RCE in Agpt Autogpt Classic
CVE-2024-6091
A vulnerability in significant-gravitas/autogpt version 0.5.1 allows an attacker to bypass the shell commands denylist settings. The issue arises when the denylist is configured to block specific commands, such as 'whoami' and '/bin/whoami'. An attacker can circumvent this restriction by executing commands with a modified path, such as '/bin/./whoami', which is not recognized by the denylist.
Vulnerability class: Command Injection (OS Command Injection)
Published · last modified .
CVSS v3 metric
CVSS v3 base score 9.8 (Critical). Vector: CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:H/I:H/A:H.
EPSS exploit prediction
EPSS 0.008, 57.7th percentile (scored ); unchanged over 30 days. About EPSS.
Affected products
- Agpt Autogpt Classic — versions 0.5.1
- Significant-Gravitas Significant-Gravitas/autogpt — versions unspecified
Weakness classification (CWE)
Public exploits and proof-of-concept code
Also mentioned in 3 other repositories (rahg0/python-intentionally-vuln-package, rahg0/python-weather-app, rahg0/python-weather-app-with-pipenv). These reference the CVE in a list, notes or a scanner and are not exploit code for it.
References
- huntr.com/bounties/8a742c13-bb5e-4bc9-8b86-049d8a386050 (Exploit, Third Party Advisory)
- github.com/significant-gravitas/autogpt/commit/ef691359b774a1f9f80cf4f5ace98219… (Patch)