Command Injection in run-llama/llama_index
run-llama/llama_index
Affected versions
| Product | Affected | Fixed in |
|---|---|---|
| llama-index PyPI |
|
|
| llama-index-llms-rungpt PyPI |
|
|
run-llama/llama_index
| Product | Affected | Fixed in |
|---|---|---|
| llama-index PyPI |
|
|
| llama-index-llms-rungpt PyPI |
|
|
A command injection vulnerability exists in the RunGptLLM class of the llama_index library, version 0.9.47, used by the RunGpt framework from JinaAI to connect to Language Learning Models (LLMs). The vulnerability arises from the improper use of the eval function, allowing a malicious or compromised LLM hosting provider to execute arbitrary commands on the client's machine. This issue was fixed in version 0.10.13. The exploitation of this vulnerability could lead to a hosting provider gaining full control over client machines.
From the CVE record · @huntr_ai
RunGptLLM in llama_index improperly uses eval when connecting to an LLM hosted by a provider. A malicious or compromised LLM hosting provider can execute arbitrary commands on a client's machine when the client uses this class.
Check deployments using RunGptLLM with PyPI llama-index versions before 0.10.13 or llama-index-llms-rungpt versions before 0.1.3.