namespace-guard 0.8.1 → 0.10.0

package/README.md

@@ -5,7 +5,7 @@
 [![TypeScript](https://img.shields.io/badge/TypeScript-5.0+-blue.svg)](https://www.typescriptlang.org/)
 [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
 
-**[Live Demo](https://paultendo.github.io/namespace-guard/)** — try it in your browser | **[Blog Post](https://paultendo.github.io/posts/namespace-guard-launch/)** — why this exists
+**[Live Demo](https://paultendo.github.io/namespace-guard/)** - try it in your browser | **[Blog Post](https://paultendo.github.io/posts/namespace-guard-launch/)** - why this exists
 
 **Check slug/handle uniqueness across multiple database tables with reserved name protection.**
 
@@ -249,7 +249,7 @@ const result = await guard.check("admin");
 // { available: false, reason: "reserved", category: "system", message: "That's a system route." }
 ```
 
-You can also use a single string message for all categories, or mix — categories without a specific message fall back to the default.
+You can also use a single string message for all categories, or mix - categories without a specific message fall back to the default.
 
 ## Async Validators
 
@@ -279,7 +279,7 @@ Validators run sequentially and stop at the first rejection. They receive the no
 
 ### Built-in Profanity Validator
 
-Use `createProfanityValidator` for a turnkey profanity filter — supply your own word list:
+Use `createProfanityValidator` for a turnkey profanity filter - supply your own word list:
 
