CVE-2025-67744

9.6 CRITICAL

📋 TL;DR

DeepChat versions before 0.5.3 contain a critical vulnerability where unsafe Mermaid diagram rendering allows arbitrary JavaScript execution. This XSS flaw escalates to full remote code execution due to exposed Electron IPC interfaces, enabling attackers to execute system commands. All users running vulnerable DeepChat instances are affected.

💻 Affected Systems

Products:
  • DeepChat
Versions: All versions before 0.5.3
Operating Systems: All platforms where DeepChat runs
Default Config Vulnerable: ⚠️ Yes
Notes: Any DeepChat instance with Mermaid diagram rendering enabled is vulnerable.

📦 What is this software?

⚠️ Risk & Real-World Impact

🔴

Worst Case

Complete system compromise allowing attacker to execute arbitrary commands, steal data, install malware, or pivot to other systems.

🟠

Likely Case

Attacker gains remote code execution on the DeepChat server, potentially accessing sensitive AI models, user data, and system resources.

🟢

If Mitigated

If proper network segmentation and least privilege are implemented, impact may be limited to the DeepChat application and its data.

🌐 Internet-Facing: HIGH
🏢 Internal Only: HIGH

🎯 Exploit Status

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

Exploitation requires crafting malicious Mermaid diagrams but does not require authentication.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: 0.5.3

Vendor Advisory: https://github.com/ThinkInAIXYZ/deepchat/security/advisories/GHSA-w8w8-82pv-5rg9

Restart Required: Yes

Instructions:

1. Stop DeepChat service. 2. Update to version 0.5.3 using package manager or manual installation. 3. Restart DeepChat service.

🔧 Temporary Workarounds

Disable Mermaid rendering

all

Temporarily disable Mermaid diagram rendering functionality

Modify DeepChat configuration to disable Mermaid diagram support

Network isolation

all

Restrict network access to DeepChat instances

Configure firewall rules to limit access to trusted IPs only

🧯 If You Can't Patch

  • Isolate DeepChat instances on separate network segments with strict firewall rules
  • Implement web application firewall (WAF) rules to block suspicious Mermaid content

🔍 How to Verify

Check if Vulnerable:

Check DeepChat version - if version is below 0.5.3, system is vulnerable

Check Version:

Check DeepChat configuration file or run: deepchat --version

Verify Fix Applied:

Verify DeepChat version is 0.5.3 or higher after update

📡 Detection & Monitoring

Log Indicators:

  • Unusual Mermaid diagram rendering requests
  • Suspicious JavaScript execution in logs
  • Unexpected system command execution

Network Indicators:

  • Unusual outbound connections from DeepChat server
  • Traffic patterns suggesting command and control

SIEM Query:

source="deepchat" AND (mermaid OR diagram) AND (javascript OR eval OR exec)

🔗 References

📤 Share & Export