CVE-2025-66648

7.2 HIGH

📋 TL;DR

CVE-2025-66648 is a cross-site scripting (XSS) vulnerability in vega-functions library that allows attackers to execute arbitrary JavaScript code by exploiting an internal function not part of the public API. This affects any application using vulnerable versions of vega-functions that processes untrusted user input. The vulnerability is particularly dangerous for web applications that allow user-supplied Vega expressions.

💻 Affected Systems

Products:
  • vega-functions
Versions: All versions prior to 6.1.1
Operating Systems: All
Default Config Vulnerable: ⚠️ Yes
Notes: Only affects applications that process untrusted user input through Vega expressions. Using vega.expressionInterpreter in CSP safe mode does NOT prevent this vulnerability.

📦 What is this software?

⚠️ Risk & Real-World Impact

🔴

Worst Case

Complete account takeover, session hijacking, data theft, or malware distribution through client-side code execution in users' browsers.

🟠

Likely Case

Stealing user sessions, performing actions on behalf of authenticated users, or defacing web applications.

🟢

If Mitigated

Limited impact if proper input validation and output encoding are implemented, though the vulnerability bypasses some standard XSS protections.

🌐 Internet-Facing: HIGH
🏢 Internal Only: MEDIUM

🎯 Exploit Status

Public PoC: ✅ No
Weaponized: UNKNOWN
Unauthenticated Exploit: ⚠️ Yes
Complexity: LOW

Exploitation requires the attacker to supply malicious input to a vulnerable Vega expression parser. No authentication is needed if the vulnerable endpoint is publicly accessible.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: 6.1.1

Vendor Advisory: https://github.com/vega/vega/security/advisories/GHSA-m9rg-mr6g-75gm

Restart Required: No

Instructions:

1. Update vega-functions dependency to version 6.1.1 or later. 2. For npm: run 'npm update vega-functions'. 3. For yarn: run 'yarn upgrade vega-functions'. 4. Rebuild and redeploy your application.

🔧 Temporary Workarounds

No workaround available

all

The vendor advisory states there is no workaround besides upgrading. Input validation and CSP do not prevent this specific vulnerability.

🧯 If You Can't Patch

  • Disable user-supplied Vega expression functionality entirely
  • Implement strict input validation and sanitization for all user inputs that could reach Vega expression parser

🔍 How to Verify

Check if Vulnerable:

Check package.json or package-lock.json for vega-functions version. If version is below 6.1.1, the system is vulnerable.

Check Version:

npm list vega-functions | grep vega-functions

Verify Fix Applied:

Verify vega-functions version is 6.1.1 or higher in package.json and that the updated version is actually being used in production.

📡 Detection & Monitoring

Log Indicators:

  • Unusual Vega expression patterns in user inputs
  • JavaScript code patterns in Vega expression logs
  • Multiple failed expression parsing attempts

Network Indicators:

  • POST/PUT requests containing suspicious JavaScript patterns in Vega expression parameters
  • Unusual content-type headers for Vega expression endpoints

SIEM Query:

source="web_logs" AND (vega OR expression) AND (javascript OR script OR eval OR alert)

🔗 References

📤 Share & Export