Am I affected by Langroid: SQLChatAgent dangerous-function blocklist can be bypassed with quoted or…? What should I do?
The SQLChatAgent SQL-injection mitigation, with default allow_dangerous_operations=False, combines a raw-text regex blocklist (_DANGEROUS_SQL_PATTERNS) with a sqlglot SELECT-only statement allowlist.
Affected: pip/langroid <= 0.65.0. Fixed in: 0.65.1. What should you check, and what should you do about it?
Full incident page with sources: https://agentarmorhq.com/incidents/2026/07/cve-2026-54760-langroid (opens external site)
1 answer
Am I affected? Affected: pip/langroid <= 0.65.0. Fixed in: 0.65.1.
- Check your version: run
pip show langroid. Versions <= 0.65.0 of langroid are affected.
What to do now
- Upgrade to the fixed version: run
pip install --upgrade "langroid>=0.65.1"(first patched version 0.65.1). - Run SQLChatAgent with a least-privilege database role that lacks pg_read_server_files.
General practice, not from the source: if the affected software could read tokens or credentials, rotate them after upgrading; pin versions and review changes before updating tools your agents run.
Sources
- GitHub Advisory Database: https://github.com/advisories/GHSA-6xc5-4r68-67fc (opens external site)
Full incident page: https://agentarmorhq.com/incidents/2026/07/cve-2026-54760-langroid (opens external site)
Did this work for you?- Check your version: run