GHSA-2pq5-3q89-j7cc: Langroid: Neo4jChatAgent executes LLM-generated Cypher without validation (prompt-to-Cypher injection; config-conditional RCE), mirroring the SQLChatAgent bug fixed in CVE-2026-25879

Summary

Langroid's Neo4jChatAgent sends LLM-generated Cypher queries directly to the Neo4j driver through its read_query and write_query paths, with no validation, statement-type allowlist, or opt-out gate. Because prompt injection can influence the query text, an attacker can read or destroy graph data and, when APOC or dbms.security procedures are enabled, reach OS-command and filesystem access. The report says this is the same defect class as the SQLChatAgent issue fixed in version 0.63.0 (CVE-2026-25879), and that the fix did not extend to the neo4j module.

Mitigation

The source does not state a fix yet. Check the original advisory for updates.

Affected packages in the Exposure Registry

Matched by package name and ecosystem. Each entry shows whether the package delegates to a language model and how many tracked packages depend on it.

• langroidPyPILLM dependency since 2023-07-14

Related items

• HighCVE-2026-76066: IntentKit authorization bypass lets callers execute owner-restricted toolsSimilar attack · NVD/CVE Database
• MediumAnthropic 'SneakPeek' Flaw Is a Silent Zero-Click ThreatSimilar attack · Dark Reading
• HighCVE-2026-104103: Ludwig archive extraction allows writes outside extraction directorySimilar attack · NVD/CVE Database
• LowAgents at the Edge: Extending AI Security Governance to Physical AI with DefenseClawSimilar attack · Cisco AI Blog
• InfoFrom Training to Escalation: Forensic Analysis and Local Recreation of RL/RLHF-Induced Privilege Escalation in AI Agents (CVE-2026-65616)Similar attack · Arxiv (cs.CR + cs.CL + cs.LG)