crs-setup-demo.conf

# ------------------------------------------------------------------------
# OWASP ModSecurity Core Rule Set ver.4.0.0-rc1
# Copyright (c) 2006-2020 Trustwave and contributors. All rights reserved.
# Copyright (c) 2021-2022 Core Rule Set project. All rights reserved.
#
# The OWASP ModSecurity Core Rule Set is distributed under
# Apache Software License (ASL) version 2
# Please see the enclosed LICENSE file for full details.
# ------------------------------------------------------------------------


#
# -- [[ Introduction ]] --------------------------------------------------------
#
# The OWASP ModSecurity Core Rule Set (CRS) is a set of generic attack
# detection rules that provide a base level of protection for any web
# application. They are written for the open source, cross-platform
# ModSecurity Web Application Firewall.
#
# See also:
# https://coreruleset.org/
# https://github.com/coreruleset/coreruleset
# https://owasp.org/www-project-modsecurity-core-rule-set/
#


#
# -- [[ System Requirements ]] -------------------------------------------------
#
# CRS requires ModSecurity version 2.8.0 or above.
# We recommend to always use the newest ModSecurity version.
#
# The configuration directives/settings in this file are used to control
# the OWASP ModSecurity CRS. These settings do **NOT** configure the main
# ModSecurity settings (modsecurity.conf) such as SecRuleEngine,
# SecRequestBodyAccess, SecAuditEngine, SecDebugLog, and XML processing.
#
# The CRS assumes that modsecurity.conf has been loaded. It is bundled with
# ModSecurity. If you don't have it, you can get it from:
# 2.x: https://raw.githubusercontent.com/SpiderLabs/ModSecurity/v2/master/modsecurity.conf-recommended
# 3.x: https://raw.githubusercontent.com/SpiderLabs/ModSecurity/v3/master/modsecurity.conf-recommended
#
# The order of file inclusion in your webserver configuration should always be:
# 1. modsecurity.conf
# 2. crs-setup.conf (this file)
# 3. rules/*.conf (the CRS rule files)
#
# Please refer to the INSTALL file for detailed installation instructions.
#


#
# -- [[ Mode of Operation: Anomaly Scoring vs. Self-Contained ]] ---------------
#
# The CRS can run in two modes:
#
# -- [[ Anomaly Scoring Mode (default) ]] --
# In CRS3, anomaly mode is the default and recommended mode, since it gives the
# most accurate log information and offers the most flexibility in setting your
# blocking policies. It is also called "collaborative detection mode".
# In this mode, each matching rule increases an 'anomaly score'.
# At the conclusion of the inbound rules, and again at the conclusion of the
# outbound rules, the anomaly score is checked, and the blocking evaluation
# rules apply a disruptive action, by default returning an error 403.
#
# -- [[ Self-Contained Mode ]] --
# In this mode, rules apply an action instantly. This was the CRS2 default.
# It can lower resource usage, at the cost of less flexibility in blocking policy
# and less informative audit logs (only the first detected threat is logged).
# Rules inherit the disruptive action that you specify (i.e. deny, drop, etc).
# The first rule that matches will execute this action. In most cases this will
# cause evaluation to stop after the first rule has matched, similar to how many
# IDSs function.
#
# -- [[ Alert Logging Control ]] --
# In the mode configuration, you must also adjust the desired logging options.
# There are three common options for dealing with logging. By default CRS enables
# logging to the webserver error log (or Event viewer) plus detailed logging to
# the ModSecurity audit log (configured under SecAuditLog in modsecurity.conf).
#
# - To log to both error log and ModSecurity audit log file, use: "log,auditlog"
# - To log *only* to the ModSecurity audit log file, use: "nolog,auditlog"
# - To log *only* to the error log file, use: "log,noauditlog"
#
# Examples for the various modes follow.
# You must leave one of the following options enabled.
# Note that you must specify the same line for phase:1 and phase:2.
#