 ```typescript
 import { createNamespaceGuard, createProfanityValidator } from "namespace-guard";
@@ -295,7 +295,7 @@ const guard = createNamespaceGuard({
 }, adapter);
 ```
 
-No words are bundled — use any word list you like (e.g., the `bad-words` npm package, your own list, or an external API wrapped in a custom validator).
+No words are bundled - use any word list you like (e.g., the `bad-words` npm package, your own list, or an external API wrapped in a custom validator).
 
 ### Built-in Homoglyph Validator
 
@@ -324,6 +324,44 @@ createHomoglyphValidator({
 
 The built-in `CONFUSABLE_MAP` contains 613 character pairs generated from [Unicode TR39 confusables.txt](https://unicode.org/reports/tr39/) plus supplemental Latin small capitals. It covers Cyrillic, Greek, Armenian, Cherokee, IPA, Coptic, Lisu, Canadian Syllabics, Georgian, and 20+ other scripts. The map is exported for inspection or extension, and is regenerable for new Unicode versions with `npx tsx scripts/generate-confusables.ts`.
 
+#### CONFUSABLE_MAP_FULL
+
+For standalone use without NFKC normalization, `CONFUSABLE_MAP_FULL` (~1,400 entries) includes every single-character-to-Latin mapping from TR39 with no NFKC filtering. This is the right map when your pipeline does not run NFKC before confusable detection, which is the case for most real-world systems: TR39's skeleton algorithm uses NFD, Chromium's IDN spoof checker uses NFD, Rust's `confusable_idents` lint runs on NFC, and django-registration applies the confusable map to raw input with no normalization at all.
+
+```typescript
+import { CONFUSABLE_MAP_FULL } from "namespace-guard";
+
+// Contains everything in CONFUSABLE_MAP, plus:
+// - ~766 entries where NFKC agrees with TR39 (mathematical alphanumerics, fullwidth forms)
+// - 31 entries where TR39 and NFKC disagree on the target letter
+CONFUSABLE_MAP_FULL["\u017f"]; // "f" (Long S: TR39 visual mapping)
+CONFUSABLE_MAP_FULL["\u{1D41A}"]; // "a" (Mathematical Bold Small A)
+```
+
+#### `skeleton()` and `areConfusable()`
+
+The TR39 Section 4 skeleton algorithm computes a normalized form of a string for confusable comparison. Two strings that look alike will produce the same skeleton. This is the same algorithm used by ICU's SpoofChecker, Chromium's IDN spoof checker, and the Rust compiler's `confusable_idents` lint.
+
+```typescript
+import { skeleton, areConfusable, CONFUSABLE_MAP } from "namespace-guard";
+
+// Compute skeletons for comparison
+skeleton("paypal");           // "paypal"
+skeleton("\u0440\u0430ypal"); // "paypal" (Cyrillic р and а)
+skeleton("pay\u200Bpal");     // "paypal" (zero-width space stripped)
+skeleton("\u017f");            // "f"      (Long S via TR39 visual mapping)
+
+// Compare two strings directly
+areConfusable("paypal", "\u0440\u0430ypal"); // true
+areConfusable("google", "g\u043e\u043egle"); // true  (Cyrillic о)
+areConfusable("hello", "world");             // false
+
+// Use CONFUSABLE_MAP for NFKC-first pipelines
+skeleton("\u017f", { map: CONFUSABLE_MAP }); // "\u017f" (Long S not in filtered map)
+```
+
+By default, `skeleton()` uses `CONFUSABLE_MAP_FULL` (the complete TR39 map), which matches the NFD-based pipeline specified by TR39. Pass `{ map: CONFUSABLE_MAP }` if your pipeline runs NFKC normalization before calling `skeleton()`.
+
 ### How the anti-spoofing pipeline works
 
 Most confusable-detection libraries apply a character map in isolation. namespace-guard uses a three-stage pipeline where each stage is aware of the others:
@@ -335,13 +373,13 @@ Input  →  NFKC normalize  →  Confusable map  →  Mixed-script reject
 
 **Stage 1: NFKC normalization** collapses full-width characters (`Ｉ` → `I`), ligatures (`fi` → `fi`), superscripts, and other Unicode compatibility forms to their canonical equivalents. This runs first, before any confusable check.
 
-**Stage 2: Confusable map** catches characters that survive NFKC but visually mimic Latin letters — Cyrillic `а` for `a`, Greek `ο` for `o`, Cherokee `Ꭺ` for `A`, and 600+ others from the Unicode Consortium's [confusables.txt](https://unicode.org/Public/security/latest/confusables.txt).
+**Stage 2: Confusable map** catches characters that survive NFKC but visually mimic Latin letters - Cyrillic `а` for `a`, Greek `ο` for `o`, Cherokee `Ꭺ` for `A`, and 600+ others from the Unicode Consortium's [confusables.txt](https://unicode.org/Public/security/latest/confusables.txt).
 
 **Stage 3: Mixed-script rejection** (`rejectMixedScript: true`) blocks identifiers that mix Latin with non-Latin scripts (Hebrew, Arabic, Devanagari, Thai, Georgian, Ethiopic, etc.) even if the specific characters aren't in the confusable map. This catches novel homoglyphs that the map doesn't cover.
 
 #### Why NFKC-aware filtering matters
 
-The key insight: TR39's confusables.txt and NFKC normalization sometimes disagree. For example, Unicode says capital `I` (U+0049) is confusable with lowercase `l` — visually true in many fonts. But NFKC maps Mathematical Bold `𝐈` (U+1D408) to `I`, not `l`. If you naively ship the TR39 mapping (`𝐈` → `l`), the confusable check will never see that character — NFKC already converted it to `I` in stage 1.
+The key insight: TR39's confusables.txt and NFKC normalization sometimes disagree. For example, Unicode says capital `I` (U+0049) is confusable with lowercase `l` - visually true in many fonts. But NFKC maps Mathematical Bold `𝐈` (U+1D408) to `I`, not `l`. If you naively ship the TR39 mapping (`𝐈` → `l`), the confusable check will never see that character - NFKC already converted it to `I` in stage 1.
 
 We found 31 entries where this happens:
 
@@ -354,7 +392,7 @@ We found 31 entries where this happens:
 | 11 Mathematical I variants | `l` | `i` | NFKC |
 | 12 Mathematical 0/1 variants | `o`/`l` | `0`/`1` | NFKC |
 
-These entries are dead code in any pipeline that runs NFKC first — and worse, they encode the *wrong* mapping. The generate script (`scripts/generate-confusables.ts`) automatically detects and excludes them.
+These entries are dead code in any pipeline that runs NFKC first - and worse, they encode the *wrong* mapping. The generate script (`scripts/generate-confusables.ts`) automatically detects and excludes them.
 
 ## Unicode Normalization
 
@@ -429,7 +467,7 @@ const result = await guard.check("acme-corp");
 
 ### Composing Strategies
 
-Combine multiple strategies — candidates are interleaved round-robin:
+Combine multiple strategies - candidates are interleaved round-robin:
 
