diff --git a/packages/inflekt/__tests__/inflection.test.ts b/packages/inflekt/__tests__/inflection.test.ts
index 6d86a36..dcd44f4 100644
--- a/packages/inflekt/__tests__/inflection.test.ts
+++ b/packages/inflekt/__tests__/inflection.test.ts
@@ -46,6 +46,14 @@ describe('singularize', () => {
     expect(singularize('apiSchemata')).toBe('apiSchema');
     expect(singularize('SCHEMATA')).toBe('SCHEMA');
   });
+
+  it('should canonicalize malformed trailing triple-s words', () => {
+    expect(singularize('classs')).toBe('class');
+    expect(singularize('Classs')).toBe('Class');
+    expect(singularize('hazardClasss')).toBe('hazardClass');
+    expect(singularize('hazardClassses')).toBe('hazardClass');
+    expect(singularize('CLASSS')).toBe('CLASS');
+  });
 });
 
 describe('pluralize', () => {
@@ -55,6 +63,36 @@ describe('pluralize', () => {
     expect(pluralize('Person')).toBe('People');
     expect(pluralize('Category')).toBe('Categories');
   });
+
+  it('should normalize class variants', () => {
+    expect(pluralize('class')).toBe('classes');
+    expect(pluralize('Class')).toBe('Classes');
+    expect(pluralize('hazardClass')).toBe('hazardClasses');
+    expect(pluralize('HazardClass')).toBe('HazardClasses');
+    expect(pluralize('hazardClasss')).toBe('hazardClasses');
+    expect(pluralize('classs')).toBe('classes');
+  });
+
+  it('should preserve already-plural Latin words', () => {
+    expect(pluralize('Schemata')).toBe('Schemata');
+    expect(pluralize('schemata')).toBe('schemata');
+  });
+
+  it.each([
+    ['class', 'classes'],
+    ['glass', 'glasses'],
+    ['boss', 'bosses'],
+    ['process', 'processes'],
+    ['address', 'addresses'],
+    ['witness', 'witnesses'],
+    ['abyss', 'abysses'],
+  ])(
+    'should handle -ss noun roundtrip for %s -> %s',
+    (singularWord, pluralWord) => {
+      expect(pluralize(singularWord)).toBe(pluralWord);
+      expect(singularize(pluralWord)).toBe(singularWord);
+    }
+  );
 });
 
 describe('singularizeLast', () => {
@@ -69,6 +107,11 @@ describe('singularizeLast', () => {
     expect(singularizeLast('api_schemata')).toBe('api_schema');
     expect(singularizeLast('ApiSchemata')).toBe('ApiSchema');
   });
+
+  it('should normalize malformed class suffixes in the final segment', () => {
+    expect(singularizeLast('hazardClassses')).toBe('hazardClass');
+    expect(singularizeLast('HazardClassses')).toBe('HazardClass');
+  });
 });
 
 describe('pluralizeLast', () => {
@@ -78,6 +121,13 @@ describe('pluralizeLast', () => {
     expect(pluralizeLast('order_item')).toBe('order_items');
     expect(pluralizeLast('OrderItem')).toBe('OrderItems');
   });
+
+  it('should normalize malformed class suffixes in the final segment', () => {
+    expect(pluralizeLast('hazardClass')).toBe('hazardClasses');
+    expect(pluralizeLast('HazardClass')).toBe('HazardClasses');
+    expect(pluralizeLast('hazardClasss')).toBe('hazardClasses');
+    expect(pluralizeLast('HazardClasss')).toBe('HazardClasses');
+  });
 });
 
 describe('distinctPluralize', () => {
@@ -95,6 +145,11 @@ describe('distinctPluralize', () => {
     expect(distinctPluralize('bus')).toBe('buses');
     expect(distinctPluralize('box')).toBe('boxes');
   });
+
+  it('should normalize malformed class variants', () => {
+    expect(distinctPluralize('classs')).toBe('classes');
+    expect(distinctPluralize('hazardClasss')).toBe('hazardClasses');
+  });
 });
 
 describe('distinctPluralizeLast', () => {
@@ -102,6 +157,11 @@ describe('distinctPluralizeLast', () => {
     expect(distinctPluralizeLast('user_profile')).toBe('user_profiles');
     expect(distinctPluralizeLast('UserProfile')).toBe('UserProfiles');
   });
+
+  it('should normalize malformed class variants in the final segment', () => {
+    expect(distinctPluralizeLast('hazardClasss')).toBe('hazardClasses');
+    expect(distinctPluralizeLast('HazardClasss')).toBe('HazardClasses');
+  });
 });
 