# Default: Anomaly Scoring mode, log to error log, log to ModSecurity audit log
# - By default, offending requests are blocked with an error 403 response.
# - To change the disruptive action, see RESPONSE-999-EXCLUSION-RULES-AFTER-CRS.conf.example
#   and review section 'Changing the Disruptive Action for Anomaly Mode'.
# - In Apache, you can use ErrorDocument to show a friendly error page or
#   perform a redirect: https://httpd.apache.org/docs/2.4/custom-error.html
#
SecDefaultAction "phase:1,log,auditlog,pass"
SecDefaultAction "phase:2,log,auditlog,pass"

# Example: Anomaly Scoring mode, log only to ModSecurity audit log
# - By default, offending requests are blocked with an error 403 response.
# - To change the disruptive action, see RESPONSE-999-EXCLUSION-RULES-AFTER-CRS.conf.example
#   and review section 'Changing the Disruptive Action for Anomaly Mode'.
# - In Apache, you can use ErrorDocument to show a friendly error page or
#   perform a redirect: https://httpd.apache.org/docs/2.4/custom-error.html
#
# SecDefaultAction "phase:1,nolog,auditlog,pass"
# SecDefaultAction "phase:2,nolog,auditlog,pass"

# Example: Self-contained mode, return error 403 on blocking
# - In this configuration the default disruptive action becomes 'deny'. After a
#   rule triggers, it will stop processing the request and return an error 403.
# - You can also use a different error status, such as 404, 406, et cetera.
# - In Apache, you can use ErrorDocument to show a friendly error page or
#   perform a redirect: https://httpd.apache.org/docs/2.4/custom-error.html
#
# SecDefaultAction "phase:1,log,auditlog,deny,status:403"
# SecDefaultAction "phase:2,log,auditlog,deny,status:403"

# Example: Self-contained mode, redirect back to homepage on blocking
# - In this configuration the 'tag' action includes the Host header data in the
#   log. This helps to identify which virtual host triggered the rule (if any).
# - Note that this might cause redirect loops in some situations; for example
#   if a Cookie or User-Agent header is blocked, it will also be blocked when
#   the client subsequently tries to access the homepage. You can also redirect
#   to another custom URL.
# SecDefaultAction "phase:1,log,auditlog,redirect:'http://%{request_headers.host}/',tag:'Host: %{request_headers.host}'"
# SecDefaultAction "phase:2,log,auditlog,redirect:'http://%{request_headers.host}/',tag:'Host: %{request_headers.host}'"


#
# -- [[ Paranoia Level Initialization ]] ---------------------------------------
#
# The Paranoia Level (PL) setting allows you to choose the desired level
# of rule checks that will add to your anomaly scores.
#
# With each paranoia level increase, the CRS enables additional rules
# giving you a higher level of security. However, higher paranoia levels
# also increase the possibility of blocking some legitimate traffic due to
# false alarms (also named false positives or FPs). If you use higher
# paranoia levels, it is likely that you will need to add some exclusion
# rules for certain requests and applications receiving complex input.
#
# - A paranoia level of 1 is default. In this level, most core rules
#   are enabled. PL1 is advised for beginners, installations
#   covering many different sites and applications, and for setups
#   with standard security requirements.
#   At PL1 you should face FPs rarely. If you encounter FPs, please
#   open an issue on the CRS GitHub site and don't forget to attach your
#   complete Audit Log record for the request with the issue.
# - Paranoia level 2 includes many extra rules, for instance enabling
#   many regexp-based SQL and XSS injection protections, and adding
#   extra keywords checked for code injections. PL2 is advised
#   for moderate to experienced users desiring more complete coverage
#   and for installations with elevated security requirements.
#   PL2 comes with some FPs which you need to handle.
# - Paranoia level 3 enables more rules and keyword lists, and tweaks
#   limits on special characters used. PL3 is aimed at users experienced
#   at the handling of FPs and at installations with a high security
#   requirement.
# - Paranoia level 4 further restricts special characters.
#   The highest level is advised for experienced users protecting
#   installations with very high security requirements. Running PL4 will
#   likely produce a very high number of FPs which have to be
#   treated before the site can go productive.
#
# All rules will log their PL to the audit log;
# example: [tag "paranoia-level/2"]. This allows you to deduct from the
# audit log how the WAF behavior is affected by paranoia level.
#
# It is important to also look into the variable
# tx.enforce_bodyproc_urlencoded (Enforce Body Processor URLENCODED)
# defined below. Enabling it closes a possible bypass of CRS.
#
# Uncomment this rule to change the default:
#
#SecAction \
#    "id:900000,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.blocking_paranoia_level=1"


