As cited
Copy frozen at (site build).
vulnerabilities
From SQLi to RCE - Exploiting LangGraph’s Checkpointer
Check Point Research identified three vulnerabilities in LangGraph, an open-source AI agent framework with over 50 million monthly downloads, affecting its SQLite and Redis checkpointers. Two vulnerabilities chain together to enable remote code execution through SQL injection and unsafe msgpack deserialization, while a third introduces SQL injection to the Redis checkpointer. LangChain has released patches for all three issues.
Why it matters: Teams self-hosting LangGraph with SQLite or Redis checkpointers that expose get_state_history() with user-controlled filters face remote code execution risk and must update to patched versions immediately; LangSmith Deployment (managed cloud) users are not affected.
- Source published
- First seen by Cybersecurity Tracker
Source attribution
Correction
Correction recorded as of .
vulnerabilities
From SQLi to RCE - Exploiting LangGraph’s Checkpointer
No summary had been written when this copy was frozen.
- First seen by Cybersecurity Tracker
Source attribution
Correction
Correction recorded as of .
vulnerabilities
From SQLi to RCE - Exploiting LangGraph’s Checkpointer
Check Point Research identified three vulnerabilities in LangGraph, an open-source framework for building stateful artificial intelligence (AI) agents: SQL injection in the SQLite checkpointer (CVE-2025-67644), unsafe msgpack deserialization (CVE-2026-28277), and a parallel injection flaw in the Redis checkpointer (CVE-2026-27022). The first two vulnerabilities chain together to enable remote code execution (RCE). LangChain has released patches for all three issues across multiple package versions.
Why it matters: Teams running LangGraph on-premises with SQLite or Redis checkpointers that expose the get_state_history() function to user input face RCE risk and must update to the patched versions immediately; LangChain's managed LangSmith Deployment service is unaffected.
- Source published
- First seen by Cybersecurity Tracker
Source attribution
Correction
Correction recorded as of .
vulnerabilities
From SQLi to RCE - Exploiting LangGraph’s Checkpointer
Check Point Research identified three vulnerabilities in LangGraph, an open-source framework for building stateful artificial intelligence (AI) agents: SQL injection in the SQLite checkpointer (CVE-2025-67644), unsafe msgpack deserialization (CVE-2026-28277), and a parallel injection flaw in the Redis checkpointer (CVE-2026-27022). The first two vulnerabilities chain together to enable remote code execution (RCE). LangChain has released patches for all three issues across multiple package versions.
Why it matters: Teams running LangGraph on-premises with SQLite or Redis checkpointers that expose the get_state_history() function to user input face RCE risk and must update to the patched versions immediately; LangChain's managed LangSmith Deployment service is unaffected.
- Source published
- First seen by Cybersecurity Tracker
Source attribution
Correction
Correction recorded as of .
vulnerabilities
From SQLi to RCE - Exploiting LangGraph’s Checkpointer
Check Point Research identified three vulnerabilities in LangGraph, an open-source framework for building stateful artificial intelligence (AI) agents: SQL injection in the SQLite checkpointer (CVE-2025-67644), unsafe msgpack deserialization (CVE-2026-28277), and a parallel injection flaw in the Redis checkpointer (CVE-2026-27022). The first two vulnerabilities chain together to enable remote code execution (RCE). LangChain has released patches for all three issues across multiple package versions.
Why it matters: Teams running LangGraph on-premises with SQLite or Redis checkpointers that expose the get_state_history() function to user input face RCE risk and must update to the patched versions immediately; LangChain's managed LangSmith Deployment service is unaffected.
- Source published
- First seen by Cybersecurity Tracker