feat: add detect sensitive info rule

.github/.release-please-manifest.json

@@ -6,6 +6,7 @@
   "arcjet-next": "1.0.0-alpha.21",
   "arcjet-node": "1.0.0-alpha.21",
   "arcjet-sveltekit": "1.0.0-alpha.21",
+  "body": "1.0.0-alpha.21",
   "decorate": "1.0.0-alpha.21",
   "duration": "1.0.0-alpha.21",
   "env": "1.0.0-alpha.21",

.github/dependabot.yml

@@ -109,6 +109,31 @@ updates:
       - dependency-name: eslint
         versions: [">=9"]
 
+  - package-ecosystem: npm
+    directory: /examples/nextjs-14-sensitive-info
+    schedule:
+      # Our dependencies should be checked daily
+      interval: daily
+    assignees:
+      - blaine-arcjet
+    reviewers:
+      - blaine-arcjet
+    commit-message:
+      prefix: deps(example)
+      prefix-development: deps(example)
+    groups:
+      dependencies:
+        patterns:
+          - "*"
+    ignore:
+      # Ignore updates to the @types/node package due to conflict between
+      # Headers in DOM.
+      - dependency-name: "@types/node"
+        versions: [">18.18"]
+      # TODO(#539): Upgrade to eslint 9
+      - dependency-name: eslint
+        versions: [">=9"]
+
   - package-ecosystem: npm
     directory: /examples/nextjs-14-app-dir-validate-email
     schedule:
@@ -405,6 +430,23 @@ updates:
         patterns:
           - "*"
 
+  - package-ecosystem: npm
+    directory: /examples/express-sensitive-info
+    schedule:
+      # Our dependencies should be checked daily
+      interval: daily
+    assignees:
+      - blaine-arcjet
+    reviewers:
+      - blaine-arcjet
+    commit-message:
+      prefix: deps(example)
+      prefix-development: deps(example)
+    groups:
+      dependencies:
+        patterns:
+          - "*"
+
   - package-ecosystem: npm
     directory: /examples/nodejs-express-launchdarkly
     schedule:

.github/release-please-config.json

@@ -52,6 +52,10 @@
       "component": "@arcjet/sveltekit",
       "skip-github-release": true
     },
+    "body": {
+      "component": "@arcjet/body",
+      "skip-github-release": true
+    },
     "decorate": {
       "component": "@arcjet/decorate",
       "skip-github-release": true
@@ -127,6 +131,7 @@
         "@arcjet/eslint-config",
         "@arcjet/headers",
         "@arcjet/ip",
+        "@arcjet/body",
         "@arcjet/logger",
         "@arcjet/protocol",
         "@arcjet/rollup-config",

.github/workflows/reusable-examples.yml

@@ -467,6 +467,48 @@ jobs:
         working-directory: examples/nextjs-14-permit
         run: npm run build
 
+  nextjs-14-sensitive-info:
+    name: Next.js 14 + Sensitive Info
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+    steps:
+      # Environment security
+      - name: Harden Runner
+        uses: step-security/harden-runner@5c7944e73c4c2a096b17a9cb74d65b6c2bbafbde # v2.9.1
+        with:
+          disable-sudo: true
+          egress-policy: block
+          allowed-endpoints: >
+            fonts.googleapis.com:443
+            fonts.gstatic.com:443
+            github.com:443
+            registry.npmjs.org:443
+
+      # Checkout
+      # Most toolchains require checkout first
+      - name: Checkout
+        uses: actions/checkout@v4
+
+      # Language toolchains
+      - name: Install Node
+        uses: actions/[email protected]
+        with:
+          node-version: 20
+
+      # Workflow
+
+      - name: Install dependencies
+        run: npm ci
+
+      - name: Install example dependencies
+        working-directory: examples/nextjs-14-sensitive-info
+        run: npm ci
+
+      - name: Build
+        working-directory: examples/nextjs-14-sensitive-info
+        run: npm run build
+
   nodejs-hono-rl:
     name: Node.js + Hono + Rate Limit
     runs-on: ubuntu-latest

analyze/edge-light.ts

@@ -1,13 +1,18 @@
 import type { ArcjetLogger, ArcjetRequestDetails } from "@arcjet/protocol";
 