# It is possible to execute rules from a higher paranoia level but not include
# them in the anomaly scoring. This allows you to take a well-tuned system on
# paranoia level 1 and add rules from paranoia level 2 without having to fear
# the new rules would lead to false positives that raise your score above the
# threshold.
# This optional feature is enabled by uncommenting the following rule and
# setting the tx.detection_paranoia_level.
# Technically, rules up to the level defined in tx.detection_paranoia_level
# will be executed, but only the rules up to tx.blocking_paranoia_level affect the
# anomaly scores.
# By default, tx.detection_paranoia_level is set to tx.blocking_paranoia_level.
# tx.detection_paranoia_level must not be lower than tx.blocking_paranoia_level.
#
# Please notice that setting tx.detection_paranoia_level to a higher paranoia
# level results in a performance impact that is equally high as setting
# tx.blocking_paranoia_level to said level.
#
#SecAction \
#    "id:900001,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.detection_paranoia_level=1"


#
# -- [[ Enforce Body Processor URLENCODED ]] -----------------------------------
#
# ModSecurity selects the body processor based on the Content-Type request
# header. But clients are not always setting the Content-Type header for their
# request body payloads. This will leave ModSecurity with limited vision into
# the payload.  The variable tx.enforce_bodyproc_urlencoded lets you force the
# URLENCODED body processor in these situations. This is off by default, as it
# implies a change of the behaviour of ModSecurity beyond CRS (the body
# processor applies to all rules, not only CRS) and because it may lead to
# false positives already on paranoia level 1. However, enabling this variable
# closes a possible bypass of CRS so it should be considered.
#
# Uncomment this rule to change the default:
#
#SecAction \
#    "id:900010,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.enforce_bodyproc_urlencoded=1"


#
# -- [[ Anomaly Scoring Mode Severity Levels ]] --------------------------------
#
# Each rule in the CRS has an associated severity level.
# These are the default scoring points for each severity level.
# These settings will be used to increment the anomaly score if a rule matches.
# You may adjust these points to your liking, but this is usually not needed.
#
# - CRITICAL severity: Anomaly Score of 5.
#       Mostly generated by the application attack rules (93x and 94x files).
# - ERROR severity: Anomaly Score of 4.
#       Generated mostly from outbound leakage rules (95x files).
# - WARNING severity: Anomaly Score of 3.
#       Generated mostly by malicious client rules (91x files).
# - NOTICE severity: Anomaly Score of 2.
#       Generated mostly by the protocol rules (92x files).
#
# In anomaly mode, these scores are cumulative.
# So it's possible for a request to hit multiple rules.
#
# (Note: In this file, we use 'phase:1' to set CRS configuration variables.
# In general, 'phase:request' is used. However, we want to make absolutely sure
# that all configuration variables are set before the CRS rules are processed.)
#
#SecAction \
#    "id:900100,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.critical_anomaly_score=5,\
#    setvar:tx.error_anomaly_score=4,\
#    setvar:tx.warning_anomaly_score=3,\
#    setvar:tx.notice_anomaly_score=2"


