian UCE tag
 226
 227 # Korean UCE Subject: lines are usually 8-bit, but are occasionally encoded
 228 # with quoted-printable or base64.
 229 #
 230 # \xbc\xba\xc0\xce means "adult"
 231 # \xb1\xa4\xb0\xed means "advertisement"
 232 # \xc1\xa4\xba\xb8 means "information"
 233 # \xc8\xab\xba\xb8 means "publicity"
 234 #
 235 # Each two byte sequence is one Korean letter; the spaces and periods are
 236 # sometimes used to obscure the words.  \xb1\xa4\xb0\xed is the most common
 237 # tag and is sometimes very obscured so we look harder.
 238 #
 239 header KOREAN_UCE_SUBJECT       Subject =~ /[({[<][. ]*(?-i:\xbc\xba[. ]*\xc0\xce[. ]*)?(?-i:\xb1\xa4(?:[. ]*|[\x00-\x7f]{0,3})\xb0\xed|\xc1\xa4[. ]*\xba\xb8|\xc8\xab[. ]*\xba\xb8)[. ]*[)}\]>]/
 240 describe KOREAN_UCE_SUBJECT     Subject: contains Korean unsolicited email tag
 241
 242 ###########################################################################
 243
 244 # two reliable signatures
 245 header __DOUBLE_IP_SPAM_1       Received =~ /from \[\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\] by \d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3} with/
 246 header __DOUBLE_IP_SPAM_2       Received =~ /from\s+\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}\s+by\s+\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3};/
 247 # loose match
 248 header __DOUBLE_IP_LOOSE        Received =~ /(?:\b(?:from|by)\b.{1,4}\b\d{1,3}[._-]\d{1,3}[._-]\d{1,3}[._-]\d{1,3}(?<!127\.0\.0\.1)\b.{0,4}){2}/i
 249 # spam signature
 250 meta RCVD_DOUBLE_IP_SPAM        (__DOUBLE_IP_SPAM_1 || __DOUBLE_IP_SPAM_2)
 251 describe RCVD_DOUBLE_IP_SPAM    Bulk email fingerprint (double IP) found
 252 # other matches
 253 meta RCVD_DOUBLE_IP_LOOSE       (__DOUBLE_IP_LOOSE && !RCVD_DOUBLE_IP_SPAM)
 254 describe RCVD_DOUBLE_IP_LOOSE   Received: by and from look like IP addresses
 255
 256 header FORGED_TELESP_RCVD       Received =~ /\.(?!br).. \(\d+-\d+-\d+-\d+\.dsl\.telesp\.net\.br /
 257 describe FORGED_TELESP_RCVD     Contains forged hostname for a DSL IP in Brazil
 258
 259 # forgery meta-rules: more reliable than their inputs
 260 meta CONFIRMED_FORGED           (__FORGED_RCVD_TRAIL && (__FORGED_AOL_RCVD || __FORGED_HOTMAIL_RCVD || __FORGED_EUDORAMAIL_RCVD || FORGED_YAHOO_RCVD || __FORGED_JUNO_RCVD || FORGED_GMAIL_RCVD))
 261 describe CONFIRMED_FORGED       Received headers are forged
 262
 263 meta MULTI_FORGED               ((__FORGED_AOL_RCVD + __FORGED_HOTMAIL_RCVD + __FORGED_EUDORAMAIL_RCVD + FORGED_YAHOO_RCVD + __FORGED_JUNO_RCVD + FORGED_GMAIL_RCVD) > 1)
 264 describe MULTI_FORGED           Received headers indicate multiple forgeries
 265
 266 header NONEXISTENT_CHARSET      Content-Type =~ /charset=.?DEFAULT/
 267 describe NONEXISTENT_CHARSET    Character set doesn't exist
 268
 269 header __HAS_MESSAGE_ID         exists:Message-Id
 270 priority __HAS_MESSAGE_ID       -2000 # Bug 8078
 271 meta MISSING_MID                !__HAS_MESSAGE_ID
 272 describe MISSING_MID            Missing Message-Id: header
 273
 274 header __HAS_DATE               exists:Date
 275 priority __HAS_DATE             -2000 # Bug 8078
 276 meta MISSING_DATE               !__HAS_DATE
 277 describe MISSING_DATE           Missing Date: header
 278
 279 header __HAS_SUBJECT            exists:Subject
 280 priority __HAS_SUBJECT          -2000 # Bug 8078