-import * as core from "./wasm/arcjet_analyze_js_req.component.js";
+import { instantiate } from "./wasm/arcjet_analyze_js_req.component.js";
 import type {
   ImportObject,
   EmailValidationConfig,
   BotDetectionResult,
   BotType,
   EmailValidationResult,
+  DetectedSensitiveInfoEntity,
+  SensitiveInfoEntities,
+  SensitiveInfoEntity,
+  SensitiveInfoResult,
 } from "./wasm/arcjet_analyze_js_req.component.js";
+import type { ArcjetJsReqSensitiveInformationIdentifier } from "./wasm/interfaces/arcjet-js-req-sensitive-information-identifier.js";
 
 import componentCoreWasm from "./wasm/arcjet_analyze_js_req.component.core.wasm?module";
 import componentCore2Wasm from "./wasm/arcjet_analyze_js_req.component.core2.wasm?module";
@@ -26,6 +31,9 @@ interface AnalyzeContext {
   characteristics: string[];
 }
 
+type DetectSensitiveInfoFunction =
+  typeof ArcjetJsReqSensitiveInformationIdentifier.detect;
+
 async function moduleFromPath(path: string): Promise<WebAssembly.Module> {
   if (path === "arcjet_analyze_js_req.component.core.wasm") {
     return componentCoreWasm;
@@ -40,9 +48,20 @@ async function moduleFromPath(path: string): Promise<WebAssembly.Module> {
   throw new Error(`Unknown path: ${path}`);
 }
 
-async function init(context: AnalyzeContext) {
+function noOpDetect(): SensitiveInfoEntity[] {
+  return [];
+}
+
+async function init(
+  context: AnalyzeContext,
+  detectSensitiveInfo?: DetectSensitiveInfoFunction,
+) {
   const { log } = context;
 
+  if (typeof detectSensitiveInfo !== "function") {
+    detectSensitiveInfo = noOpDetect;
+  }
+
   const coreImports: ImportObject = {
     "arcjet:js-req/logger": {
       debug(msg) {
@@ -69,10 +88,13 @@ async function init(context: AnalyzeContext) {
         return "unknown";
       },
     },
+    "arcjet:js-req/sensitive-information-identifier": {
+      detect: detectSensitiveInfo,
+    },
   };
 
   try {
-    return core.instantiate(moduleFromPath, coreImports);
+    return instantiate(moduleFromPath, coreImports);
   } catch {
     log.debug("WebAssembly is not supported in this runtime");
   }
@@ -94,6 +116,9 @@ export {
    * almost certain this request was not a bot.
    */
   type BotDetectionResult,
+  type DetectedSensitiveInfoEntity,
+  type SensitiveInfoEntity,
+  type DetectSensitiveInfoFunction,
 };
 
 /**
@@ -161,3 +186,25 @@ export async function detectBot(
     };
   }
 }
+export async function detectSensitiveInfo(
+  context: AnalyzeContext,
+  candidate: string,
+  entities: SensitiveInfoEntities,
+  contextWindowSize: number,
+  detect?: DetectSensitiveInfoFunction,
+): Promise<SensitiveInfoResult> {
+  const analyze = await init(context, detect);
+
+  if (typeof analyze !== "undefined") {
+    const skipCustomDetect = typeof detect !== "function";
+    return analyze.detectSensitiveInfo(candidate, {
+      entities,
+      contextWindowSize,
+      skipCustomDetect,
+    });
+  } else {
+    throw new Error(
+      "SENSITIVE_INFO rule failed to run because Wasm is not supported in this environment.",
+    );
+  }
+}

analyze/index.ts

@@ -1,13 +1,18 @@
 import type { ArcjetLogger, ArcjetRequestDetails } from "@arcjet/protocol";
 
-import * as core from "./wasm/arcjet_analyze_js_req.component.js";
+import { instantiate } from "./wasm/arcjet_analyze_js_req.component.js";
 import type {
   ImportObject,
   EmailValidationConfig,
   BotDetectionResult,
   BotType,
   EmailValidationResult,
+  DetectedSensitiveInfoEntity,
+  SensitiveInfoEntities,
+  SensitiveInfoEntity,
+  SensitiveInfoResult,
 } from "./wasm/arcjet_analyze_js_req.component.js";
+import type { ArcjetJsReqSensitiveInformationIdentifier } from "./wasm/interfaces/arcjet-js-req-sensitive-information-identifier.js";
 
 import { wasm as componentCoreWasm } from "./wasm/arcjet_analyze_js_req.component.core.wasm?js";
 import { wasm as componentCore2Wasm } from "./wasm/arcjet_analyze_js_req.component.core2.wasm?js";
@@ -26,6 +31,9 @@ interface AnalyzeContext {
   characteristics: string[];
 }
 
+type DetectSensitiveInfoFunction =
+  typeof ArcjetJsReqSensitiveInformationIdentifier.detect;
+
 // TODO: Do we actually need this wasmCache or does `import` cache correctly?
 const wasmCache = new Map<string, WebAssembly.Module>();
 
@@ -54,9 +62,20 @@ async function moduleFromPath(path: string): Promise<WebAssembly.Module> {
   throw new Error(`Unknown path: ${path}`);
 }
 
