Skip to content

Managed rule sets

The OWASP Core Rule Set and EZGH's rule sets for PHP, Node.js and known bad inputs, their modes and paranoia levels, and rule overrides for false positives.

Updated View as Markdown

A managed rule set is a set of rules EZGH writes and keeps up to date. A WAF runs the sets it lists, in that order, after its custom rules.

Set Name Blocks
Core rule set (OWASP CRS) ezgh/core Common web attacks: SQL injection, cross-site scripting, path traversal, remote code and file inclusion, protocol abuse. Rules add to a request’s anomaly score, and the request is blocked when its score is high enough.
PHP ezgh/php Exploits of known vulnerabilities in PHP and PHP applications, such as WordPress and popular plugins.
Node.js ezgh/node Exploits of known vulnerabilities in Node.js and npm server frameworks, such as Next.js.
Known bad inputs ezgh/known-bad-inputs Exploits that work across runtimes, such as Log4Shell JNDI lookups and Spring4Shell.

The PHP, Node.js and Known bad inputs sets block on a single matching rule.

ezgh waf managed-rule-sets list

ListManagedRuleSets lists each set with its description, current version and updatedAt.

Versions

A WAF always runs the latest version of each set. EZGH updates the sets as new vulnerabilities are exploited, and rolls each update out gradually. There’s nothing to upgrade.

Turn a set on

A WAF’s managedRuleSets is an ordered list. Each entry has:

Field
name The set’s name, such as ezgh/core. Each set at most once.
mode block (the default) or log: record matches and let requests through.
paranoiaLevel ezgh/core only: 1 (the default) or 2. Level 2 adds rules that catch more attacks and match more legitimate requests. null for the other sets.
ruleOverrides Up to 100 rules of the set to turn off or log, below.
"managedRuleSets": [
  { "name": "ezgh/core", "mode": "block", "paranoiaLevel": 1, "ruleOverrides": [] },
  { "name": "ezgh/node", "mode": "log", "paranoiaLevel": null, "ruleOverrides": [] }
]

A new WAF runs no managed rule sets until you list them. A WAF’s mode of log overrides every set’s mode.

Rule overrides for false positives

When a managed rule blocks legitimate requests to your application, override that one rule instead of turning the whole set off. Find its ID in the request’s log record (matchedRules, or terminatingRule), then add it to the set’s ruleOverrides:

{
  "name": "ezgh/core",
  "mode": "block",
  "paranoiaLevel": 1,
  "ruleOverrides": [
    { "ruleId": 942100, "action": "off" }
  ]
}
action
off The rule doesn’t run.
log The rule records its matches and doesn’t block. Not for ezgh/core, whose rules can only be turned off.

An ID that isn’t a rule of the set is 400 invalid_request.

Navigation

Type to search…

↑↓ navigate↵ selectEsc close