Arbitrary File Read via ImagePromptTemplate
A vulnerability in langchain-core versions >=0.1.17,<0.1.53 || >=0.2.0,<0.2.43 || >=0.3.0,<0.3.15 allows unauthorized file reading from the host system. This issue was patched in version 0.3.15.
Available publicly on Feb 09 2025 | Available with Premium on Nov 06 2024
Remediation Steps
- Update to langchain-core version 0.3.15 or later.
- Ensure that user inputs to prompt templates are properly sanitized and validated.
- Avoid exposing the outputs of prompt templates directly to users or through downstream models without proper checks.
Patch Details
- Fixed Version: 0.3.15
- Patch Commit: https://github.com/langchain-ai/langchain/commit/c1e742347f9701aadba8920e4d1f79a636e50b68
Want more out of Sightline?
Sightline offers even more for premium customers
Go Premium
We have - related security advisories that are available with Sightline Premium.