-async function init(context: AnalyzeContext) {
+function noOpDetect(): SensitiveInfoEntity[] {
+  return [];
+}
+
+async function init(
+  context: AnalyzeContext,
+  detectSensitiveInfo?: DetectSensitiveInfoFunction,
+) {
   const { log } = context;
 
+  if (typeof detectSensitiveInfo !== "function") {
+    detectSensitiveInfo = noOpDetect;
+  }
+
   const coreImports: ImportObject = {
     "arcjet:js-req/logger": {
       debug(msg) {
@@ -83,10 +102,13 @@ async function init(context: AnalyzeContext) {
         return "unknown";
       },
     },
+    "arcjet:js-req/sensitive-information-identifier": {
+      detect: detectSensitiveInfo,
+    },
   };
 
   try {
-    return core.instantiate(moduleFromPath, coreImports);
+    return instantiate(moduleFromPath, coreImports);
   } catch {
     log.debug("WebAssembly is not supported in this runtime");
   }
@@ -108,6 +130,9 @@ export {
    * almost certain this request was not a bot.
    */
   type BotDetectionResult,
+  type DetectedSensitiveInfoEntity,
+  type SensitiveInfoEntity,
+  type DetectSensitiveInfoFunction,
 };
 
 /**
@@ -175,3 +200,26 @@ export async function detectBot(
     };
   }
 }
+
+export async function detectSensitiveInfo(
+  context: AnalyzeContext,
+  candidate: string,
+  entities: SensitiveInfoEntities,
+  contextWindowSize: number,
+  detect?: DetectSensitiveInfoFunction,
+): Promise<SensitiveInfoResult> {
+  const analyze = await init(context, detect);
+
+  if (typeof analyze !== "undefined") {
+    const skipCustomDetect = typeof detect !== "function";
+    return analyze.detectSensitiveInfo(candidate, {
+      entities,
+      contextWindowSize,
+      skipCustomDetect,
+    });
+  } else {
+    throw new Error(
+      "SENSITIVE_INFO rule failed to run because Wasm is not supported in this environment.",
+    );
+  }
+}

analyze/wasm/arcjet_analyze_js_req.component.d.ts

@@ -1,3 +1,5 @@
+import type { SensitiveInfoEntity } from './interfaces/arcjet-js-req-sensitive-information-identifier.js';
+export { SensitiveInfoEntity };
 /**
 * # Variants
 * 
@@ -35,16 +37,42 @@ export interface EmailValidationConfig {
   allowDomainLiteral: boolean,
   blockedEmails: Array<string>,
 }
+export type SensitiveInfoEntities = SensitiveInfoEntitiesAllow | SensitiveInfoEntitiesDeny;
+export interface SensitiveInfoEntitiesAllow {
+  tag: 'allow',
+  val: Array<SensitiveInfoEntity>,
+}
+export interface SensitiveInfoEntitiesDeny {
+  tag: 'deny',
+  val: Array<SensitiveInfoEntity>,
+}
+export interface SensitiveInfoConfig {
+  entities: SensitiveInfoEntities,
+  contextWindowSize?: number,
+  skipCustomDetect: boolean,
+}
+export interface DetectedSensitiveInfoEntity {
+  start: number,
+  end: number,
+  identifiedType: SensitiveInfoEntity,
+}
+export interface SensitiveInfoResult {
+  allowed: Array<DetectedSensitiveInfoEntity>,
+  denied: Array<DetectedSensitiveInfoEntity>,
+}
 import { ArcjetJsReqEmailValidatorOverrides } from './interfaces/arcjet-js-req-email-validator-overrides.js';
 import { ArcjetJsReqLogger } from './interfaces/arcjet-js-req-logger.js';
+import { ArcjetJsReqSensitiveInformationIdentifier } from './interfaces/arcjet-js-req-sensitive-information-identifier.js';
 export interface ImportObject {
   'arcjet:js-req/email-validator-overrides': typeof ArcjetJsReqEmailValidatorOverrides,
   'arcjet:js-req/logger': typeof ArcjetJsReqLogger,
+  'arcjet:js-req/sensitive-information-identifier': typeof ArcjetJsReqSensitiveInformationIdentifier,
 }
 export interface Root {
   detectBot(headers: string, patternsAdd: string, patternsRemove: string): BotDetectionResult,
   generateFingerprint(request: string, characteristics: Array<string>): string,
   isValidEmail(candidate: string, options: EmailValidationConfig): EmailValidationResult,
+  detectSensitiveInfo(content: string, options: SensitiveInfoConfig): SensitiveInfoResult,
 }
 
 /**