 ```typescript
 suggest: {
@@ -503,10 +541,10 @@ npx namespace-guard check acme-corp
 # ✓ acme-corp is available
 
 npx namespace-guard check admin
-# ✗ admin — That name is reserved. Try another one.
+# ✗ admin - That name is reserved. Try another one.
 
 npx namespace-guard check "a"
-# ✗ a — Use 2-30 lowercase letters, numbers, or hyphens.
+# ✗ a - Use 2-30 lowercase letters, numbers, or hyphens.
 ```
 
 ### With a config file
@@ -648,7 +686,8 @@ Enable in-memory caching to reduce database calls during rapid checks (e.g., liv
 const guard = createNamespaceGuard({
   sources: [/* ... */],
   cache: {
-    ttl: 5000, // milliseconds (default: 5000)
+    ttl: 5000,     // milliseconds (default: 5000)
+    maxSize: 1000, // max cached entries before LRU eviction (default: 1000)
   },
 }, adapter);
 
@@ -760,7 +799,10 @@ import {
   createNamespaceGuard,
   createProfanityValidator,
   createHomoglyphValidator,
+  skeleton,
+  areConfusable,
   CONFUSABLE_MAP,
+  CONFUSABLE_MAP_FULL,
   normalize,
   type NamespaceConfig,
   type NamespaceSource,
@@ -770,6 +812,7 @@ import {
   type FindOneOptions,
   type OwnershipScope,
   type SuggestStrategyName,
+  type SkeletonOptions,
 } from "namespace-guard";
 ```
 

package/dist/adapters/raw.d.mts

@@ -3,32 +3,6 @@ import { NamespaceAdapter } from '../index.mjs';
 type QueryExecutor = (sql: string, params: unknown[]) => Promise<{
     rows: Record<string, unknown>[];
 }>;
-/**
- * Create a namespace adapter for raw SQL queries
- *
- * Works with PostgreSQL-compatible clients that use $1-style parameter placeholders (pg).
- * For MySQL or SQLite, wrap the executor to translate parameter syntax.
- *
- * @example
- * ```ts
- * import { Pool } from "pg";
- * import { createNamespaceGuard } from "namespace-guard";
- * import { createRawAdapter } from "namespace-guard/adapters/raw";
- *
- * const pool = new Pool();
- *
- * const guard = createNamespaceGuard(
- *   {
- *     reserved: ["admin", "api", "settings"],
- *     sources: [
- *       { name: "users", column: "handle", scopeKey: "id" },
- *       { name: "organizations", column: "slug", scopeKey: "id" },
- *     ],
- *   },
- *   createRawAdapter((sql, params) => pool.query(sql, params))
- * );
- * ```
- */
 declare function createRawAdapter(execute: QueryExecutor): NamespaceAdapter;
 
 export { createRawAdapter };

package/dist/adapters/raw.d.ts

@@ -3,32 +3,6 @@ import { NamespaceAdapter } from '../index.js';
 type QueryExecutor = (sql: string, params: unknown[]) => Promise<{
     rows: Record<string, unknown>[];
 }>;
-/**
- * Create a namespace adapter for raw SQL queries
- *
- * Works with PostgreSQL-compatible clients that use $1-style parameter placeholders (pg).
- * For MySQL or SQLite, wrap the executor to translate parameter syntax.
- *
- * @example
- * ```ts
- * import { Pool } from "pg";
- * import { createNamespaceGuard } from "namespace-guard";
- * import { createRawAdapter } from "namespace-guard/adapters/raw";
- *
- * const pool = new Pool();
- *
- * const guard = createNamespaceGuard(
- *   {
- *     reserved: ["admin", "api", "settings"],
- *     sources: [
- *       { name: "users", column: "handle", scopeKey: "id" },
- *       { name: "organizations", column: "slug", scopeKey: "id" },
- *     ],
- *   },
- *   createRawAdapter((sql, params) => pool.query(sql, params))
- * );
- * ```
- */
 declare function createRawAdapter(execute: QueryExecutor): NamespaceAdapter;
 