#
# -- [[ Anomaly Scoring Mode Blocking Threshold Levels ]] ----------------------
#
# Here, you can specify at which cumulative anomaly score an inbound request,
# or outbound response, gets blocked.
#
# Most detected inbound threats will give a critical score of 5.
# Smaller violations, like violations of protocol/standards, carry lower scores.
#
# [ At default value ]
# If you keep the blocking thresholds at the defaults, the CRS will work
# similarly to previous CRS versions: a single critical rule match will cause
# the request to be blocked and logged.
#
# [ Using higher values ]
# If you want to make the CRS less sensitive, you can increase the blocking
# thresholds, for instance to 7 (which would require multiple rule matches
# before blocking) or 10 (which would require at least two critical alerts - or
# a combination of many lesser alerts), or even higher. However, increasing the
# thresholds might cause some attacks to bypass the CRS rules or your policies.
#
# [ New deployment strategy: Starting high and decreasing ]
# It is a common practice to start a fresh CRS installation with elevated
# anomaly scoring thresholds (>100) and then lower the limits as your
# confidence in the setup grows. You may also look into the Sampling
# Percentage section below for a different strategy to ease into a new
# CRS installation.
#
# [ Anomaly Threshold / Paranoia Level Quadrant ]
#
#     High Anomaly Limit   |   High Anomaly Limit
#     Low Paranoia Level   |   High Paranoia Level
#     -> Fresh Site        |   -> Experimental Site
# ------------------------------------------------------
#     Low Anomaly Limit    |   Low Anomaly Limit
#     Low Paranoia Level   |   High Paranoia Level
#     -> Standard Site     |   -> High Security Site
#
# Uncomment this rule to change the defaults:
#
#SecAction \
#    "id:900110,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.inbound_anomaly_score_threshold=5,\
#    setvar:tx.outbound_anomaly_score_threshold=4"


#
# -- [[ Application Specific Rule Exclusions ]] --------------------------------
#
# CRS 3.x contained exclusion packages to tweak the CRS for use with common
# web applications, lowering the number of false positives.
#
# In CRS 4, these are no longer part of the CRS itself, but they are available
# as "CRS plugins". Some plugins improve support for web applications, and others
# may bring new functionality. Plugins are not installed by default, but can be
# downloaded from the plugin registry:
# https://github.com/coreruleset/plugin-registry
#
# For detailed information about using and installing plugins, please see:
# https://coreruleset.org/docs/concepts/plugins/


#
# -- [[ Anomaly Score Reporting Level ]] ---------------------------------------
#
# When a request is blocked due to the anomaly score meeting or exceeding the
# anomaly threshold then the blocking rule will also report the anomaly score.
# This applies to the separate inbound and outbound anomaly scores.
#
# In phase 5, there are additional rules that can perform additional reporting
# of anomaly scores with a verbosity that depends on the reporting level defined
# below.
#
# By setting the reporting level you control whether you want additional
# reporting beyond the blocking rule or not and, if yes, which requests should
# be covered. The higher the reporting level, the more verbose the reporting is.
#
# There are 6 reporting levels:
#
# 0 - Reporting disabled
# 1 - Reporting for requests with a blocking anomaly score >= a threshold
# 2 - Reporting for requests with a detection anomaly score >= a threshold
# 3 - Reporting for requests with a blocking anomaly score greater than 0
# 4 - Reporting for requests with a detection anomaly score greater than 0
# 5 - Reporting for all requests
#
# Note: Reporting levels 1 and 2 make it possible to differentiate between
# requests that are blocked and requests that are *not* blocked but would have
# been blocked if the blocking PL was equal to detection PL. This may be useful
# for certain FP tuning methodologies, for example moving to a higher PL.
#
# A value of 5 can be useful on platforms where you are interested in logging
# non-scoring requests, yet it is not possible to report this information in
# the request/access log. This applies to Nginx, for example.
#
#SecAction \
#    "id:900115,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.reporting_level=4"


#
# -- [[ Early Anomaly Scoring Mode Blocking ]] ------------------------------
#
# The anomaly scores for the request and the responses are generally summed up
# and evaluated at the end of phase:2 and at the end of phase:4 respectively.
# However, it is possible to enable an early evaluation of these anomaly scores
# at the end of phase:1 and at the end of phase:3.
#
# If a request (or a response) hits the anomaly threshold in this early
# evaluation, then blocking happens immediately (if blocking is enabled) and
# the phase 2 (and phase 4 respectively) will no longer be executed.
#
# Enable the rule 900120 that sets the variable tx.early_blocking to 1 in order
# to enable early blocking. The variable tx.early_blocking is set to 0 by
# default. Early blocking is thus disabled by default.
#
# Please note that early blocking will hide potential alerts from you. This
# means that a payload that would appear in an alert in phase 2 (or phase 4)
# does not get evaluated if the request is being blocked early. So when you
# disabled early blocking again at some point in the future, then new alerts
# from phase 2 might pop up.
SecAction \
    "id:900120,\
    phase:1,\
    pass,\
    t:none,\
    nolog,\
    setvar:tx.early_blocking=1"


