cve.bz

CVE-2026-45134 langchain-ai langsmith-sdk: Deserialization of Untrusted Data

7.1 highEPSS 0.2%Fix available

High (7.1) in langchain-ai langsmith-sdk. EPSS 0.2% estimated probability of exploitation activity in the next 30 days. Fixed in langsmith 0.6.0, 0.8.0.

On this page
LangSmith Client SDKs provide SDK's for interacting with the LangSmith platform. Prior to LangSmith SDK Python 0.8.0 and JS/TS 0.6.0, the LangSmith SDK's prompt pull methods (pull_prompt / pull_prompt_commit in Python, pullPrompt / pullPromptCommit in JS/TS) fetch and deserialize prompt manifests from the LangSmith Hub. These manifests may contain serialized LangChain objects and model configuration that affect runtime behavior. When pulling a public prompt by owner/name identifier, the manifest content is controlled by an external party, but prior versions of the SDK did not distinguish this from pulling a prompt within the caller's own organization. This vulnerability is fixed in LangSmith SDK Python 0.8.0 and JS/TS 0.6.0.

Severity

CVSS 3.1

7.1 high
CVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:U/C:H/I:L/A:N
EPSS
0.2% (unchanged from 0.2% a week ago; 0.3% 30 days ago) estimated probability of exploitation activity in the next 30 days; 10th percentile. Scored 2026-11-22. A forecast for this CVE across the internet, not for your own systems.
CISA's assessment (SSVC)
Exploitation: none · Automatable: no · Technical impact: partial, assessed 2026-06-02. From CISA's Vulnrichment in the CVE List: whether it is exploited, whether exploiting it can be automated, and how much control it gives.
CWE
CWE-502 Deserialization of Untrusted Data
OWASP Top 10
A08:2025 Software or Data Integrity Failures
Published
Modified
NVD analysis
Awaiting analysis
Record from
the NVD, synced

Fix and patched versions

  • langsmith (PyPI): fixed in 0.8.0
  • langsmith (npm): fixed in 0.6.0
  • langchain-classic (PyPI): fixed in 1.0.7
  • langchain (PyPI): fixed in 0.3.30
  • openSUSE: fixed package versions per release

Ask an AI about this CVE

Show the promptExplain CVE-2026-45134 for an engineer deciding whether to act. Use https://cve.bz/api/public/cves/CVE-2026-45134?via=ask-ai-copy and https://cve.bz/cve/CVE-2026-45134?via=ask-ai-copy as the source of truth: what is affected and which versions fix it, whether it is in CISA KEV, public exploits, and its EPSS (an internet-wide 30-day estimate of exploitation activity, not a forecast for my systems). If they give no fixed versions, use the CVE record https://cveawg.mitre.org/api/cve/CVE-2026-45134 or the vendor advisory, not third-party mirrors. Tell me how to check whether I am affected and how to verify a fix. Cite the sources and say where they are uncertain.

Opens a third-party assistant; its answer is its own, not cve.bz's. The prompt holds only this CVE's id and public links.

Affected

Versions

From the CNA's record in the CVE List.

ProductAffected
langchain-ai langsmith-sdk< 0.8.0
langchain-ai langsmith-sdk< 0.6.0

Packages (OSV)

AdvisoryEcosystemPackageAffectedFixed in
GHSA-3644-q5cj-c5c7PyPIlangsmith<0.8.00.8.0
GHSA-3644-q5cj-c5c7npmlangsmith<0.6.00.6.0
GHSA-3644-q5cj-c5c7PyPIlangchain-classic<1.0.71.0.7
GHSA-3644-q5cj-c5c7PyPIlangchain<0.3.300.3.30
PYSEC-2026-2555PyPIlangchain<0.3.300.3.30
PYSEC-2026-2560PyPIlangchain-classic<1.0.71.0.7
PYSEC-2026-2582PyPIlangsmith<0.8.00.8.0

Linux distributions

DistributionReleasePackageFixed inAdvisory
openSUSETumbleweedpython-langsmith0.10.3-1.1openSUSE-SU-2026:11284-1

From the distributions' data via OSV.dev: SUSE security data, CC BY 4.0. Modified: reformatted and linked to CVE records. Credits. On the distributions' own pages: openSUSE.

Red Hat

Red Hat rates this Moderate.

Red Hat's statement

This Moderate impact vulnerability in LangSmith Client SDKs can lead to information disclosure and potential integrity loss. Red Hat products using these SDKs are affected when pulling public prompts from the LangSmith Hub, as untrusted manifest content may be deserialized. This could result in the execution of malicious LangChain objects or model configurations, requiring user interaction for exploitation.

Red Hat's mitigation

To mitigate this vulnerability, avoid pulling public prompts from the LangSmith Hub using the LangSmith Client SDKs without explicit trust. The SDK provides a `dangerously_pull_public_prompt` flag (Python) or `dangerouslyPullPublicPrompt` (JS/TS) that must be set to `True` to enable pulling public prompts by `owner/name`. Do not enable this flag unless the prompt's contents have been independently reviewed and trusted. Additionally, when pulling prompts, avoid setting `include_model=True` or `secrets_from_env=True` if the prompt source is untrusted, as these options expand the deserialization allowlist and allow reading environment variables.

ProductPackageStateAdvisory
OpenShift Lightspeedopenshift-lightspeed/lightspeed-service-api-rhel9Affected
Red Hat Ansible Automation Platform 2ansible-automation-platform-24/lightspeed-rhel8Will not fix
Red Hat Ansible Automation Platform 2ansible-automation-platform-25/lightspeed-chatbot-rhel8Will not fix
Red Hat Ansible Automation Platform 2ansible-automation-platform-25/lightspeed-rhel8Will not fix
Red Hat Ansible Automation Platform 2ansible-automation-platform-26/lightspeed-rhel9Not affected
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/bootc-cuda-rhel9Affected
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/bootc-rocm-rhel9Affected
Red Hat Enterprise Linux AI (RHEL AI) 3rhelai3/disk-image-cuda-rhel9Affected
Red Hat OpenShift AI (RHOAI)rhoai/odh-llama-stack-core-rhel9Not affected
Red Hat OpenShift AI (RHOAI)rhoai/odh-mlflow-rhel9Not affected
Red Hat OpenShift AI (RHOAI)rhoai/odh-trustyai-nemo-guardrails-server-rhel9Affected

Contains Red Hat Security Data © Red Hat, Inc., licensed under CC BY 4.0. Source: https://access.redhat.com/security/data. Modified: reformatted. Red Hat's page for CVE-2026-45134.

Attack patterns (CAPEC)

  • CAPEC-586 Object Injection Likelihood: Medium · Severity: High

The attack patterns MITRE's CAPEC relates to this record's weaknesses, as MITRE names them. Credits.

References

Check your stackJSON