Why Security Teams Choose AI-First

Your AI-first
security platform.

Transform your security operations with continuous threat modeling, agentic vulnerability triage and automated discovery of security gaps as your systems evolve.

Intelligence that evolves with you.

We constantly learn from your security operations, building sophisticated understanding that enhances protection.

Adaptive Threat Modeling

System Assets & Associated Threats

Detailed analysis of system components and their associated security threats

PDF Viewer Component

Frontend component using PDF.js for rendering customer documents like invoices and catalogs

Identified Threats

1 threats
VC: HIGH
VI: MEDIUM
VA: LOW
95% confidence

Confidentiality

HIGH

Integrity

MEDIUM

Availability

LOW

Mitigations

Update PDF.js to Latest Version

Upgrade PDF.js from version 3.11.174 to 4.2.67 or later to patch the vulnerability

98% effective

Implement Content Security Policy

Deploy strict CSP headers to prevent inline script execution and limit XSS impact

85% effective

PDF Viewer Sandboxing

Isolate PDF viewer in sandboxed iframe with restricted permissions

90% effective

Document Upload Service

Backend microservice handling document uploads, validation, and storage

Identified Threats

1 threats
VC: MEDIUM
VI: MEDIUM
VA: LOW
88% confidence

Confidentiality

MEDIUM

Integrity

MEDIUM

Availability

LOW

Mitigations

Enhanced File Validation

Implement deep content inspection and malware scanning for uploaded files

92% effective

File Type Restrictions

Enforce strict file type validation beyond MIME type checking

88% effective

Web Application Frontend

React-based customer interface with product browsing and checkout functionality

Identified Threats

1 threats
VC: HIGH
VI: HIGH
VA: MEDIUM
92% confidence

Confidentiality

HIGH

Integrity

HIGH

Availability

MEDIUM

Mitigations

Session Security Hardening

Implement secure session management with HttpOnly and SameSite cookie attributes

95% effective

Authentication State Validation

Add real-time validation of authentication state for sensitive operations

87% effective

Intelligent Vulnerability Triage

CVE-2024-4367

HIGH
NOT TRIAGED
View in NVD
Autonomous Analysis Enabled

Security Questionnaire

AI agent automatically analyzes the vulnerability and provides security assessment. Review findings and provide additional context where needed.

Completion Progress

1 of 2 questions answered • Triage will auto-complete when all questions are resolved

50%

Initial Question:

Analyze the codebase and deployment artifacts to determine if PDF.js library is present and what version is being used for document rendering functionality.

DISMISSED
AI Agent10:00:00 AM
⏳ Queued for autonomous analysis...
AI Agent10:30:00 AM
🚀 Autonomous analysis process has begun. I will now inspect the container images and gather evidence...
AI Agent11:00:00 AM
After comprehensive autonomous inspection of all available container images and deployment artifacts:

Analysis Results

PDF.js Dependencies Scan:
  • ❌ No PDF.js library found in package.json dependencies
  • ❌ No PDF.js imports detected in JavaScript/TypeScript files
  • ❌ No PDF rendering components found in React codebase
  • ❌ No PDF.js related configuration files detected
Alternative Document Handling:
  • ✅ Found native browser PDF viewer usage via <embed> and <iframe> tags
  • ✅ Server-side PDF generation using puppeteer for reports
  • ✅ All PDF processing happens server-side with no client-side PDF.js usage

Conclusion

CVE-2024-4367 is NOT APPLICABLE to this environment. The vulnerability specifically affects PDF.js library implementations, but this application:
  1. Does not use PDF.js library anywhere in the codebase
  2. Relies on native browser PDF rendering capabilities
  3. Uses server-side PDF generation that is not affected by this CVE
Risk Level: NONE - This vulnerability does not apply to the current architecture.

Initial Question:

Examine the production environment to identify any client-side PDF processing capabilities that might use vulnerable PDF.js versions through third-party integrations or CDN resources.

