BETA Shoulder ist in der Beta — Befunde können manchmal falsch sein. Dein Feedback bestimmt, was wir als Nächstes beheben. Feedback teilen
👁️

Exposure of Sensitive Information to an Unauthorized Actor

🛡️ 14 Regeln erkennen dies

Exposure of Sensitive Information to an Unauthorized Actor

The product exposes sensitive information to an actor that is not explicitly authorized to have access to that information.

There are many different kinds of mistakes that introduce information exposures. The severity of the error can range widely, depending on the context in which the product operates, the type of sensitive information that is revealed, and the benefits it may provide to an attacker.

Verbreitung
Hoch
Häufig ausgenutzt
Auswirkung
Kritisch
4 Regeln mit kritischem Schweregrad
Prävention
Dokumentiert
14 Fix-Beispiele
2 Prävention
2 Prävention

So behebst du diese Schwachstelle

Präventionsstrategien für Information Exposure basierend auf 14 Shoulder-Erkennungsregeln.

Environment Variable Secret Exposure HIGH

Use environment variables for configuration only; never log or return their values

+5 -1 go
  package main
  
  import (
      "log"
      "os"
  )
  
  func main() {
      apiKey := os.Getenv("API_KEY")
-     log.Printf("API Key: %s", apiKey)
+     if apiKey == "" {
+         log.Fatal("API_KEY not configured")
+     }
+     log.Println("API key configured:", len(apiKey) > 0)
+     // Use apiKey internally, never log or return it
  }
  
LLM Model Theft HIGH

Store API keys in environment variables, never log them, and protect model endpoints with authentication

+5 -3 go
- client := openai.NewClient("sk-proj-1234567890abcdefghijklmnop")
- log.Printf("Using key: %s", apiKey)
- http.Handle("/models/", http.FileServer(http.Dir("./models")))
+ client := openai.NewClient(os.Getenv("OPENAI_API_KEY"))
+ log.Printf("Request completed: model=%s tokens=%d", model, usage.TotalTokens)
+ 
+ modelsHandler := http.FileServer(http.Dir("./models"))
+ http.Handle("/models/", authMiddleware(rateLimiter(modelsHandler)))
  
LLM Sensitive Information Disclosure HIGH

Mask PII and redact credentials before sending data to LLM APIs, and use structured logging

+7 -6 go
- resp, _ := client.CreateChatCompletion(ctx, openai.ChatCompletionRequest{
-     Messages: []openai.ChatCompletionMessage{{
-         Content: fmt.Sprintf("User SSN: %s, Password: %s", user.SSN, password),
-     }},
- })
- log.Printf("Request: %v", messages)
+ safeMessage := maskPII(userInput)
+ safeMessage = redactCredentials(safeMessage)
+ 
+ resp, _ := client.CreateChatCompletion(ctx, openai.ChatCompletionRequest{
+     Messages: []openai.ChatCompletionMessage{{Content: safeMessage}},
+ })
+ log.Printf("Completed: model=%s tokens=%d", model, resp.Usage.TotalTokens)
  
Environment Variable Secret Exposure HIGH

Use secrets internally without exposing them in logs, responses, or client-side code

+4 -3 javascript
- app.get('/config', (req, res) => {
-   res.json({ apiKey: process.env.API_KEY });
- });
+ const apiKey = process.env.API_KEY;
+ // Use the key server-side only
+ const result = await externalApi.call({ key: apiKey });
+ res.json({ data: result });
  
LLM Model Theft HIGH

Load API keys from environment variables and proxy LLM calls through your server

+1 -1 javascript
  const openai = new OpenAI({
-   apiKey: 'sk-proj-1234567890abcdefghijklmnop'
+   apiKey: process.env.OPENAI_API_KEY
  });
  
LLM Sensitive Information Disclosure HIGH

Mask or redact PII and credentials before sending data to LLM APIs

+3 -2 javascript
- const response = await openai.chat.completions.create({
-   messages: [{ role: 'user', content: `Process: ${userRecord}` }]
+ const masked = maskPII(userRecord);
+ const response = await openai.chat.completions.create({
+   messages: [{ role: 'user', content: `Process: ${masked}` }]
  });
  
Information Disclosure MEDIUM

Return generic error messages to users; log detailed errors server-side only

+12 -8 python
- from flask import jsonify
- 
- @app.route('/api/data')
- def handler():
-     try:
-         return jsonify(process())
-     except Exception as e:
-         return jsonify({'error': str(e), 'trace': traceback.format_exc()}), 500
+ import logging
+ from flask import jsonify
+ 
+ logger = logging.getLogger(__name__)
+ 
+ @app.route('/api/data')
+ def handler():
+     try:
+         return jsonify(process())
+     except Exception as e:
+         logger.error(f"Error: {e}", exc_info=True)
+         return jsonify({'error': 'Internal server error'}), 500
  
