Skip to main content
ALTCHA®

Classifier rules & scores

This page lists all 53 classifier rules and 22 email (EML) rules, with their scores and triggers, and the formulas that turn rule scores into a score and verdict. For the concepts behind them, see Scoring and Classifier.

Score computation

StepFormula
Rule scoreEach rule reports score: 0 when it didn't trigger, null when it is disabled or doesn't apply (for example, no input for it).
Category scoreSum of the category's non-null rule scores. Text and HIS sums are rounded to 2 decimals. Similarity uses max(SIMILARITY, NEAR_DUPLICATE) + CAMPAIGN instead.
Counted categoriesA category counts if it ran and has at least one enabled rule. Injection, Similarity, Rate limit, and AI count only when their score is > 0, so a clean check or a limit that wasn't exceeded doesn't lower the mean.
Overall scoreMean of the counted category scores, rounded down to 2 decimals (0 when no category counts). Then raised by the floors.
VerdictGOOD for a score < 1, NEUTRAL for a score < 2, BAD for anything higher.
triggeredRulesEvery rule that scored > 0, highest score first. Ties keep the order shown on this page. Each rule appears once.

Categories and when they run:

CategoryResponse fieldRuns when
LocationlocationA country is known (IP geolocation, otherwise the request's timeZone)
DevicedeviceDevice input (request headers) is present
EmailemailAn email address is present
IPipAn IP address is present
TexttextNon-blank text is present (the body's text, otherwise text from the form fields)
InjectioninjectionForm fields, text, or an email address are present to scan
SimilaritysimilarityText is present and at least one similarity rule is enabled
Rate limitrateLimitThe body has rateLimit and RATE_LIMIT is enabled
Human Interaction Signature (HIS)hisPOST /v2/verify only, with an HIS score in the challenge, outside HIS monitor mode, and at least one HIS rule enabled
AIaiThe security group enables classifier.ai, [classifier.ai] has a provider, text is not empty, and at least one AI rule is requested

Floors

After the mean is computed, these conditions raise the overall score to at least a minimum value. Floors only raise the score; they never lower it.

ConditionMinimum scoreResulting verdict
Any injection rule triggered (injection score > 0)2BAD
A similarity match (SIMILARITY or NEAR_DUPLICATE) against a sample with weight ≥ 12BAD
URL_PHISHING triggered2BAD
HIS_INTEGRITY triggered2BAD
HIS_REPLAY triggered, or HIS_AUTOMATION ≥ 1.81at least NEUTRAL
Highest AI answer probability ≥ 0.92BAD
Highest AI answer probability ≥ 0.7 (and below 0.9)1at least NEUTRAL

Enabling and disabling rules

Every rule in the classifier tables below can be disabled. To do so, use the security-group settings classifier.disableRules / classifier.enableRules, or the request body's disableRules / enableRules on POST /v2/classifier (each body list replaces the group's list of the same kind). If a disable list is set, every rule except those runs and the enable list is ignored. Otherwise, if an enable list is set, only those rules run. Otherwise all rules run. Because the lists are replaced separately, a body enableRules has no effect while the group sets classifier.disableRules; send "disableRules": [] with it to clear the group's disable list. Rule names are case-insensitive and can be separated by ,, ;, or whitespace. See Rules.

Email address

RuleScoreTriggerCan be disabled
DISPOSABLE1The domain is on the disposable-email list and not on the trusted-provider listYes
DMARC0.5The domain has no DMARC recordYes
FREE_PROVIDER1 / 0.5The domain is on the free-email list: 1, or 0.5 if it is also on the trusted-provider listYes
MX5The domain has no MX recordYes
PRIVACY_RELAY1The domain is on the email privacy-relay listYes

Device

RuleScoreTriggerCan be disabled
BOT2The request is signed by a verified Web Bot Auth agent. Never triggers on POST /v2/classifier, which has no signed request to check. BOT is also an IP rule (see IP).Yes
ACCEPT_HEADER_MISSING1No Accept header. null when the header is present but there's no User-Agent.Yes
ACCEPT_LANGUAGE_HEADER_MISSING1No Accept-Language header. null when the header is present but there's no User-Agent.Yes
USER_AGENT_HEADER_MISSING1No User-Agent headerYes
USER_AGENT_HEADER_UNCOMMON1The User-Agent isn't on the common User-Agent list. Never triggers while the list is empty, or for a trusted app (classifier.trustedApps). null without a User-Agent.Yes
CLI1The User-Agent is a command-line HTTP client, unless it's a trusted app. null without a User-Agent.Yes
LIBRARY1The User-Agent is an HTTP client library, unless it's a trusted app. null without a User-Agent.Yes

IP

RuleScoreTriggerCan be disabled
BOT2A threat list classifies the IP as a bot, crawler, AI training/search agent, or user-fetch agentYes
HOSTING2The IP is a hosting/data-center IP. On POST /v2/eml, it's ignored when the sending server is authorized by the sender's domain.Yes
MALICIOUS5A threat list classifies the IP as maliciousYes
PROXY0.5The IP is a proxyYes
TOR1The IP is a Tor exit nodeYes
THREAT_SCORE0–10The IP's threat score (the highest of the matching threat-list entries); 0 without oneYes

