by Kastro
Paper/Archive/CVE-2024-3271

CVE-2024-3271

run-llama run-llama/llama_index — Command Injection in run-llama/llama_index

EPSS 3%public exploitCVSS 9.8
critical.is priorityof 100
57
CVSS 9.8 (CNA)9.8 / 10 × 3029.4Not in CISA KEV+300.0EPSS percentile 86.9.869 × 2017.4Exploit: SSVC poc+1010.0SSVC exploitation: poc+100.0
Description · from the CVE record

A command injection vulnerability exists in the run-llama/llama_index repository, specifically within the safe_eval function. Attackers can bypass the intended security mechanism, which checks for the presence of underscores in code generated by LLM, to execute arbitrary code. This is achieved by crafting input that does not contain an underscore but still results in the execution of OS commands. The vulnerability allows for remote code execution (RCE) on the server hosting the application.

Affected

ProductAffectedFixed inSource
run-llama/llama_indexunspecified – < 10.2610.26CNA
llama_indexall versions–CISA-ADP

Scores by source

@huntr_ai (CNA)9.8CVSS:3.0/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H
CISA ADPSSVCExploitation: poc · Automatable: yes · Technical impact: total

References

refhuntr.com/bounties/9b32490e-7cf9-470e-8d49-ba083ae7a279refgithub.com/run-llama/llama_index/commit/5fbcb5a8b9f20f81b791c7fc8849e352613ab475