PraisonAI hat ein Beliebiges Datei Schreiben (Zip Slip) bei Vorlagen Extraktion

The Arbitrary File Access vulnerability in PraisonAI arises from the insecure use of Python's zipfile.extractall() function during template installation. When a user downloads and extracts template archives (e.g., from GitHub), the application fails to properly validate the file paths within the archive. This allows an attacker to craft a malicious archive containing files with strategically crafted paths that, when extracted, overwrite critical system files or inject malicious code into the PraisonAI installation directory. Successful exploitation could lead to remote code execution, complete system compromise, and data exfiltration. The potential blast radius extends to any system running a vulnerable version of PraisonAI and utilizing external template sources.

Organizations using PraisonAI for collaborative AI development and deployment are at risk, particularly those who allow users to import templates from external sources like GitHub. Shared hosting environments where multiple users share the same PraisonAI installation are also at increased risk, as a compromised template from one user could potentially affect other users.

• python: Monitor PraisonAI logs for suspicious file extraction activity, particularly attempts to write files outside the designated template directory. Look for patterns involving zipfile.extractall() and unusual file paths.

# Example: Monitor for file extraction attempts
import os
import logging

logging.basicConfig(filename='praisona.log', level=logging.INFO)

for root, _, files in os.walk('/path/to/praisona/templates'): # Replace with actual path
    for file in files:
        logging.info(f'File accessed: {os.path.join(root, file)}')

• generic web: Monitor web server access logs for requests to download template archives from external sources. Analyze the downloaded archives for suspicious filenames or directory traversal sequences. • generic web: Check PraisonAI's configuration files for any settings related to template download locations or extraction directories. Ensure these settings are properly secured and restricted.

1. 2026-04-07Disclosure

   disclosure

1. Reserviert2026-04-06
2. Veröffentlicht2026-04-07
3. Geändert2026-04-09
4. EPSS aktualisiert2026-04-15

The primary mitigation for CVE-2026-39307 is to immediately upgrade PraisonAI to version 1.5.113 or later, which includes the fix for this vulnerability. If upgrading is not immediately feasible, consider implementing a temporary workaround by disabling the automatic template installation feature and manually verifying the contents of any downloaded template archives before extraction. Implement strict file access controls on the PraisonAI installation directory to limit the potential impact of a successful exploit. Consider using a Web Application Firewall (WAF) to filter potentially malicious template archives based on file extensions or path patterns. After upgrading, confirm the fix by attempting to extract a known malicious archive and verifying that the files are extracted only within the intended directory.