Location

RuleScoreTriggerCan be disabled
HIGH_RISK_COUNTRY2The country is in the security group's classifier.highRiskCountries (default list high_risk)Yes
LOCATION_DISTANCE0–5Distance between the IP's coordinates and lastKnownLocation: 0 up to 100 km, otherwise log2(km / 100 + 1), capped at 5 and rounded to 1 decimal. null without lastKnownLocation.Yes
TIMEZONE_MISMATCH1The request's timeZone belongs to a different country than the IP. null unless both are known.Yes

Rate limit

The rate-limit category counts toward the mean only when the limit is exceeded (score > 0).

RuleScoreTriggerCan be disabled
RATE_LIMIT10The request exceeds a rateLimit.limit from the request bodyYes

Text

The text engine scores each rule from 0 to 1. Sentinel multiplies that by 5 and rounds to 2 decimals, so these rules range from 0 to 5. URL_PHISHING is the exception. "Relative to text length" means the same number of hits scores lower in a longer text.

RuleScoreTriggerCan be disabled
SPAM_WORDS0–5Spam dictionary terms and phrases, weighted, relative to text lengthYes
PROFANITY0–5Profanity dictionary terms, each distinct term counted onceYes
URL0–5URLs (up to 5 counted), relative to text lengthYes
URL_PHISHING10 per URLEach distinct URL on the phishing list. 0 while the list is disabled or not loaded. Triggers the BAD floor.Yes
LINK_MARKUP0–5Link markup (href=, [url, [link), relative to text lengthYes
CURRENCY0–5Currency symbols next to a digit ($5, 5 €), relative to text lengthYes
EXCLAMATION0–5Runs of two or more !, relative to text lengthYes
REPEATED_CHARS0–5The same letter four or more times in a row, relative to text lengthYes
INVISIBLE_CHARS0–5Zero-width or other invisible characters between letters or digitsYes
MIXED_SCRIPT0–5Words that mix Latin with Cyrillic or Greek lettersYes
RANDOM_CHARS0–5Words with erratic letter case, or random letter-digit codes, relative to text lengthYes
EMOJI0–5Emojis, relative to text lengthYes
OBFUSCATION0–5Dictionary terms found only through obfuscation (leet, separators, repeated letters, * masking, look-alike characters)Yes
CAPITALIZATION0–5Mostly capital letters: at least 12 cased letters, 60% of them uppercaseYes

Injection

The injection category counts toward the mean only when it scores > 0, and any hit forces BAD (see Floors). POST /v2/eml doesn't run injection checks.

RuleScoreTriggerCan be disabled
SQL_INJECTION5An SQL lexer reads a value as SQL, either as is or after a quote the value breaks out of: comment truncation, tautology, UNION SELECT, stacked query, or SQL functionYes
HTML_INJECTION5A dangerous tag, an on… event-handler attribute, or a javascript:, vbscript:, or data:text/html URLYes
EMAIL_HEADER_INJECTION5A line break in an email value, or a line break followed by a mail header in a single-line fieldYes
TEMPLATE_INJECTION5A JNDI lookup, an arithmetic probe, or code access inside template delimiters ({{ }}, ${ }, …)Yes
NOSQL_INJECTION5A MongoDB operator key or object, or a JavaScript injectionYes

Similarity

Matches the text against the training groups selected by classifier.similarityGroups or the body's similarityGroups. The category score is max(SIMILARITY, NEAR_DUPLICATE) + CAMPAIGN, rounded to 2 decimals, and counts only when it is > 0. See Similarity.

RuleScoreTriggerCan be disabled
SIMILARITY0–10Semantic match: the highest 5 × weight × cosine similarity among the matched groups, capped at 10. null when no selected group has embedded samples or the text couldn't be embedded. A match against a sample with weight ≥ 1 forces BAD.Yes
NEAR_DUPLICATE0–10Near-duplicate (SimHash) match: the highest 5 × weight among the matched groups, capped at 10. null when the selected groups have no samples. A match against a sample with weight ≥ 1 forces BAD.Yes
CAMPAIGN0–10At least min_sources distinct sources sent this text (or a near-duplicate of it) within the campaign window: 5 × sources / min_sources, capped at 10. Otherwise 0. null for short texts or when campaign detection is off.Yes

Human Interaction Signature

Runs only on POST /v2/verify. See Human Interaction Signature.

