Signal is active with 4 mentions in latest observed window
Recommended action window: Monitor and triage in normal cycle
NVD description
langgraph-api implements the LangGraph API for rapid development and testing. From 0.1.2 until 0.15.0, JavaScript or TypeScript custom authentication Store authorization handlers can return a rewritten namespace that is not carried across the runtime boundary and applied before the Store operation. An authenticated user can therefore access another user's Store namespace when a deployment relies on namespace rewriting for isolation, allowing unauthorized reads, overwrites, searches, or deletions. Python custom authentication and handlers that reject unauthorized namespaces instead of rewriting them are not affected. This issue is fixed in version 0.15.0.
Priority
LOW
Exploitation
NONE
PoC
NONE
Patch
NONE
Momentum
STABLE
What stops one AI agent from accessing another agent's data? Coinspect researchers discovered CVE-2026-118668, a LangGraph authorization flaw that could allow one agent to read and modify another agent’s stored data. Our first case study in agentic-systems security.
Browser use is growing fast. Related note on agent isolation, we recently found a high-severity issue in LangGraph’s Store auth path (CVE-2026-118668) where JS custom handlers could fail to properly scope namespaces between agents. Fixed in langgraph-api >= 0.15.0. Write-up: https://www.coinspect.com/blog/langgraph-agent-security/
Nice production example. On the security side, just flagging a recent high-severity issue in LangGraph’s Store auth path (CVE-2026-118668) that could let one authenticated agent access another’s data under certain JS custom-auth setups. Fixed in langgraph-api >= 0.15.0. Write-up: https://www.coinspect.com/blog/langgraph-agent-security/
On the multi-agent isolation side, just a heads-up that there was a recent high-severity issue in LangGraph’s Store auth path (CVE-2026-118668) where JS/TS custom handlers could fail to properly apply namespace rewriting. Fixed in langgraph-api >= 0.15.0. Write-up: https://www.coinspect.com/blog/langgraph-agent-security/
Interesting point on tighter harness + infra integration. On the isolation side, we recently found a high-severity issue in LangGraph’s Store auth path (CVE-2026-118668) where JS/TS custom handlers could fail to properly apply namespace rewriting, letting one authenticated agent read or modify another’s data. Fixed in langgraph-api >= 0.15.0. Write-up: https://www.coinspect.com/blog/langgraph-agent-security/