 export { createRawAdapter };

package/dist/adapters/raw.js

@@ -23,10 +23,20 @@ __export(raw_exports, {
   createRawAdapter: () => createRawAdapter
 });
 module.exports = __toCommonJS(raw_exports);
+var SAFE_IDENTIFIER = /^[a-zA-Z_][a-zA-Z0-9_]*$/;
+function assertSafeIdentifier(name, label) {
+  if (!SAFE_IDENTIFIER.test(name)) {
+    throw new Error(`Unsafe ${label}: ${JSON.stringify(name)}. Use only letters, digits, and underscores.`);
+  }
+}
 function createRawAdapter(execute) {
   return {
     async findOne(source, value, options) {
       const idColumn = source.idColumn ?? "id";
+      assertSafeIdentifier(source.name, "table name");
+      assertSafeIdentifier(source.column, "column name");
+      assertSafeIdentifier(idColumn, "id column name");
+      if (source.scopeKey) assertSafeIdentifier(source.scopeKey, "scope key");
       const columns = source.scopeKey && source.scopeKey !== idColumn ? `"${idColumn}", "${source.scopeKey}"` : `"${idColumn}"`;
       const whereClause = options?.caseInsensitive ? `LOWER("${source.column}") = LOWER($1)` : `"${source.column}" = $1`;
       const sql = `SELECT ${columns} FROM "${source.name}" WHERE ${whereClause} LIMIT 1`;

package/dist/adapters/raw.mjs

@@ -1,8 +1,18 @@
 // src/adapters/raw.ts
+var SAFE_IDENTIFIER = /^[a-zA-Z_][a-zA-Z0-9_]*$/;
+function assertSafeIdentifier(name, label) {
+  if (!SAFE_IDENTIFIER.test(name)) {
+    throw new Error(`Unsafe ${label}: ${JSON.stringify(name)}. Use only letters, digits, and underscores.`);
+  }
+}
 function createRawAdapter(execute) {
   return {
     async findOne(source, value, options) {
       const idColumn = source.idColumn ?? "id";
+      assertSafeIdentifier(source.name, "table name");
+      assertSafeIdentifier(source.column, "column name");
+      assertSafeIdentifier(idColumn, "id column name");
+      if (source.scopeKey) assertSafeIdentifier(source.scopeKey, "scope key");
       const columns = source.scopeKey && source.scopeKey !== idColumn ? `"${idColumn}", "${source.scopeKey}"` : `"${idColumn}"`;
       const whereClause = options?.caseInsensitive ? `LOWER("${source.column}") = LOWER($1)` : `"${source.column}" = $1`;
       const sql = `SELECT ${columns} FROM "${source.name}" WHERE ${whereClause} LIMIT 1`;

package/dist/cli.js

@@ -144,14 +144,12 @@ function createScrambleStrategy(_pattern) {
     const candidates = [];
     const chars = identifier.split("");
     for (let i = 0; i < chars.length - 1; i++) {
-      if (chars[i] !== chars[i + 1]) {
-        const swapped = [...chars];
-        [swapped[i], swapped[i + 1]] = [swapped[i + 1], swapped[i]];
-        const candidate = swapped.join("");
-        if (candidate !== identifier && !seen.has(candidate)) {
-          seen.add(candidate);
-          candidates.push(candidate);
-        }
+      const swapped = [...chars];
+      [swapped[i], swapped[i + 1]] = [swapped[i + 1], swapped[i]];
+      const candidate = swapped.join("");
+      if (candidate !== identifier && !seen.has(candidate)) {
+        seen.add(candidate);
+        candidates.push(candidate);
       }
     }
     return candidates;
@@ -298,7 +296,7 @@ function createNamespaceGuard(config, adapter) {
   const purelyNumericMsg = configMessages.purelyNumeric ?? "Identifiers cannot be purely numeric.";
   const cacheEnabled = !!config.cache;
   const cacheTtl = config.cache?.ttl ?? 5e3;
-  const cacheMaxSize = 1e3;
+  const cacheMaxSize = config.cache?.maxSize ?? 1e3;
   const cacheMap = /* @__PURE__ */ new Map();
   let cacheHits = 0;
   let cacheMisses = 0;
@@ -488,10 +486,20 @@ function createNamespaceGuard(config, adapter) {
 }
 