 281 meta MISSING_SUBJECT            !__HAS_SUBJECT
 282 describe MISSING_SUBJECT        Missing Subject: header
 283
 284 # bug 6353
 285 header __HAS_FROM               exists:From
 286 priority __HAS_FROM             -2000 # Bug 8078
 287 meta MISSING_FROM               !__HAS_FROM
 288 describe MISSING_FROM           Missing From: header
 289
 290 # bug 6149: avoid common .jp false positives
 291 header __GAPPY_SUBJECT          Subject =~ /\b(?:[a-z]([-_. =~\/:,*!\@\#\$\%\^&+;\"\'<>\\])\1{0,2}){4}/i
 292 meta GAPPY_SUBJECT              (__GAPPY_SUBJECT && !__ISO_2022_JP_DELIM)
 293 describe GAPPY_SUBJECT          Subject: contains G.a.p.p.y-T.e.x.t
 294
 295 ### header existence tests (description is added automatically)
 296
 297 # X-Fix example: NTMail fixed non RFC822 compliant EMail message
 298 #
 299 # X-PMFLAGS is all caps
 300 #
 301 # Headers that seem to only be used by a single spamming software and
 302 # are found together in the same message:
 303 # 1. X-MailingID and X-ServerHost
 304 # 2. X-Stormpost-To and X-List-Unsubscribe
 305 #
 306 # not spammish: X-EM-Registration, X-EM-Version, X-Antiabuse, X-List-Host,
 307 # X-Message-Id
 308 # bad FP rate: Comment, Date-warning
 309
 310 header PREVENT_NONDELIVERY      exists:Prevent-NonDelivery-Report
 311 describe PREVENT_NONDELIVERY    Message has Prevent-NonDelivery-Report header
 312
 313 header X_IP                     exists:X-IP
 314 describe X_IP                   Message has X-IP header
 315
 316 header   __HAS_MIMEOLE          exists:X-MimeOLE
 317 header   __HAS_MSMAIL_PRI       exists:X-MSMail-Priority
 318 header   __HAS_SQUIRRELMAIL_IN_MAILER   X-Mailer =~ /SquirrelMail\b/
 319 # Ever growing Office version list without X-MimeOLE, bug 6346, 7122, 7463.
 320 header   __HAS_OFFICE1214_IN_MAILER     X-Mailer =~ /^Microsoft (?:Office )?Outlook 1[2456]\.0/
 321 # CGP MAPI module fingerprint, to protect from MISSING_MIMEOLE
 322 header   __HAS_CGP_MAPI_IN_MAILER X-Mailer =~ /CommuniGate Pro MAPI/
 323 meta     MISSING_MIMEOLE        (__HAS_MSMAIL_PRI && !__HAS_MIMEOLE && !__HAS_SQUIRRELMAIL_IN_MAILER && !__HAS_OFFICE1214_IN_MAILER && !__HAS_CGP_MAPI_IN_MAILER && !__HDR_RCVD_TONLINEDE && !__MIME_BASE64 && !__DKIM_EXISTS)
 324 describe MISSING_MIMEOLE        Message has X-MSMail-Priority, but no X-MimeOLE
 325
 326 header __HAS_X_MAILER           exists:X-Mailer
 327
 328 header __IS_EXCH                X-MimeOLE =~ /Produced By Microsoft Exchange V/
 329
 330 header SUBJ_AS_SEEN             Subject =~ /\bAs Seen/i
 331 describe SUBJ_AS_SEEN           Subject contains "As Seen"
 332
 333 header SUBJ_DOLLARS             Subject =~ /^\$[0-9.,]+\b/
 334 describe SUBJ_DOLLARS           Subject starts with dollar amount
 335
 336
 337
 338
 339
 340
 341 #DISABLING DUE TO POOR S/O 2012-09-27
 342 #header SUBJ_YOUR_DEBT          Subject =~ /Your (?:Bills|Debt|Credit)/i
 343 #describe SUBJ_YOUR_DEBT                Subject contains "Your Bills" or similar
 344
 345 header SUBJ_YOUR_FAMILY         Subject =~ /Your Family/i
 346 describe SUBJ_YOUR_FAMILY       Subject contains "Your Family"
 347
 348
 349 # the real services never HELO as 'foo.com', instead 'mail.foo.com' or
 350 # something like that.  Note: be careful when expanding this... legit dotcom
 351 # HELOers include: hotmail.com, drizzle.com, lockergnome.com.