 describe('lcFirst', () => {
@@ -189,6 +249,12 @@ describe('toFieldName', () => {
     expect(toFieldName('Schemata')).toBe('schema');
     expect(toFieldName('ApiSchemata')).toBe('apiSchema');
   });
+
+  it('should normalize malformed class variants', () => {
+    expect(toFieldName('Classes')).toBe('class');
+    expect(toFieldName('HazardClasses')).toBe('hazardClass');
+    expect(toFieldName('HazardClassses')).toBe('hazardClass');
+  });
 });
 
 describe('toQueryName', () => {
@@ -196,5 +262,7 @@ describe('toQueryName', () => {
     expect(toQueryName('User')).toBe('users');
     expect(toQueryName('OrderItem')).toBe('orderItems');
     expect(toQueryName('Category')).toBe('categories');
+    expect(toQueryName('Class')).toBe('classes');
+    expect(toQueryName('HazardClass')).toBe('hazardClasses');
   });
 });
diff --git a/packages/inflekt/src/pluralize.ts b/packages/inflekt/src/pluralize.ts
index 01faa91..4701469 100644
--- a/packages/inflekt/src/pluralize.ts
+++ b/packages/inflekt/src/pluralize.ts
@@ -25,18 +25,59 @@ const LATIN_SUFFIX_OVERRIDES: Array<[string, string]> = [
   ['data', 'datum'],
 ];
 
+const TRAILING_TRIPLE_S_REGEX = /[sS]{3,}$/;
+const TRAILING_TRIPLE_S_BEFORE_ES_REGEX = /[sS]{3,}(?=e[sS]$)/;
+
+function normalizeTrailingSRun(suffix: string): string {
+  return suffix === suffix.toUpperCase() ? 'SS' : 'ss';
+}
+
+function normalizeTripleSBeforeEs(word: string): string {
+  return word.replace(TRAILING_TRIPLE_S_BEFORE_ES_REGEX, normalizeTrailingSRun);
+}
+
+function normalizeTrailingTripleS(word: string): string {
+  const match = word.match(TRAILING_TRIPLE_S_REGEX);
+  if (!match) {
+    return word;
+  }
+
+  const suffix = match[0];
+  const prefix = word.slice(0, -suffix.length);
+  const normalizedSuffix = normalizeTrailingSRun(suffix);
+  return `${prefix}${normalizedSuffix}`;
+}
+
+function normalizeMalformedDoubleS(word: string): string {
+  return normalizeTrailingTripleS(normalizeTripleSBeforeEs(word));
+}
+
+function enforceDoubleSPlural(singularWord: string, pluralWord: string): string {
+  if (!singularWord.toLowerCase().endsWith('ss')) {
+    return pluralWord;
+  }
+
+  // Defensive normalization for malformed outputs like "hazardClasss".
+  if (pluralWord === `${singularWord}s`) {
+    return `${singularWord}es`;
+  }
+
+  return pluralWord;
+}
+
 /**
  * Convert a word to its singular form with PostGraphile-compatible Latin handling
  * @example "Users" -> "User", "People" -> "Person", "Schemata" -> "Schema", "ApiSchemata" -> "ApiSchema"
  */
 export function singularize(word: string): string {
-  const lowerWord = word.toLowerCase();
+  const normalizedWord = normalizeMalformedDoubleS(word);
+  const lowerWord = normalizedWord.toLowerCase();
 
   for (const [pluralSuffix, singularSuffix] of LATIN_SUFFIX_OVERRIDES) {
     if (lowerWord.endsWith(pluralSuffix)) {
-      const suffixStart = word.length - pluralSuffix.length;
-      const prefix = word.slice(0, suffixStart);
-      const originalSuffix = word.slice(suffixStart);
+      const suffixStart = normalizedWord.length - pluralSuffix.length;
+      const prefix = normalizedWord.slice(0, suffixStart);
+      const originalSuffix = normalizedWord.slice(suffixStart);
 
       const isAllCaps = originalSuffix === originalSuffix.toUpperCase();
       const isUpperSuffix =
@@ -55,7 +96,13 @@ export function singularize(word: string): string {
     }
   }
 
-  return inflection.singularize(word);
+  return normalizeMalformedDoubleS(inflection.singularize(normalizedWord));
+}
+
+function pluralizeCanonical(word: string): string {
+  const normalizedWord = normalizeMalformedDoubleS(word);
+  const pluralWord = normalizeMalformedDoubleS(inflection.pluralize(normalizedWord));
+  return enforceDoubleSPlural(singularize(normalizedWord), pluralWord);
 }
 
 /**
@@ -63,7 +110,7 @@ export function singularize(word: string): string {
  * @example "User" -> "Users", "Person" -> "People"
  */
 export function pluralize(word: string): string {
-  return inflection.pluralize(word);
+  return pluralizeCanonical(word);
 }
 
 /**