Database/AI/ML frameworks & serving
LangChain (`LLMMathChain`): Prompt injection
CVSS 9.8CVE-2023-29374AI/ML frameworks & servingcurated
Impact
Prompt injection → arbitrary Python execution
Who can reach it
Untrusted prompt or retrieved document reaching an agent running on the GPU node
What to do
No patch for the pattern — LLM output feeding exec is the design. Sandbox agent execution; provider isolates the node's credentials
References
Related entries
- LangChain (PALChain): Arbitrary code execution via `os.system`/`exec` in generated codeCVE-2023-36258 · LangChain (PALChain)Critical
- LangChain (`load_prompt`): Arbitrary code execution from a JSON prompt fileCVE-2023-36281 · LangChain (`load_prompt`)Critical
- langchain-experimental (PALChain): Bypass of the CVE-2023-36258 fixCVE-2023-44467 · langchain-experimental (PALChain)Critical
- Ray (job submission API): Unauthenticated RCE — the Jobs API accepts arbitrary code by designCVE-2023-48022 · Ray (job submission API)Critical
- MLflow: Arbitrary account creation bypassing authenticationCVE-2023-6014 · MLflowCritical
- MLflow: Overwrite any file on the MLflow host without authenticationCVE-2023-6018 · MLflowCritical
This entry is curated: imported from vendor advisories with machine assistance, not yet individually verified. Confirm against your vendor's advisory before acting, and report anything wrong.