 352 header RCVD_FAKE_HELO_DOTCOM    Received =~ /^from (?:msn|yahoo|yourwebsite|lycos|excite|cs|aol|localhost|koreanmail|allexecs|mydomain|juno|eudoramail|compuserve|desertmail|excite|caramail)\.com \(/m
 353 describe RCVD_FAKE_HELO_DOTCOM  Received contains a faked HELO hostname
 354
 355 header SUBJECT_DIET             Subject =~ /\bLose .*(?:pounds|lbs|weight)/i
 356 describe SUBJECT_DIET           Subject talks about losing pounds
 357
 358
 359 # MIME boundary tests; spam tools use distinctive patterns.
 360 header MIME_BOUND_DD_DIGITS     Content-Type =~ /boundary=\"--\d+\"/
 361 describe MIME_BOUND_DD_DIGITS   Spam tool pattern in MIME boundary
 362 header MIME_BOUND_DIGITS_15     Content-Type =~ /boundary=\"\d{15,}\"/
 363 describe MIME_BOUND_DIGITS_15   Spam tool pattern in MIME boundary
 364 header MIME_BOUND_MANY_HEX      Content-Type =~ /boundary="[\da-f]{8}(?:-[\da-f]{4}){3}-[\da-f]{12}"/
 365 describe MIME_BOUND_MANY_HEX    Spam tool pattern in MIME boundary
 366
 367 # note: the first alternation is anchored for speed
 368 header TO_MALFORMED             To !~ /(?:^|[^\S"])(?:(?:\"[^\"]+\"|\S+)\@\S+\.\S+|^\s*.+:\s*;|^\s*\"[^\"]+\":\s*;|^\s*\([^\)]*\)\s*$|<\S+(?:\!\S+){1,}>|^\s*$)/ [if-unset: unset@unset.unset]
 369 describe TO_MALFORMED           To: has a malformed address
 370
 371 header __CD                     exists:Content-Disposition
 372 header __CT                     exists:Content-Type
 373 header __CTE                    exists:Content-Transfer-Encoding
 374 header __MIME_VERSION           exists:MIME-Version
 375 header __CT_TEXT_PLAIN          Content-Type =~ /^text\/plain\b/i
 376 meta MIME_HEADER_CTYPE_ONLY     (!__CD && !__CTE && __CT && !__MIME_VERSION && !__CT_TEXT_PLAIN)
 377 describe MIME_HEADER_CTYPE_ONLY 'Content-Type' found without required MIME headers
 378
 379 header WITH_LC_SMTP             Received =~ /\swith\ssmtp;\s/
 380 describe WITH_LC_SMTP           Received line contains spam-sign (lowercase smtp)
 381
 382
 383 header SUBJ_BUY                 Subject =~ /^buy/i
 384 describe SUBJ_BUY               Subject line starts with Buy or Buying
 385
 386 # seems to be ratware
 387 header RCVD_AM_PM               Received =~ /; [A-Z][a-z][a-z], \d{1,2} \d{4} \d{1,2}:\d\d:\d\d [AP]M [+-]\d{4}/
 388 describe RCVD_AM_PM             Received headers forged (AM/PM)
 389
 390 header __USER_AGENT_MSN         X-Mailer =~ /^MSN Explorer /
 391
 392 # host no longer exists according to administrator
 393 header FAKE_OUTBLAZE_RCVD       Received =~ /\.mr\.outblaze\.com/
 394 describe FAKE_OUTBLAZE_RCVD     Received header contains faked 'mr.outblaze.com'
 395
 396
 397 # thanks to David Ritz for passing this on
 398 header UNCLOSED_BRACKET         ALL =~ /\[\d+\r?\n/s
 399 describe UNCLOSED_BRACKET       Headers contain an unclosed bracket
 400
 401 header FROM_DOMAIN_NOVOWEL      From =~ /\@\S*[bcdfgjklmnpqrstvwxz]{7}/i
 402 describe FROM_DOMAIN_NOVOWEL    From: domain has series of non-vowel letters
 403 tflags FROM_DOMAIN_NOVOWEL userconf     # lock scores low
 404
 405 header FROM_LOCAL_NOVOWEL       From =~ /[bcdfgjklmnpqrstvwxz]{7}\S*\@/i
 406 describe FROM_LOCAL_NOVOWEL     From: localpart has series of non-vowel letters
 407 tflags FROM_LOCAL_NOVOWEL userconf     # lock scores low
 408
