Paper/Archive/CVE-2025-46059

CVE-2025-46059

mitre — langchain-ai v0.3.51 was discovered to contain an indirect prompt injection vulnerability in the GmailToolkit component.

EPSS 0.8%public exploit CVSS 9.8

critical.is priorityof 100

CVSS 9.8 (ADP)9.8 / 10 × 3029.4 Not in CISA KEV+300.0 EPSS percentile 57.8.578 × 2011.6 Exploit: SSVC poc+1010.0 SSVC exploitation: poc+100.0

Description · from the CVE record

langchain-ai v0.3.51 was discovered to contain an indirect prompt injection vulnerability in the GmailToolkit component. This vulnerability allows attackers to execute arbitrary code and compromise the application via a crafted email message. NOTE: this is disputed by the Supplier because the code-execution issue was introduced by user-written code that does not adhere to the LangChain security practices.

Affected

ProductAffectedFixed inSource

n/an/a–CNA

Scores by source

CISA ADP9.8CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H

CISA ADPSSVCExploitation: poc · Automatable: yes · Technical impact: total

References

refgithub.com/langchain-ai/langchain/issues/30833 refgithub.com/Jr61-star/CVEs/blob/main/CVE-2025-46059.md refgithub.com/langchain-ai/langchain-community/issues/217#issuecomment-3144824471 refpython.langchain.com/docs/security