SOC status:Duty analyst on shift

UK Cyber Defence
VulnerabilityDeferred

CVE-2024-34359

This allows `jinja2` Server Side Template Injection which leads to remote code execution by a carefully constructed payload.

CRITICAL 9.6EPSS 28.4%

Does this matter?

EPSS puts the probability of exploitation in the next 30 days at 28.4%, higher than 98% of all known CVEs. Patch or mitigate before the next change window.

Description

llama-cpp-python is the Python bindings for llama.cpp. `llama-cpp-python` depends on class `Llama` in `llama.py` to load `.gguf` llama.cpp or Latency Machine Learning Models. The `__init__` constructor built in the `Llama` takes several parameters to configure the loading and running of the model. Other than `NUMA, LoRa settings`, `loading tokenizers,` and `hardware settings`, `__init__` also loads the `chat template` from targeted `.gguf` 's Metadata and furtherly parses it to `llama_chat_format.Jinja2ChatFormatter.to_chat_handler()` to construct the `self.chat_handler` for this model. Nevertheless, `Jinja2ChatFormatter` parse the `chat template` within the Metadate with sandbox-less `jinja2.Environment`, which is furthermore rendered in `__call__` to construct the `prompt` of interaction. This allows `jinja2` Server Side Template Injection which leads to remote code execution by a carefully constructed payload.

CVSS 3.1
9.6 CRITICALCVSS:3.1/AV:N/AC:L/PR:N/UI:R/S:C/C:H/I:H/A:H
EPSS
28.42% probability · 98th percentile
CISA KEV
Not listed
Weakness
CWE-76
Source
security-advisories@github.com

Source: NVD record, EPSS from FIRST.org, KEV from CISA. Refreshed daily.