LLM Sensitive Information Disclosure HIGH

Use Presidio or similar libraries to anonymize PII before sending data to LLM APIs

+14 -6 python
- messages = [{
-     'role': 'user',
-     'content': f"User SSN: {user.ssn}, email: {user.email}. Summarize profile."
- }]
- logging.info(f"Request: {messages}")
- response = openai.chat.completions.create(model='gpt-4', messages=messages)
+ from presidio_analyzer import AnalyzerEngine
+ from presidio_anonymizer import AnonymizerEngine
+ 
+ analyzer = AnalyzerEngine()
+ anonymizer = AnonymizerEngine()
+ 
+ def anonymize_text(text: str) -> str:
+     results = analyzer.analyze(text=text, language='en')
+     return anonymizer.anonymize(text=text, analyzer_results=results).text
+ 
+ safe_message = anonymize_text(user_message)
+ messages = [{'role': 'user', 'content': safe_message}]
+ response = openai.chat.completions.create(model='gpt-4', messages=messages)
+ logger.info('Completed', extra={'model': 'gpt-4', 'tokens': response.usage.total_tokens})
  
Sensitive Field Exposure in API Response CRITICAL

Use explicit field selection or serializer schemas to exclude sensitive fields from responses

+5 -1 python
  from flask import jsonify
  from models import User
  
  @app.route('/api/users')
  def get_users():
      users = User.query.all()
-     return jsonify([u.__dict__ for u in users])
+     return jsonify([{
+         'id': u.id,
+         'email': u.email,
+         'name': u.name
+     } for u in users])
  
3 Erkennung
3 Erkennung

Finden Sie Schwachstellen in Ihrem Code

Verwenden Sie Shoulder, um Ihren Code nach Exposure of Sensitive Information to an Unauthorized Actor-Mustern zu scannen. 14 Regeln.

Terminal
# Scan with Shoulder CLI
npx @shoulderdev/cli trust --cwe=200

# Or scan entire project
npx @shoulderdev/cli trust .

Erkennungsregeln (14)