 409 header FROM_LOCAL_HEX           From =~ /[0-9a-f]{11}\S*\@/i
 410 describe FROM_LOCAL_HEX         From: localpart has long hexadecimal sequence
 411
 412 header FROM_LOCAL_DIGITS        From =~ /\d{11}\S*\@/i
 413 describe FROM_LOCAL_DIGITS      From: localpart has long digit sequence
 414
 415 header __TOCC_EXISTS            exists:ToCc
 416
 417 header X_PRIORITY_CC            ALL =~ /^X-Priority:.*?^Cc:/msi
 418 describe X_PRIORITY_CC          Cc: after X-Priority: (bulk email fingerprint)
 419
 420 # catch non-RFC2047 compliant messages
 421 # Apple Mail has a bug where headers will have whitespace around the encoded
 422 # text, so try to ignore that
 423 header BAD_ENC_HEADER           ALL:raw =~ /=\?[^?\s]+\?[^?\s]\?\s*[^?]+\s(?!\?=)/
 424 describe BAD_ENC_HEADER         Message has bad MIME encoding in the header
 425
 426
 427 header __ML1  Precedence =~ m{\b(?:list|bulk)\b}i
 428 meta   __ML2  __HAS_LIST_ID
 429 header __ML3  exists:List-Post
 430 header __ML4  exists:Mailing-List
 431 header __ML5  Return-Path:addr =~ m{^(?:[^\@]+-(?:request|bounces|admin|owner)|owner-[^\@]+)(?:\@|\z)}i
 432 meta     __VIA_ML  __ML1 || __ML2 || __ML3 || __ML4 || __ML5
 433 describe __VIA_ML  Mail from a mailing list
 434
 435
 436 # some clueless mailing lists (like zmailer with an RFC822TABS option on)
 437 # are replacing a leading space by a TAB in header fields From, To,
 438 # Cc, Date (Bug 6429)
 439 header   __ML_TURNS_SP_TO_TAB  Received =~ /\(ORCPT <rfc822;/
 440 describe __ML_TURNS_SP_TO_TAB  A mailing list changing a space to a TAB
 441
 442
 443 # must keep it in sync with https://www.iana.org/assignments/ipv4-address-space/
 444 header RCVD_ILLEGAL_IP          X-Spam-Relays-Untrusted =~ / (?:by|ip)=(?=\d+\.\d+\.\d+\.\d+ )(?:(?:0|2(?:2[4-9]|[3-5]\d)|192\.0\.2|198\.51\.100|203\.0\.113)\.|(?:\d+\.){0,3}(?!(?:2(?:[0-4]\d|5[0-5])|[01]?\d\d?)\b))/
 445 describe RCVD_ILLEGAL_IP        Received: contains illegal IP address
 446
 447
 448 ###########################################################################
 449
 450 ifplugin Mail::SpamAssassin::Plugin::HeaderEval
 451
 452 header __FORGED_AOL_RCVD                eval:check_for_fake_aol_relay_in_rcvd()
 453
 454 header CHARSET_FARAWAY_HEADER   eval:check_for_faraway_charset_in_headers()
 455 describe CHARSET_FARAWAY_HEADER A foreign language charset used in headers
 456 tflags CHARSET_FARAWAY_HEADER   userconf
 457
 458     ###################################################################
 459
 460 # illegal characters that should be MIME encoded
 461 # might want to exempt users using languages that don't use Latin
 462 # alphabets, but do it in the eval
 463
 464 # Will FP without 4.0 and UTF-8 support
 465 if (version >= 4.000000)
 466   header __SUBJ_ILLEGAL_CHARS   eval:check_illegal_chars('Subject','0.00','2')
 467   meta SUBJ_ILLEGAL_CHARS         (__SUBJ_ILLEGAL_CHARS && !__FROM_YAHOO_COM)
 468   header FROM_ILLEGAL_CHARS     eval:check_illegal_chars('From','0.20','2')
 469   header __HEAD_ILLEGAL_CHARS   eval:check_illegal_chars('ALL','0.010','2')
 470   meta HEAD_ILLEGAL_CHARS       __HEAD_ILLEGAL_CHARS && !__SUBJ_ILLEGAL_CHARS && !FROM_ILLEGAL_CHARS
 471 endif
 472 if (version < 4.000000)
 473   meta __SUBJ_ILLEGAL_CHARS 0
 474   meta SUBJ_ILLEGAL_CHARS 0
 475   meta FROM_ILLEGAL_CHARS 0
