CVE-2024-12909
📋 TL;DR
This SQL injection vulnerability in the FinanceChatLlamaPack allows attackers to execute arbitrary SQL queries through the database_agent's run_sql_query function. Exploitation can lead to remote code execution via PostgreSQL large object functionality. Users of run-llama/llama_index repository versions up to v0.12.3 are affected.
💻 Affected Systems
- run-llama/llama_index FinanceChatLlamaPack
📦 What is this software?
Llamaindex by Llamaindex
⚠️ Risk & Real-World Impact
Worst Case
Full database compromise leading to remote code execution, data exfiltration, and complete system takeover.
Likely Case
Database manipulation, data theft, and potential privilege escalation within the database environment.
If Mitigated
Limited to database-level impact if proper input validation and parameterized queries are implemented.
🎯 Exploit Status
SQL injection to RCE chain is documented in public bounty reports.
🛠️ Fix & Mitigation
✅ Official Fix
Patch Version: 0.3.0
Vendor Advisory: https://github.com/run-llama/llama_index/commit/5d03c175476452db9b8abcdb7d5767dd7b310a75
Restart Required: Yes
Instructions:
1. Update llama_index to version 0.3.0 or later. 2. Restart the application. 3. Verify the fix by checking the version and testing SQL query functionality.
🔧 Temporary Workarounds
Input Validation and Sanitization
allImplement strict input validation and parameterized queries for all database operations.
Database Privilege Reduction
allRestrict database user permissions to prevent large object functionality execution.
REVOKE ALL ON LARGE OBJECT FROM [username];
🧯 If You Can't Patch
- Implement network segmentation to isolate the vulnerable component from critical systems.
- Deploy a web application firewall with SQL injection detection rules.
🔍 How to Verify
Check if Vulnerable:
Check if using llama_index version <= v0.12.3 and FinanceChatLlamaPack is enabled.Check Version:
pip show llama_index | grep VersionVerify Fix Applied:
Confirm version is >= 0.3.0 and test SQL query functionality with malicious inputs.📡 Detection & Monitoring
Log Indicators:
- Unusual SQL query patterns
- Large object function calls in database logs
- Multiple failed query attempts
Network Indicators:
- Unusual database connection patterns
- SQL injection payloads in application traffic
SIEM Query:
source="application.logs" AND ("run_sql_query" OR "FinanceChatLlamaPack") AND ("SELECT lo_" OR "CREATE FUNCTION")