🟨 Javascript 5 rules
Environment Variable Secret Exposure HIGH
Detects when environment variables (which may contain secrets like API keys, passwords, tokens) are leaked through logging, HTTP responses, or external requests. Environment variables commonly store sensitive data: - API keys (AWS_ACCESS_KEY_ID, STRIPE_SECRET_KEY) - Database passwords (DB_PASSWORD, DATABASE_URL) - JWT secrets (JWT_SECRET) - OAuth tokens (GITHUB_TOKEN, SLACK_TOKEN) Leaking these values exposes credentials and allows unauthorized access. This rule uses taint flow analysis to de
LLM Model Theft HIGH
Detects vulnerabilities that could lead to model theft or API key exposure. OWASP LLM10 - Model Theft. Model theft can occur through: - API key exposure in client-side code or logs - Model weights exposed via insecure endpoints - Model extraction attacks via unrestricted API access - Insecure model serialization and storage - Missing access controls on model endpoints This rule detects: - Hardcoded API keys in source code - API keys in client-side JavaScript - Model files served without authen
LLM Sensitive Information Disclosure HIGH
Detects potential sensitive information disclosure in AI/LLM implementations. OWASP LLM06 - Sensitive Information Disclosure. Sensitive information can be leaked through: - PII (Personal Identifiable Information) in prompts - Credentials or secrets in prompts or system messages - Sensitive business data sent to third-party LLM APIs - Logging LLM conversations containing sensitive data - LLM responses exposed without filtering This rule detects: - Sensitive data patterns in LLM prompts - Creden
Sensitive Field Exposure in API Response CRITICAL
Detects when sensitive data fields (passwords, tokens, secrets, API keys) are exposed through API endpoint responses. This commonly happens when: 1. Mapping user data with sensitive fields: `.map(u => ({ password: u.password }))` 2. Returning entire user objects: `res.json(user)` where user has password field 3. Including sensitive fields in response objects: `res.json({ password: user.password })` This is particularly dangerous when AI-generated code returns user collections without filtering
Prisma Sensitive Field Exposure CRITICAL
Prisma returns all fields by default. Without 'select' or 'omit', password hashes and API tokens can leak to clients.
🔷 Typescript 5 rules
Environment Variable Secret Exposure HIGH
Detects when environment variables (which may contain secrets like API keys, passwords, tokens) are leaked through logging, HTTP responses, or external requests. Environment variables commonly store sensitive data: - API keys (AWS_ACCESS_KEY_ID, STRIPE_SECRET_KEY) - Database passwords (DB_PASSWORD, DATABASE_URL) - JWT secrets (JWT_SECRET) - OAuth tokens (GITHUB_TOKEN, SLACK_TOKEN) Leaking these values exposes credentials and allows unauthorized access. This rule uses taint flow analysis to de
LLM Model Theft HIGH
Detects vulnerabilities that could lead to model theft or API key exposure. OWASP LLM10 - Model Theft. Model theft can occur through: - API key exposure in client-side code or logs - Model weights exposed via insecure endpoints - Model extraction attacks via unrestricted API access - Insecure model serialization and storage - Missing access controls on model endpoints This rule detects: - Hardcoded API keys in source code - API keys in client-side JavaScript - Model files served without authen
LLM Sensitive Information Disclosure HIGH
Detects potential sensitive information disclosure in AI/LLM implementations. OWASP LLM06 - Sensitive Information Disclosure. Sensitive information can be leaked through: - PII (Personal Identifiable Information) in prompts - Credentials or secrets in prompts or system messages - Sensitive business data sent to third-party LLM APIs - Logging LLM conversations containing sensitive data - LLM responses exposed without filtering This rule detects: - Sensitive data patterns in LLM prompts - Creden
Sensitive Field Exposure in API Response CRITICAL
Detects when sensitive data fields (passwords, tokens, secrets, API keys) are exposed through API endpoint responses. This commonly happens when: 1. Mapping user data with sensitive fields: `.map(u => ({ password: u.password }))` 2. Returning entire user objects: `res.json(user)` where user has password field 3. Including sensitive fields in response objects: `res.json({ password: user.password })` This is particularly dangerous when AI-generated code returns user collections without filtering
Prisma Sensitive Field Exposure CRITICAL
Prisma returns all fields by default. Without 'select' or 'omit', password hashes and API tokens can leak to clients.
🐍 Python 5 rules
Information Disclosure MEDIUM
Detects information disclosure vulnerabilities: debug mode enabled, exposing stack traces, returning detailed error messages, or leaking sensitive data.
LLM Model Theft HIGH
Detects vulnerabilities that could lead to model theft or API key exposure. OWASP LLM10 - Model Theft. Model theft can occur through: - API key exposure in source code or logs - Model weights exposed via insecure endpoints - Model extraction attacks via unrestricted API access - Insecure model serialization and storage
LLM Sensitive Information Disclosure HIGH
Detects potential sensitive information disclosure in AI/LLM implementations. OWASP LLM06 - Sensitive Information Disclosure. Sensitive information can be leaked through: - PII (Personal Identifiable Information) in prompts - Credentials or secrets in prompts or system messages - Sensitive business data sent to third-party LLM APIs - Logging LLM conversations containing sensitive data
Sensitive Field Exposure in API Response CRITICAL
Detects when sensitive data fields (passwords, tokens, secrets) are exposed through API endpoint responses. This commonly happens when: 1. Returning user dictionaries with sensitive fields 2. Serializing ORM objects without excluding sensitive fields 3. Including sensitive fields in JSON responses Security Impact: - Password hash exposure enabling offline cracking attacks - API key/token leakage allowing account takeover - Session token exposure enabling session hijacking - PII disclosure viol
Server Information Disclosure LOW
Detects server configuration that exposes version information, framework details, or other implementation details through HTTP headers. This information helps attackers identify known vulnerabilities in specific versions.
4 Warnzeichen
4 Warnzeichen

Worauf bei Code-Reviews zu achten ist

Diese Muster weisen auf potenzielle Exposure of Sensitive Information to an Unauthorized Actor-Schwachstellen hin. Achten Sie bei Code-Reviews und Sicherheitsaudits darauf.

🟠
Model theft vulnerability: ... go-llm-model-theft
🟠
vulnerabilities leading to model theft or API key exposure such as hardcoded keys or insecure model go-llm-model-theft
🟠
Potential sensitive information disclosure: ... go-llm-sensitive-info-disclosure
🟠
sensitive information disclosure in AI/LLM implementations such as credentials or PII in prompts go-llm-sensitive-info-disclosure
🟠
when environment variables (which may contain secrets like API keys, passwords, tokens) are leaked t javascript-env-vars-secret-exposure
🟠
vulnerabilities that could lead to model theft or API key exposure javascript-llm-model-theft
🟠
potential sensitive information disclosure in AI/LLM implementations javascript-llm-sensitive-info-disclosure
🟡
information disclosure vulnerabilities: debug mode enabled, exposing stack traces, returning detaile python-information-disclosure
🔍

Scanne deine Codebasis nach Exposure of Sensitive Information to an Unauthorized Actor

Shoulder CLI findet anfällige Muster in deiner gesamten Codebasis.