Xpath Injection in Feathersjs

Xpath injection in Feathersjs applications typically occurs when user input is incorporated into XPath queries without proper sanitization. Feathersjs, being a flexible Node.js framework, allows developers to integrate with various data sources including databases that support XPath queries, such as XML-based storage systems or when using XPath within NoSQL queries.

The vulnerability manifests when Feathersjs services accept user input and directly inject it into XPath expressions. For example, consider a Feathersjs service that queries an XML database for user information:

const { authenticate } = require('@feathersjs/authentication');
const { BadRequest } = require('@feathersjs/errors');

class XmlService {
  async find(params) {
    const { username, password } = params.query;
    
    // Vulnerable XPath query construction
    const xpathQuery = `/users/user[username='${username}' and password='${password}']`;
    
    // Execute XPath query against XML data
    const result = await this.executeQuery(xpathQuery);
    
    return result;
  }
}

In this Feathersjs service, an attacker could submit a username like admin' or '1'='1 which would transform the XPath query into:

/users/user[username='admin' or '1'='1' and password='anything']

This bypasses authentication by always evaluating to true. Another common pattern in Feathersjs applications is when using XML-based configuration files or when integrating with SOAP services that use XPath internally:

const xpath = require('xpath');
const dom = require('xmldom').DOMParser;

class ConfigService {
  async getConfig(params) {
    const { configId } = params.query;
    
    // Vulnerable: direct interpolation of user input
    const xpathQuery = `/configurations/config[id='${configId}']`;
    
    const document = new dom().parseFromString(xmlData);
    const result = xpath.select(xpathQuery, document);
    
    return result;
  }
}

Feathersjs developers often fall into this trap when building custom services that need to query XML data or when working with XML-based APIs. The framework's flexibility means there's no built-in protection against XPath injection, making it the developer's responsibility to sanitize inputs.

Detecting XPath injection vulnerabilities in Feathersjs applications requires examining how user input flows through your services and gets incorporated into queries. middleBrick's black-box scanning approach is particularly effective for this, as it tests the actual runtime behavior of your Feathersjs endpoints without requiring source code access.

When middleBrick scans a Feathersjs API, it specifically looks for endpoints that might be vulnerable to XPath injection by:

• Testing common XPath injection patterns like single quotes, double quotes, and boolean operators
• Analyzing parameter names that suggest XML/XPath usage (username, id, query, path, etc.)
• Checking for XML-based content types in responses
• Examining error messages that might reveal XPath syntax

For developers who want to proactively scan their Feathersjs applications, middleBrick can be integrated directly into your development workflow:

# Scan your Feathersjs API from the CLI
middlebrick scan https://api.yourapp.com/users

# Integrate into your CI/CD pipeline
middlebrick scan --fail-below B --output json https://staging.yourapp.com

# Use the GitHub Action in your Feathersjs project
- name: Scan API Security
  uses: middleBrick/middleBrick-action@v1
  with:
    url: https://api.yourapp.com
    fail-on-grade-below: B

Within your Feathersjs codebase, you can identify potential XPath injection points by searching for patterns like:

// Search for these patterns in your codebase:
const xpath = require('xpath');
const xmldom = require('xmldom');
// String concatenation with user input
const query = `/path/to/element[name='${userInput}']`;
// Template literals with user input
const query = `/path/to/element[name="${userInput}"]`;

middleBrick's scoring system specifically flags these issues with severity levels based on the potential impact. A Feathersjs service that allows authentication bypass through XPath injection would receive a critical severity rating, while data exposure through XPath queries might be rated high severity.

Remediating XPath injection in Feathersjs applications requires a combination of input validation, parameterized queries, and proper escaping. Here are Feathersjs-specific approaches to secure your code:

1. Use Parameterized XPath Queries

Instead of string concatenation, use XPath's parameter substitution features:

const { BadRequest } = require('@feathersjs/errors');
const xpath = require('xpath');
const dom = require('xmldom').DOMParser;

class SecureXmlService {
  async find(params) {
    const { username, password } = params.query;
    
    // Validate and sanitize inputs
    if (!this.isValidInput(username) || !this.isValidInput(password)) {
      throw new BadRequest('Invalid input detected');
    }
    
    const document = new dom().parseFromString(xmlData);
    
    // Use parameterized XPath with proper escaping
    const xpathQuery = `//user[username=$username and password=$password]`;
    const result = xpath.select(xpathQuery, document, {
      username: username,
      password: password
    });
    
    return result;
  }
  
  isValidInput(input) {
    // Allow only alphanumeric and basic punctuation
    return /^[a-zA-Z0-9_.-]+$/.test(input);
  }
}

2. Input Validation and Whitelisting

Implement strict input validation in your Feathersjs services:

const { HookContext } = require('@feathersjs/feathers');

// Create a validation hook for XPath injection prevention
const validateXPathInput = () => {
  return async (context: HookContext) => {
    const { params } = context;
    
    // Define allowed characters
    const allowedPattern = /^[[email protected]]+$/;
    
    // Check all query parameters
    Object.values(params.query).forEach(value => {
      if (typeof value === 'string' && !allowedPattern.test(value)) {
        throw new Error('Input contains invalid characters');
      }
    });
    
    return context;
  };
};

// Apply to your Feathersjs service
class XmlService {
  setup(app) {
    this.before = {
      find: [validateXPathInput()]
    };
  }
}

3. Use Feathersjs Hooks for Centralized Security

Implement security checks as reusable hooks:

const { hooks } = require('@feathersjs/feathers');

// XPath injection prevention hook
const xpathSecurityHook = () => {
  return async (context) => {
    const { method, type, params } = context;
    
    // Only apply to external queries
    if (type === 'before' && method === 'find') {
      const query = params.query || {};
      
      // Check for suspicious patterns
      const suspiciousPatterns = [
        /'.*or.*=/i,
        /'.*and.*=/i,
        /'.*union.*'/i,
        /'.*select.*'/i
      ];
      
      for (const [key, value] of Object.entries(query)) {
        if (typeof value === 'string' && 
            suspiciousPatterns.some(pattern => pattern.test(value))) {
          throw new Error('Potential XPath injection detected');
        }
      }
    }
    
    return context;
  };
};

// Apply to all services
app.hooks({
  before: {
    all: [xpathSecurityHook()]
  }
});

4. Alternative: Use Safe XML Libraries

Consider using libraries that provide safer XML querying:

const fastXmlParser = require('fast-xml-parser');

class SafeXmlService {
  async find(params) {
    const { userId } = params.query;
    
    // Parse XML and query in memory
    const parsedData = fastXmlParser.parse(xmlData);
    
    // Use JavaScript filtering instead of XPath
    const result = parsedData.users.user.find(u => u.id === userId);
    
    return result;
  }
}