 // src/adapters/raw.ts
+var SAFE_IDENTIFIER = /^[a-zA-Z_][a-zA-Z0-9_]*$/;
+function assertSafeIdentifier(name, label) {
+  if (!SAFE_IDENTIFIER.test(name)) {
+    throw new Error(`Unsafe ${label}: ${JSON.stringify(name)}. Use only letters, digits, and underscores.`);
+  }
+}
 function createRawAdapter(execute) {
   return {
     async findOne(source, value, options) {
       const idColumn = source.idColumn ?? "id";
+      assertSafeIdentifier(source.name, "table name");
+      assertSafeIdentifier(source.column, "column name");
+      assertSafeIdentifier(idColumn, "id column name");
+      if (source.scopeKey) assertSafeIdentifier(source.scopeKey, "scope key");
       const columns = source.scopeKey && source.scopeKey !== idColumn ? `"${idColumn}", "${source.scopeKey}"` : `"${idColumn}"`;
       const whereClause = options?.caseInsensitive ? `LOWER("${source.column}") = LOWER($1)` : `"${source.column}" = $1`;
       const sql = `SELECT ${columns} FROM "${source.name}" WHERE ${whereClause} LIMIT 1`;
@@ -574,6 +582,10 @@ async function createDatabaseAdapter(url) {
     process.exit(1);
   }
   const Pool = pg.default?.Pool ?? pg.Pool;
+  if (!Pool) {
+    console.error("Could not find Pool export from pg package. Check your pg version.");
+    process.exit(1);
+  }
   const pool = new Pool({ connectionString: url });
   return {
     adapter: createRawAdapter((sql, params) => pool.query(sql, params)),

package/dist/cli.mjs

@@ -121,14 +121,12 @@ function createScrambleStrategy(_pattern) {
     const candidates = [];
     const chars = identifier.split("");
     for (let i = 0; i < chars.length - 1; i++) {
-      if (chars[i] !== chars[i + 1]) {
-        const swapped = [...chars];
-        [swapped[i], swapped[i + 1]] = [swapped[i + 1], swapped[i]];
-        const candidate = swapped.join("");
-        if (candidate !== identifier && !seen.has(candidate)) {
-          seen.add(candidate);
-          candidates.push(candidate);
-        }
+      const swapped = [...chars];
+      [swapped[i], swapped[i + 1]] = [swapped[i + 1], swapped[i]];
+      const candidate = swapped.join("");
+      if (candidate !== identifier && !seen.has(candidate)) {
+        seen.add(candidate);
+        candidates.push(candidate);
       }
     }
     return candidates;
@@ -275,7 +273,7 @@ function createNamespaceGuard(config, adapter) {
   const purelyNumericMsg = configMessages.purelyNumeric ?? "Identifiers cannot be purely numeric.";
   const cacheEnabled = !!config.cache;
   const cacheTtl = config.cache?.ttl ?? 5e3;
-  const cacheMaxSize = 1e3;
+  const cacheMaxSize = config.cache?.maxSize ?? 1e3;
   const cacheMap = /* @__PURE__ */ new Map();
   let cacheHits = 0;
   let cacheMisses = 0;
@@ -465,10 +463,20 @@ function createNamespaceGuard(config, adapter) {
 }
 