RuleScoreTriggerCan be disabled
HIS_AUTOMATION0–3Bot probability of 0.5 or higher: (probability − 0.5) × 6, rounded to 2 decimals. Otherwise 0. A score ≥ 1.8 forces at least NEUTRAL.Yes
HIS_INTEGRITY3Inconsistent signature (fabricated or tampered samples, or a webdriver browser). Forces BAD.Yes
HIS_NO_INTERACTION1The visitor didn't move, touch, scroll, click, or typeYes
HIS_REPLAY2The same signature was already submitted. Forces at least NEUTRAL.Yes

AI

Each AI rule asks the model a yes/no question. The rule scores 5 × p, rounded to 2 decimals, where p is the model's probability of "yes", if p ≥ 0.7. Otherwise it scores 0. Rules that weren't asked are null. The highest probability also drives the AI floors. If the provider fails, ai.error is AI_UNAVAILABLE and no AI rule scores. See AI decider.

RuleScoreTriggerCan be disabled
AI_SPAM0 or 3.5–5Is the text spam?Yes
AI_SCAM0 or 3.5–5Is it a scam or fraud attempt?Yes
AI_ABUSE0 or 3.5–5Is it abusive, harassing, or hateful?Yes
AI_GIBBERISH0 or 3.5–5Is it gibberish?Yes
AI_OFF_TOPIC0 or 3.5–5Is it off-topic for the form? Asked only when classifier.aiPurpose is set.Yes
AI_CUSTOM0 or 3.5–5The group's own question. Asked only when classifier.aiInstructions is set.Yes

Email (EML) rules

POST /v2/eml scores a raw message with the EML rules below. It also runs the classifier on the sender address, subject and text, and originating IP. Email analysis is aggregated differently from the classifier:

AspectEML behavior
ScoreSum of the non-null EML rule scores (not a mean), rounded down to 2 decimals, with a minimum of 0
Classifier resultIts overall score becomes the CLASSIFICATION rule. The classifier's own rules appear in classification.triggeredRules.
Verdictspam is true when score is 2 or higher. There are no floors.
Classifier differencesNo device, rate-limit, HIS, or injection checks. HOSTING is skipped when the sending server is authorized by the sender's domain.
triggeredRulesEML rules that scored > 0, highest first. LIST_UNSUBSCRIBE never appears, because it only lowers the score.
DisablingX-Disable-Rules request header: up to 200 EML and classifier rule names, separated by ,, ;, or whitespace. Classifier names in it replace the security group's classifier.disableRules.
RuleScoreTriggerCan be disabled
ARC4 / 0.5ARC result fail: 4; neutral: 0.5; any other result: 0. null when not checked.Yes
ATTACHMENT_RISK0–4The highest risk among all attachments: deceptive name or executable content 4; archived executable, disk image, or executable 3; macros 2.5; encrypted archive 2; HTML/SVG or type mismatch 1.5Yes
BULK_MAILER1.5X-Mailer or User-Agent names a bulk-mailing toolYes
CLASSIFICATIONclassifier scoreThe classifier's overall score for the sender, subject, text, and originating IPYes
DATE_ANOMALY1 / 0.51 when Date is missing, invalid, or more than a day after the topmost Received time (or now, without one); 0.5 when it's more than 7 days before that timeYes
DELIVERED_TO_MISMATCH1The first Delivered-To differs from the first To addressYes
DKIM4 / 0.54 when any DKIM signature fails; otherwise 0.5 when any is neutral; otherwise 0. null when not checked.Yes
DMARC_ALIGNMENT4 / 1DMARC fail: 4 under a quarantine or reject policy, 1 under none or an unknown policy. null when not checked.Yes
FROM_SPOOFING2A From display name contains a different address than the sender's ("admin@paypal.com" <evil@attacker.example>)Yes
HIDDEN_TEXT1.5At least 200 characters of hidden HTML text, making up more than half of the HTML's textYes
HTML_ONLY0.5An HTML body without a text/plain alternativeYes
LINK_PHISHING10 per linkEach distinct link target on the phishing list that the text's URL_PHISHING didn't already scoreYes
LINK_TEXT_MISMATCH1An anchor's text shows a URL or domain on a different site than its target (subdomains count as the same site)Yes
LIST_UNSUBSCRIBE-1A List-Unsubscribe header on mail that passes DMARC. Lowers the score.Yes
MALICIOUS_ATTACHMENT10 per attachmentEach distinct attachment (by checksum) on the malicious file hash listYes
MESSAGE_ID_ANOMALY1 / 0.51 when Message-ID is missing; 0.5 when it's malformed or its domain is unrelated to the sender and relaysYes
NO_SUBJECT0.5No subject, or the subject (No Subject)Yes
NO_TEXT2No text bodyYes
REPLY_TO_SPOOFING2The first Reply-To differs from the first From: by address for a free-provider sender, otherwise by domainYes
SPF4 / 0.5SPF result fail: 4; neutral: 0.5; any other result: 0. null when not checked.Yes
SUSPICIOUS_LINK2 / 0.5The most suspicious link: 2 for a javascript: URL, a data: URL outside img, or an IP-address host; 0.5 for a link shortenerYes
UNDISCLOSED_RECIPIENTS1No To recipientsYes

See Email analysis and the API reference.

Start typing to search...

↑ ↓ Navigate ↵ Select