LangGrinch Vulnerability Analysis: CVE-2025-68664

Overview

This repository contains a technical report analyzing CVE-2025-68664, also known as LangGrinch, a serialization injection vulnerability affecting LangChain Core.

The report explains the vulnerability background, root cause, exploitation flow, threat model, mitigation strategies, and real-world security impact on LLM-based applications.

Topics Covered

• LangChain serialization and deserialization
• Insecure deserialization / CWE-502
• LLM application security
• Prompt injection as a delivery path
• Secret exposure risk
• STRIDE threat modeling
• Mitigation and secure coding practices

Report

The full report is available here:

• LangGrinch CVE-2025-68664 Report

Key Takeaways

• Untrusted user input, LLM output, tool responses, and metadata should not be treated as trusted serialized objects.
• Applications should upgrade vulnerable LangChain Core versions.
• Sensitive features such as environment-based secret resolution should be restricted.
• Developers should validate and sanitize structured data before deserialization.