 // src/adapters/raw.ts
+var SAFE_IDENTIFIER = /^[a-zA-Z_][a-zA-Z0-9_]*$/;
+function assertSafeIdentifier(name, label) {
+  if (!SAFE_IDENTIFIER.test(name)) {
+    throw new Error(`Unsafe ${label}: ${JSON.stringify(name)}. Use only letters, digits, and underscores.`);
+  }
+}
 function createRawAdapter(execute) {
   return {
     async findOne(source, value, options) {
       const idColumn = source.idColumn ?? "id";
+      assertSafeIdentifier(source.name, "table name");
+      assertSafeIdentifier(source.column, "column name");
+      assertSafeIdentifier(idColumn, "id column name");
+      if (source.scopeKey) assertSafeIdentifier(source.scopeKey, "scope key");
       const columns = source.scopeKey && source.scopeKey !== idColumn ? `"${idColumn}", "${source.scopeKey}"` : `"${idColumn}"`;
       const whereClause = options?.caseInsensitive ? `LOWER("${source.column}") = LOWER($1)` : `"${source.column}" = $1`;
       const sql = `SELECT ${columns} FROM "${source.name}" WHERE ${whereClause} LIMIT 1`;
@@ -551,6 +559,10 @@ async function createDatabaseAdapter(url) {
     process.exit(1);
   }
   const Pool = pg.default?.Pool ?? pg.Pool;
+  if (!Pool) {
+    console.error("Could not find Pool export from pg package. Check your pg version.");
+    process.exit(1);
+  }
   const pool = new Pool({ connectionString: url });
   return {
     adapter: createRawAdapter((sql, params) => pool.query(sql, params)),

