CVE-2026-22792

9.6 CRITICAL

📋 TL;DR

This vulnerability allows attackers to inject malicious HTML that executes arbitrary JavaScript in the 5ire desktop AI assistant renderer context. This JavaScript can then call exposed bridge APIs to create unauthorized MCP servers, potentially leading to remote command execution. All users running 5ire versions before 0.15.3 are affected.

💻 Affected Systems

Products:
  • 5ire desktop AI assistant
Versions: All versions prior to 0.15.3
Operating Systems: Windows, macOS, Linux
Default Config Vulnerable: ⚠️ Yes
Notes: Cross-platform vulnerability affecting all operating systems where 5ire runs.

📦 What is this software?

⚠️ Risk & Real-World Impact

🔴

Worst Case

Remote command execution on the victim's system, allowing full system compromise, data theft, and lateral movement.

🟠

Likely Case

Unauthorized creation of MCP servers leading to arbitrary code execution within the application context.

🟢

If Mitigated

Limited impact if proper input validation and sandboxing prevent JavaScript execution.

🌐 Internet-Facing: HIGH - Attackers can exploit via crafted content delivered through the application's interfaces.
🏢 Internal Only: MEDIUM - Requires user interaction with malicious content but can be exploited internally.

🎯 Exploit Status

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

Exploitation requires user interaction with malicious content but uses simple HTML injection techniques.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: 0.15.3

Vendor Advisory: https://github.com/nanbingxyz/5ire/security/advisories/GHSA-p5fm-wm8g-rffx

Restart Required: Yes

Instructions:

1. Download version 0.15.3 from https://github.com/nanbingxyz/5ire/releases/tag/v0.15.3
2. Install the update
3. Restart the application

🔧 Temporary Workarounds

Disable HTML rendering features

all

Temporarily disable or restrict HTML rendering capabilities in the application if available.

Network segmentation

all

Restrict network access to prevent external exploitation attempts.

🧯 If You Can't Patch

  • Isolate affected systems from untrusted networks and content sources.
  • Implement strict content filtering for all inputs to the application.

🔍 How to Verify

Check if Vulnerable:

Check the application version in settings or about dialog.

Check Version:

Check application settings or run with --version flag if available.

Verify Fix Applied:

Confirm version is 0.15.3 or later in the application settings.

📡 Detection & Monitoring

Log Indicators:

  • Unusual MCP server creation events
  • JavaScript execution errors in renderer logs

Network Indicators:

  • Unexpected network connections from 5ire process
  • MCP server creation requests

SIEM Query:

Process creation events from 5ire.exe or similar with unusual command-line arguments

🔗 References

📤 Share & Export