#
# -- [[ HTTP Policy Settings ]] ------------------------------------------------
#
# This section defines your policies for the HTTP protocol, such as:
# - allowed HTTP versions, HTTP methods, allowed request Content-Types
# - forbidden file extensions (e.g. .bak, .sql) and request headers (e.g. Proxy)
#
# These variables are used in the following rule files:
# - REQUEST-911-METHOD-ENFORCEMENT.conf
# - REQUEST-920-PROTOCOL-ENFORCEMENT.conf

# HTTP methods that a client is allowed to use.
# Default: GET HEAD POST OPTIONS
# Example: for RESTful APIs, add the following methods: PUT PATCH DELETE
# Example: for WebDAV, add the following methods: CHECKOUT COPY DELETE LOCK
#          MERGE MKACTIVITY MKCOL MOVE PROPFIND PROPPATCH PUT UNLOCK
# Uncomment this rule to change the default.
#SecAction \
#    "id:900200,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.allowed_methods=GET HEAD POST OPTIONS'"

# Content-Types that a client is allowed to send in a request.
# Default: |application/x-www-form-urlencoded| |multipart/form-data| |multipart/related|
# |text/xml| |application/xml| |application/soap+xml| |application/json|
# |application/cloudevents+json| |application/cloudevents-batch+json|
#
# Please note, that the rule where CRS uses this variable (920420) evaluates it with operator
# `@within`, which is case sensitive, but uses t:lowercase. You must add your whole custom
# Content-Type with lowercase.
#
# Bypass Warning: some applications may not rely on the content-type request header in order
# to parse the request body. This could make an attacker able to send malicious URLENCODED/JSON/XML
# payloads without being detected by the WAF. Allowing request content-type that doesn't activate any
# body processor (for example: "text/plain", "application/x-amf", "application/octet-stream", etc..)
# could lead to a WAF bypass. For example, a malicious JSON payload submitted with a "text/plain"
# content type may still be interpreted as JSON by a backend application but would not trigger the
# JSON body parser at the WAF, leading to a bypass.
#
# To prevent blocking request with not allowed content-type by default, you can create an exclusion
# rule that removes rule 920420. For example:
#SecRule REQUEST_HEADERS:Content-Type "@rx ^text/plain" \
#    "id:1234,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    ctl:ruleRemoveById=920420,\
#    chain"
#    SecRule REQUEST_URI "@rx ^/foo/bar" "t:none"
#
# Uncomment this rule to change the default.
#
#SecAction \
#    "id:900220,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.allowed_request_content_type=|application/x-www-form-urlencoded| |multipart/form-data| |multipart/related| |text/xml| |application/xml| |application/soap+xml| |application/json| |application/cloudevents+json| |application/cloudevents-batch+json|'"

# Allowed HTTP versions.
# Default: HTTP/1.0 HTTP/1.1 HTTP/2 HTTP/2.0
# Example for legacy clients: HTTP/0.9 HTTP/1.0 HTTP/1.1 HTTP/2 HTTP/2.0
# Note that some web server versions use 'HTTP/2', some 'HTTP/2.0', so
# we include both version strings by default.
# Uncomment this rule to change the default.
#SecAction \
#    "id:900230,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.allowed_http_versions=HTTP/1.0 HTTP/1.1 HTTP/2 HTTP/2.0'"