 476   meta __HEAD_ILLEGAL_CHARS 0
 477   meta HEAD_ILLEGAL_CHARS 0
 478 endif
 479
 480 describe SUBJ_ILLEGAL_CHARS     Subject: has too many raw illegal characters
 481 describe FROM_ILLEGAL_CHARS     From: has too many raw illegal characters
 482 describe HEAD_ILLEGAL_CHARS     Headers have too many raw illegal characters
 483
 484     ###################################################################
 485
 486 # a forged Hotmail message; host HELO'd as hotmail.com, but it wasn't
 487 header __FORGED_HOTMAIL_RCVD    eval:check_for_forged_hotmail_received_headers()
 488
 489 # this, by comparison is more common: from was @hotmail.com, but it wasn't
 490 header FORGED_HOTMAIL_RCVD2     eval:check_for_no_hotmail_received_headers()
 491 describe FORGED_HOTMAIL_RCVD2 hotmail.com 'From' address, but no 'Received:'
 492
 493 header __FORGED_EUDORAMAIL_RCVD eval:check_for_forged_eudoramail_received_headers()
 494
 495 header FORGED_YAHOO_RCVD        eval:check_for_forged_yahoo_received_headers()
 496 describe FORGED_YAHOO_RCVD      'From' yahoo.com does not match 'Received' headers
 497
 498 if (version >= 3.004002)
 499 header FORGED_GMAIL_RCVD        eval:check_for_forged_gmail_received_headers()
 500 describe FORGED_GMAIL_RCVD      'From' gmail.com does not match 'Received' headers
 501 endif
 502
 503 header __FORGED_JUNO_RCVD       eval:check_for_forged_juno_received_headers()
 504
 505
 506
 507 header SORTED_RECIPS            eval:sorted_recipients()
 508 describe SORTED_RECIPS          Recipient list is sorted by address
 509
 510 header SUSPICIOUS_RECIPS        eval:similar_recipients('0.65','undef')
 511 describe SUSPICIOUS_RECIPS      Similar addresses in recipient list
 512
 513 # this is a quite common false positive, as it's legal to remove a To but leave
 514 # a CC. so don't score it high.
 515 header MISSING_HEADERS          eval:check_for_missing_to_header()
 516 describe MISSING_HEADERS        Missing To: header
 517
 518 header DATE_IN_PAST_03_06       eval:check_for_shifted_date('-6', '-3')
 519 describe DATE_IN_PAST_03_06     Date: is 3 to 6 hours before Received: date
 520
 521 header DATE_IN_PAST_06_12       eval:check_for_shifted_date('-12', '-6')
 522 describe DATE_IN_PAST_06_12     Date: is 6 to 12 hours before Received: date
 523
 524 header DATE_IN_PAST_12_24       eval:check_for_shifted_date('-24', '-12')
 525 describe DATE_IN_PAST_12_24     Date: is 12 to 24 hours before Received: date
 526
 527 header DATE_IN_PAST_24_48       eval:check_for_shifted_date('-48', '-24')
 528 describe DATE_IN_PAST_24_48     Date: is 24 to 48 hours before Received: date
 529
 530
 531 header DATE_IN_PAST_96_XX       eval:check_for_shifted_date('undef', '-96')
 532 describe DATE_IN_PAST_96_XX     Date: is 96 hours or more before Received: date
 533
 534 header DATE_IN_FUTURE_03_06     eval:check_for_shifted_date('3', '6')
 535 describe DATE_IN_FUTURE_03_06   Date: is 3 to 6 hours after Received: date
 536
 537 header DATE_IN_FUTURE_06_12     eval:check_for_shifted_date('6', '12')
 538 describe DATE_IN_FUTURE_06_12   Date: is 6 to 12 hours after Received: date
 539
 540 header DATE_IN_FUTURE_12_24     eval:check_for_shifted_date('12', '24')
 541 describe DATE_IN_FUTURE_12_24   Date: is 12 to 24 hours after Received: date
 542
