Cyber Resilience

CVE-2024-28088

HighPublic PoC

Published: 04 March 2024

Published
04 March 2024
Modified
08 January 2025
KEV Added
Patch
CVSS Score v3.1 8.1 CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:H/I:H/A:N
EPSS Score 0.1343 94.4th percentile
Risk Priority 24 60% EPSS · 20% KEV · 20% CVSS

Summary

CVE-2024-28088 is a high-severity Path Traversal (CWE-22) vulnerability in Langchain Langchain. Its CVSS base score is 8.1 (High).

Operationally, exploitation aligns with the MITRE ATT&CK technique File and Directory Discovery (T1083); ranked in the top 5.6% of CVEs by exploit likelihood; it is not currently listed in the CISA KEV catalog; a public proof-of-concept is referenced.

This vulnerability is AI-related — categorised as AI Agent Protocols and Integrations; in the Supply Chain and Deployment risk domain; MITRE ATLAS techniques in scope: AI Supply Chain Compromise (AML.T0010), AI Model Inference API Access (AML.T0040), LLM Prompt Injection (AML.T0051).

Deeper analysis

LangChain through version 0.1.10 contains a directory traversal flaw tracked as CVE-2024-28088 that affects the load_chain function in the langchain-core component. An actor who can supply the final portion of the path parameter is able to inject ../ sequences, which bypasses the intended restriction that configurations may be loaded only from the hwchase17/langchain-hub GitHub repository. The issue is assigned CWE-22 and carries a CVSS 3.1 score of 8.1.

An authenticated caller who controls the path argument can therefore retrieve arbitrary files from the local filesystem or cause the application to load attacker-supplied configuration, resulting in exposure of large-language-model API keys or remote code execution. The vulnerability is exploitable over the network without user interaction once the attacker can invoke load_chain with a crafted path.

A fix is present in langchain-core 0.1.29, delivered via the changes merged in pull request 18600 and reflected in the updated loading.py implementation. Public proof-of-concept code demonstrating both information disclosure and code execution has been published. The EPSS score remains at 0.1343 with no material increase from its recorded peak.

EU & UK References

Vulnerability details

LangChain through 0.1.10 allows ../ directory traversal by an actor who is able to control the final part of the path parameter in a load_chain call. This bypasses the intended behavior of loading configurations only from the hwchase17/langchain-hub GitHub repository.…

more

The outcome can be disclosure of an API key for a large language model online service, or remote code execution. (A patch is available as of release 0.1.29 of langchain-core.)

CWE(s)

AI Security AnalysisAI

AI Category
AI Agent Protocols and Integrations
Risk Domain
Supply Chain and Deployment
OWASP Top 10 for LLMs 2025
None mapped
Classification Reason
LangChain is a framework for building applications powered by LLMs, including chains, agents, and integrations with models, tools, and hubs, making it a fit for AI Agent Protocols and Integrations.

Related Threats

MITRE ATT&CK Enterprise TechniquesAI

T1083 File and Directory Discovery Discovery
Adversaries may enumerate files and directories or may search in specific locations of a host or network share for certain information within a file system.
T1210 Exploitation of Remote Services Lateral Movement
Adversaries may exploit remote services to gain unauthorized access to internal systems once inside of a network.
T1552.001 Credentials In Files Credential Access
Adversaries may search local file systems and remote file shares for files containing insecurely stored credentials.
Why these techniques?

Directory traversal in load_chain enables arbitrary file reads (T1083, T1552.001 for API key disclosure) and remote code execution (T1210) by loading malicious local chain configurations.

MITRE ATLAS TechniquesAI

MITRE ATLAS techniques

AML.T0010: AI Supply Chain CompromiseAML.T0040: AI Model Inference API AccessAML.T0051: LLM Prompt Injection

Affected Assets

langchain
langchain
≤ 0.1.12

Mitigating Controls

Likely Mitigating Controls AI

Per-CVE control mapping for this CVE has not run yet; the list below is derived from the weakness types (CWEs) cited in the NVD entry.

addresses: CWE-22

Validates pathnames and filenames to prevent traversal outside intended directories.

References