AGENT REPLIED
AI Agent11:30:00 AM
⏳ Queued for autonomous analysis...
AI Agent11:40:00 AM
🚀 Autonomous analysis process has begun. I will now inspect the container images and gather evidence...
AI Agent12:15:00 PM
After analyzing the production environment and third-party integrations:

Critical Finding: Vulnerable PDF.js Integration

DocuSign Integration Discovery:
  • 🚨 Found PDF.js v3.8.162 embedded in DocuSign web integration
  • 🚨 Vulnerable version - affects all versions before 4.2.67
  • 🚨 Customer-facing exposure through contract signing workflow
Technical Details:
// Found in production HTML
<script 
	src="https://demo.docusign.net/assets/pdf.js/3.8/pdf.min.js"
/>
// This version is vulnerable to CVE-2024-4367
Risk Assessment:
  • Severity: HIGH
  • Exposure: Customer contract signing process
  • Attack Vector: Malicious PDF uploaded to DocuSign could execute XSS
  • Impact: Session hijacking, data theft, unauthorized contract modifications

Recommended Remediation Plan

Immediate Actions (0-24 hours):
  1. Contact DocuSign support to request PDF.js version upgrade
  2. Implement additional CSP headers to restrict script execution
  3. Add iframe sandboxing for DocuSign integration
Short-term (1-7 days):
  1. Evaluate alternative e-signature providers with updated PDF.js
  2. Implement client-side validation for PDF uploads before DocuSign
  3. Add monitoring for suspicious PDF-related JavaScript execution
Long-term (1-4 weeks):
  1. Migrate to DocuSign API-only integration (no embedded PDF viewer)
  2. Implement comprehensive third-party dependency scanning
  3. Create security review process for all vendor integrations

Questions for Security Team

What is your current process for vetting third-party integrations? Understanding this will help prioritize the remediation approach and prevent similar issues. How critical is the embedded DocuSign PDF viewer to your business workflow? This affects whether we should pursue immediate workarounds or plan a more comprehensive migration.

Vulnerability Details

CVE ID

CVE-2024-4367

Description

A flaw was found in the PDF.js library where an attacker could execute arbitrary JavaScript when a crafted PDF is viewed in the browser, potentially leading to cross-site scripting (XSS) attacks.

Severity

HIGH

Status

NOT TRIAGED

Autonomous Analysis

AI agent automatically analyzes container images and codebase for vulnerability assessment

Agentic Discovery & Testing

Autonomous Discovery Summary

Real-time security assessment and penetration testing results

2
Critical Vulnerabilities
4
Recon Targets
15
Endpoints Scanned
6
Services Analyzed
Coverage Progress75% of infrastructure mapped
Last scan: Jan 15, 2024 • Next scan in 15 minutes

Recent Findings

6 total findings
https://shop.example.com/viewer/pdf
VULNERABILITY
Jan 15, 2024
Vulnerability: CVE-2024-4367
Risk Level: HIGH
Impact: XSS via crafted PDF document
https://api.shop.example.com/documents/upload
RECON TARGET
Jan 15, 2024
Risk Level: MEDIUM
Finding: Insufficient file validation
https://cdn.shop.example.com
RECON TARGET
Jan 15, 2024
Risk Level: MEDIUM
Finding: CDN security headers analysis
Impact: Missing CSP increases XSS vulnerability impact
https://api.shop.example.com/auth/token
VULNERABILITY
Jan 15, 2024
Vulnerability: Weak JWT Implementation
Risk Level: HIGH
Impact: JWT tokens can be cracked offline, session hijacking possible
postgresql://10.0.1.100:5432
RECON TARGET
Jan 15, 2024
Risk Level: MEDIUM
Finding: Database configuration assessment
redis://10.0.1.200:6379
RECON TARGET
Jan 15, 2024
Risk Level: LOW
Finding: Cache security analysis

Learns from your work, continuously

  • Understands specs, notes, tickets and commits
  • Evolves with feedback and refinements
  • Keeps system assets and risks up to date
Powerful Integrations

Built for modern security teams

Seamlessly integrate with your existing security stack and development workflow. Our AI-first platform adapts to your infrastructure and scales with your team.

and many more

Get In Touch

Ready to transform your security operations?

See how Primesight's AI-first platform automates threat modeling, streamlines vulnerability triage, and maintains real-time visibility into your attack surface..

Let's chat about it

Schedule a meet

Send us a message