package/dist/index.d.mts

@@ -6,14 +6,14 @@ type NamespaceSource = {
     column: string;
     /** Column name for the primary key (default: "id", or "_id" for Mongoose) */
     idColumn?: string;
-    /** Scope key for ownership checks — allows users to update their own slug without a false collision */
+    /** Scope key for ownership checks - allows users to update their own slug without a false collision */
     scopeKey?: string;
 };
 /** Built-in suggestion strategy names. */
 type SuggestStrategyName = "sequential" | "random-digits" | "suffix-words" | "short-random" | "scramble" | "similar";
 /** Configuration for a namespace guard instance. */
 type NamespaceConfig = {
-    /** Reserved names — flat list, Set, or categorized record */
+    /** Reserved names - flat list, Set, or categorized record */
     reserved?: Set<string> | string[] | Record<string, string[]>;
     /** Data sources to check for collisions */
     sources: NamespaceSource[];
@@ -35,7 +35,7 @@ type NamespaceConfig = {
         /** Message shown when a purely numeric identifier is rejected (default: "Identifiers cannot be purely numeric.") */
         purelyNumeric?: string;
     };
-    /** Async validation hooks — run after format/reserved checks, before DB */
+    /** Async validation hooks - run after format/reserved checks, before DB */
     validators?: Array<(value: string) => Promise<{
         available: false;
         message: string;
@@ -53,6 +53,8 @@ type NamespaceConfig = {
     cache?: {
         /** Time-to-live in milliseconds (default: 5000) */
         ttl?: number;
+        /** Maximum number of cached entries before LRU eviction (default: 1000) */
+        maxSize?: number;
     };
 };
 /** Options passed to adapter `findOne` calls. */
@@ -60,7 +62,7 @@ type FindOneOptions = {
     /** Use case-insensitive matching */
     caseInsensitive?: boolean;
 };
-/** Database adapter interface — implement this for your ORM or query builder. */
+/** Database adapter interface - implement this for your ORM or query builder. */
 type NamespaceAdapter = {
     findOne: (source: NamespaceSource, value: string, options?: FindOneOptions) => Promise<Record<string, unknown> | null>;
 };
@@ -80,6 +82,13 @@ type CheckResult = {
     category?: string;
     suggestions?: string[];
 };
+/** Options for the `skeleton()` and `areConfusable()` functions. */
+type SkeletonOptions = {
+    /** Confusable character map to use.
+     *  Default: `CONFUSABLE_MAP_FULL` (complete TR39 map, no NFKC filtering).
+     *  Pass `CONFUSABLE_MAP` if your pipeline runs NFKC before calling skeleton(). */
+    map?: Record<string, string>;
+};
 /**
  * Normalize a raw identifier: trims whitespace, applies NFKC Unicode normalization,
  * lowercases, and strips leading `@` symbols.
@@ -106,7 +115,7 @@ declare function normalize(raw: string, options?: {
 /**
  * Create a validator that rejects identifiers containing profanity or offensive words.
  *
- * Supply your own word list — no words are bundled with the library.
+ * Supply your own word list - no words are bundled with the library.
  * The returned function is compatible with `config.validators`.
  *
  * @param words - Array of words to block
@@ -144,6 +153,22 @@ declare function createProfanityValidator(words: string[], options?: {
  * Regenerate: `npx tsx scripts/generate-confusables.ts`
  */
 declare const CONFUSABLE_MAP: Record<string, string>;
+/**
+ * Complete TR39 confusable mapping: every single-character mapping to a
+ * lowercase Latin letter or digit from confusables.txt, with no NFKC filtering.
+ *
+ * Use this when your pipeline does NOT run NFKC normalization before confusable
+ * detection (which is most real-world systems: TR39 skeleton uses NFD, Chromium
+ * uses NFD, Rust uses NFC, django-registration uses no normalization at all).
+ *
+ * Includes ~1,400 entries vs CONFUSABLE_MAP's ~613 NFKC-deduped entries.
+ * The additional entries cover characters that NFKC normalization would handle
+ * (mathematical alphanumerics, fullwidth forms, etc.) plus the 31 entries where
+ * TR39 and NFKC disagree on the target letter.
+ *
+ * Regenerate: `npx tsx scripts/generate-confusables.ts`
+ */
+declare const CONFUSABLE_MAP_FULL: Record<string, string>;
 /**
  * Create a validator that rejects identifiers containing homoglyph/confusable characters.
  *
@@ -177,6 +202,48 @@ declare function createHomoglyphValidator(options?: {
     available: false;
     message: string;
 } | null>;
+/**
+ * Compute the TR39 Section 4 skeleton of a string for confusable comparison.
+ *
+ * Implements `internalSkeleton`:
+ * 1. NFD normalize
+ * 2. Remove Default_Ignorable_Code_Point characters
+ * 3. Replace each character via the confusable map
+ * 4. Reapply NFD
+ * 5. Lowercase
+ *
+ * The default map is `CONFUSABLE_MAP_FULL` (the complete TR39 mapping without
+ * NFKC filtering), which matches the NFD-based pipeline used by ICU, Chromium,
+ * and the TR39 spec itself. Pass `{ map: CONFUSABLE_MAP }` if your pipeline
+ * runs NFKC normalization before calling skeleton().
+ *
+ * @param input - The string to skeletonize
+ * @param options - Optional settings (custom confusable map)
+ * @returns The skeleton string for comparison
+ *
+ * @example
+ * ```ts
+ * skeleton("paypal") === skeleton("\u0440\u0430ypal") // true (Cyrillic р/а)
+ * skeleton("pay\u200Bpal") === skeleton("paypal")     // true (zero-width stripped)
+ * ```
+ */
+declare function skeleton(input: string, options?: SkeletonOptions): string;
+/**
+ * Check whether two strings are visually confusable by comparing their TR39 skeletons.
+ *
+ * @param a - First string
+ * @param b - Second string
+ * @param options - Optional settings (custom confusable map)
+ * @returns `true` if the strings produce the same skeleton
+ *
+ * @example
+ * ```ts
+ * areConfusable("paypal", "\u0440\u0430ypal") // true
+ * areConfusable("google", "g\u043e\u043egle") // true
+ * areConfusable("hello", "world")             // false
+ * ```
+ */
+declare function areConfusable(a: string, b: string, options?: SkeletonOptions): boolean;
 /**
  * Create a namespace guard instance for checking slug/handle uniqueness
  * across multiple database tables with reserved name protection.
@@ -225,4 +292,4 @@ declare function createNamespaceGuard(config: NamespaceConfig, adapter: Namespac
 /** The guard instance returned by `createNamespaceGuard`. */
 type NamespaceGuard = ReturnType<typeof createNamespaceGuard>;
 
-export { CONFUSABLE_MAP, type CheckResult, type FindOneOptions, type NamespaceAdapter, type NamespaceConfig, type NamespaceGuard, type NamespaceSource, type OwnershipScope, type SuggestStrategyName, createHomoglyphValidator, createNamespaceGuard, createProfanityValidator, normalize };
+export { CONFUSABLE_MAP, CONFUSABLE_MAP_FULL, type CheckResult, type FindOneOptions, type NamespaceAdapter, type NamespaceConfig, type NamespaceGuard, type NamespaceSource, type OwnershipScope, type SkeletonOptions, type SuggestStrategyName, areConfusable, createHomoglyphValidator, createNamespaceGuard, createProfanityValidator, normalize, skeleton };