# Forbidden file extensions.
# Guards against unintended exposure of development/configuration files.
# Default: .asa/ .asax/ .ascx/ .backup/ .bak/ .bat/ .cdx/ .cer/ .cfg/ .cmd/ .com/ .config/ .conf/ .cs/ .csproj/ .csr/ .dat/ .db/ .dbf/ .dll/ .dos/ .htr/ .htw/ .ida/ .idc/ .idq/ .inc/ .ini/ .key/ .licx/ .lnk/ .log/ .mdb/ .old/ .pass/ .pdb/ .pol/ .printer/ .pwd/ .rdb/ .resources/ .resx/ .sql/ .swp/ .sys/ .vb/ .vbs/ .vbproj/ .vsdisco/ .webinfo/ .xsd/ .xsx/
# Example: .bak/ .config/ .conf/ .db/ .ini/ .log/ .old/ .pass/ .pdb/ .rdb/ .sql/
# Note that .axd was removed due to false positives (see PR 1925).
#
# To additionally guard against configuration/install archive files from being
# accidentally exposed, common archive file extensions can be added to the
# restricted extensions list. An example list of common archive file extensions
# is presented below:
# .7z/ .br/ .bz/ .bz2/ .cab/ .cpio/ .gz/ .img/ .iso/ .jar/ .rar/ .tar/ .tbz2/ .tgz/ .txz/ .xz/ .zip/ .zst/
# (Source: https://en.wikipedia.org/wiki/List_of_archive_formats)
#
# Uncomment this rule to change the default.
#SecAction \
#    "id:900240,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.restricted_extensions=.asa/ .asax/ .ascx/ .backup/ .bak/ .bat/ .cdx/ .cer/ .cfg/ .cmd/ .com/ .config/ .conf/ .cs/ .csproj/ .csr/ .dat/ .db/ .dbf/ .dll/ .dos/ .htr/ .htw/ .ida/ .idc/ .idq/ .inc/ .ini/ .key/ .licx/ .lnk/ .log/ .mdb/ .old/ .pass/ .pdb/ .pol/ .printer/ .pwd/ .rdb/ .resources/ .resx/ .sql/ .swp/ .sys/ .vb/ .vbs/ .vbproj/ .vsdisco/ .webinfo/ .xsd/ .xsx/'"

# Forbidden request headers.
# Header names should be lowercase, enclosed by /slashes/ as delimiters.
# Default: /accept-charset/ /content-encoding/ /proxy/ /lock-token/ /content-range/ /if/
#
# Note: Accept-Charset is a deprecated header that should not be used by clients and
# ignored by servers. It can be used for a response WAF bypass, by asking for a charset
# that the WAF cannot decode.
# Reference: https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Accept-Charset
#
# Note: Content-Encoding is used to list any encodings that have been applied to the
# original payload. It is only used for compression, which isn't supported by CRS by
# default since it blocks newlines and null bytes inside the request body. Most
# compression algorithms require at least null bytes per RFC. Blocking it shouldn't
# break anything and increases security since ModSecurity is incapable of properly
# scanning compressed request bodies.
#
# Note: Blocking Proxy header prevents 'httpoxy' vulnerability: https://httpoxy.org
#
# Note: Blocking the x-http-method-override,x-http-method and x-method-override headers
# prevents attacks as described here: https://www.sidechannel.blog/en/http-method-override-what-it-is-and-how-a-pentester-can-use-it
#
# Uncomment this rule to change the default.
#SecAction \
#    "id:900250,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.restricted_headers=/accept-charset/ /content-encoding/ /proxy/ /lock-token/ /content-range/ /if/ /x-http-method-override/ /x-http-method/ /x-method-override/'"

# Content-Types charsets that a client is allowed to send in a request.
# The content-types are enclosed by |pipes| as delimiters to guarantee exact matches.
# Default: |utf-8| |iso-8859-1| |iso-8859-15| |windows-1252|
# Uncomment this rule to change the default.
#SecAction \
#    "id:900280,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:'tx.allowed_request_content_type_charset=|utf-8| |iso-8859-1| |iso-8859-15| |windows-1252|'"

#
# -- [[ HTTP Argument/Upload Limits ]] -----------------------------------------
#
# Here you can define optional limits on HTTP get/post parameters and uploads.
# This can help to prevent application specific DoS attacks.
#
# These values are checked in REQUEST-920-PROTOCOL-ENFORCEMENT.conf.
# Beware of blocking legitimate traffic when enabling these limits.
#