 543 header DATE_IN_FUTURE_24_48     eval:check_for_shifted_date('24', '48')
 544 describe DATE_IN_FUTURE_24_48   Date: is 24 to 48 hours after Received: date
 545
 546 header DATE_IN_FUTURE_48_96     eval:check_for_shifted_date('48', '96')
 547 describe DATE_IN_FUTURE_48_96   Date: is 48 to 96 hours after Received: date
 548
 549 #header DATE_IN_FUTURE_96_XX    eval:check_for_shifted_date('96', 'undef')
 550 meta DATE_IN_FUTURE_96_XX       (0)
 551 describe DATE_IN_FUTURE_96_XX   Date: is 96 hours or more after Received: date
 552
 553 header UNRESOLVED_TEMPLATE      eval:check_unresolved_template()
 554 describe UNRESOLVED_TEMPLATE    Headers contain an unresolved template
 555
 556 header SUBJ_ALL_CAPS            eval:subject_is_all_caps()
 557 describe SUBJ_ALL_CAPS          Subject is all capitals
 558
 559
 560 header LOCALPART_IN_SUBJECT     eval:check_for_to_in_subject('user')
 561 describe LOCALPART_IN_SUBJECT   Local part of To: address appears in Subject
 562
 563 header MSGID_OUTLOOK_INVALID    eval:check_outlook_message_id()
 564 describe MSGID_OUTLOOK_INVALID  Message-Id is fake (in Outlook Express format)
 565
 566 header HEADER_COUNT_CTYPE       eval:check_header_count_range('Content-Type','2','999')
 567 describe HEADER_COUNT_CTYPE     Multiple Content-Type headers found
 568
 569 endif
 570
 571 ###########################################################################
 572
 573 ifplugin Mail::SpamAssassin::Plugin::MIMEEval
 574
 575 # this is also mostly-theoretical, so allow 0 hits
 576 header HEAD_LONG                eval:check_msg_parse_flags('truncated_header')
 577 describe HEAD_LONG              Message headers are very long
 578 tflags HEAD_LONG                userconf
 579
 580 header MISSING_HB_SEP           eval:check_msg_parse_flags('missing_head_body_separator')
 581 describe MISSING_HB_SEP         Missing blank line between message header and body
 582 tflags MISSING_HB_SEP           userconf
 583
 584 endif
 585
 586 ###########################################################################
 587
 588 ifplugin Mail::SpamAssassin::Plugin::RelayEval
 589
 590 header __UNPARSEABLE_RELAY_COUNT        eval:check_relays_unparseable()
 591 tflags __UNPARSEABLE_RELAY_COUNT        userconf
 592
 593 meta     UNPARSEABLE_RELAY      (__UNPARSEABLE_RELAY_COUNT >= 1)
 594 tflags   UNPARSEABLE_RELAY      userconf
 595 describe UNPARSEABLE_RELAY      Informational: message has unparseable relay lines
 596
 597
 598 header RCVD_HELO_IP_MISMATCH    eval:helo_ip_mismatch()
 599 describe RCVD_HELO_IP_MISMATCH  Received: HELO and IP do not match, but should
 600
 601 # not used directly right now due to FPs; but CONFIRMED_FORGED turns it
 602 # into a 1.0 S/O rule anyway, so that's not a problem ;)
 603 # 2.626   3.6340   1.5251    0.704   0.34    1.44  FORGED_RCVD_TRAIL
 604 # 0.956   3.3890   0.0000    1.000   0.98    4.30  CONFIRMED_FORGED
 605 header __FORGED_RCVD_TRAIL      eval:check_for_forged_received_trail()
 606
 607 header NO_RDNS_DOTCOM_HELO      eval:check_for_no_rdns_dotcom_helo()
 608 describe NO_RDNS_DOTCOM_HELO    Host HELO'd as a big ISP, but had no rDNS
 609
 610 endif
 611
 612 ifplugin Mail::SpamAssassin::Plugin::HeaderEval
 613
 614 header __ENV_AND_HDR_FROM_MATCH eval:check_for_matching_env_and_hdr_from()
 615
 616 endif
 617