CVE-2025-5276

7.4 HIGH

📋 TL;DR

This SSRF vulnerability in mcp-markdownify-server allows attackers to craft prompts that trick the server into making HTTP requests to attacker-controlled URLs when using webpage-to-markdown, bing-search-to-markdown, or youtube-to-markdown tools. This can lead to sensitive information disclosure from internal systems. All users of mcp-markdownify-server are affected.

💻 Affected Systems

Products:
  • mcp-markdownify-server
Versions: All versions
Operating Systems: All
Default Config Vulnerable: ⚠️ Yes
Notes: Vulnerability exists in default configuration when using the affected tools.

⚠️ Manual Verification Required

This CVE does not have specific version information in our database, so automatic vulnerability detection cannot determine if your system is affected.

Why? The CVE database entry doesn't specify which versions are vulnerable (no version ranges provided by the vendor/NVD).

🔒 Custom verification scripts are available for registered users. Sign up free to download automated test scripts.

Recommended Actions:
  1. Review the CVE details at NVD
  2. Check vendor security advisories for your specific version
  3. Test if the vulnerability is exploitable in your environment
  4. Consider updating to the latest version as a precaution

⚠️ Risk & Real-World Impact

🔴

Worst Case

Complete internal network reconnaissance, access to cloud metadata services, retrieval of sensitive internal API data, and potential lateral movement to other systems.

🟠

Likely Case

Information disclosure from internal HTTP services, access to cloud instance metadata, and potential credential leakage from internal APIs.

🟢

If Mitigated

Limited to external URL access only if proper network segmentation and egress filtering are implemented.

🌐 Internet-Facing: HIGH
🏢 Internal Only: MEDIUM

🎯 Exploit Status

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

Exploitation requires crafting malicious prompts that trigger the vulnerable functions, which is straightforward for attackers familiar with SSRF techniques.

🛠️ Fix & Mitigation

✅ Official Fix

Patch Version: Patched in commit 0284aa8f34d32c65e20d8cda2d429b7943c9af03

Vendor Advisory: https://security.snyk.io/vuln/SNYK-JS-MCPMARKDOWNIFYSERVER-10249387

Restart Required: Yes

Instructions:

1. Update to the latest version of mcp-markdownify-server. 2. Restart the service. 3. Verify the fix by checking that URL validation is now implemented.

🔧 Temporary Workarounds

Network Egress Filtering

all

Implement firewall rules to restrict outbound HTTP/HTTPS traffic from the server to only trusted external destinations.

Input Validation Proxy

all

Deploy a reverse proxy that validates and sanitizes all URLs before they reach the markdownify-server.

🧯 If You Can't Patch

  • Disable or restrict access to the vulnerable tools (webpage-to-markdown, bing-search-to-markdown, youtube-to-markdown)
  • Implement strict network segmentation to isolate the server from sensitive internal systems

🔍 How to Verify

Check if Vulnerable:

Check if your version is before commit 0284aa8f34d32c65e20d8cda2d429b7943c9af03 by examining the source code or package version.

Check Version:

npm list mcp-markdownify-server

Verify Fix Applied:

Test that the server now validates URLs and rejects requests to internal/private IP addresses and localhost.

📡 Detection & Monitoring

Log Indicators:

  • Unusual outbound HTTP requests from the server
  • Requests to internal IP addresses or localhost
  • Multiple failed URL validation attempts

Network Indicators:

  • HTTP requests from markdownify-server to unexpected internal endpoints
  • Traffic to cloud metadata services (169.254.169.254, etc.)
  • Outbound requests following prompt execution

SIEM Query:

source="markdownify-server" AND (dest_ip=PRIVATE_IP_RANGE OR dest_ip=127.0.0.1 OR dest_ip=169.254.169.254)

🔗 References

📤 Share & Export