# Block request if number of arguments is too high
# Default: unlimited
# Example: 255
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900300,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.max_num_args=255"

# Block request if the length of any argument name is too high
# Default: unlimited
# Example: 100
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900310,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.arg_name_length=100"

# Block request if the length of any argument value is too high
# Default: unlimited
# Example: 400
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900320,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.arg_length=400"

# Block request if the total length of all combined arguments is too high
# Default: unlimited
# Example: 64000
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900330,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.total_arg_length=64000"

# Block request if the file size of any individual uploaded file is too high
# Default: unlimited
# Example: 1048576
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900340,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.max_file_size=1048576"

# Block request if the total size of all combined uploaded files is too high
# Default: unlimited
# Example: 1048576
# Uncomment this rule to set a limit.
#SecAction \
#    "id:900350,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.combined_file_sizes=1048576"


#
# -- [[ Easing In / Sampling Percentage ]] -------------------------------------
#
# Adding the Core Rule Set to an existing productive site can lead to false
# positives, unexpected performance issues and other undesired side effects.
#
# It can be beneficial to test the water first by enabling the CRS for a
# limited number of requests only and then, when you have solved the issues (if
# any) and you have confidence in the setup, to raise the ratio of requests
# being sent into the ruleset.
#
# Adjust the percentage of requests that are funnelled into the Core Rules by
# setting TX.sampling_percentage below. The default is 100, meaning that every
# request gets checked by the CRS.  The selection of requests, which are going
# to be checked, is based on a pseudo random number generated by ModSecurity.
#
# If a request is allowed to pass without being checked by the CRS, there is no
# entry in the audit log (for performance reasons), but an error log entry is
# written.  If you want to disable the error log entry, then issue the
# following directive somewhere after the inclusion of the CRS
# (E.g., RESPONSE-999-EXCLUSION-RULES-AFTER-CRS.conf).
#
#SecRuleUpdateActionById 901450 "nolog"
#
# ATTENTION: If this TX.sampling_percentage is below 100, then some of the
# requests will bypass the Core Rules completely and you lose the ability to
# protect your service with ModSecurity.
#
# Uncomment this rule to enable this feature:
#
#SecAction \
#    "id:900400,\
#    phase:1,\
#    pass,\
#    nolog,\
#    setvar:tx.sampling_percentage=100"



#
# -- [[ Check UTF-8 encoding ]] ------------------------------------------------
#
# The CRS can optionally check request contents for invalid UTF-8 encoding.
# We only want to apply this check if UTF-8 encoding is actually used by the
# site; otherwise it will result in false positives.
#
# Uncomment this rule to use this feature:
#
#SecAction \
#    "id:900950,\
#    phase:1,\
#    pass,\
#    t:none,\
#    nolog,\
#    setvar:tx.crs_validate_utf8_encoding=1"


#
# -- [[ Collection timeout ]] --------------------------------------------------
#
# Set the SecCollectionTimeout directive from the ModSecurity default (1 hour)
# to a lower setting which is appropriate to most sites.
# This increases performance by cleaning out stale collection (block) entries.
#
# This value should be greater than or equal to any block durations or timeouts
# set by plugins that make use of ModSecurity's persistent collections (e.g. the
# DoS protection and IP reputation plugins).
#
# Ref: https://github.com/SpiderLabs/ModSecurity/wiki/Reference-Manual-(v2.x)#SecCollectionTimeout

# Please keep this directive uncommented.
# Default: 600 (10 minutes)
SecCollectionTimeout 600


#
# -- [[ End of setup ]] --------------------------------------------------------
#
# The CRS checks the tx.crs_setup_version variable to ensure that the setup
# has been loaded. If you are not planning to use this setup template,
# you must manually set the tx.crs_setup_version variable before including
# the CRS rules/* files.
#
# The variable is a numerical representation of the CRS version number.
# E.g., v3.0.0 is represented as 300.
#
SecAction \
    "id:900990,\
    phase:1,\
    pass,\
    t:none,\
    nolog,\
    setvar:tx.crs_setup_version=400"