Host command execution via Python sandbox escape
Published Sep 13, 2026 · Updated Sep 13, 2026
Code execution in CrewAI before fb2323b allows attackers to run host commands via untrusted Python passed to CodeInterpreterTool. SandboxPython blocks selected imports but exposes Python's object graph, letting untrusted code recover unrestricted runtime capabilities such as ctypes.CDLL(None). Exploitation requires CodeInterpreterTool to process attacker-controlled Python in the in-process restricted sandbox when isolated Docker execution is unavailable.
Summary
What happened
Code execution in CrewAI before fb2323b allows attackers to run host commands via untrusted Python passed to CodeInterpreterTool. SandboxPython blocks selected imports but exposes Python's object graph, letting untrusted code recover unrestricted runtime capabilities such as ctypes.CDLL(None). Exploitation requires CodeInterpreterTool to process attacker-controlled Python in the in-process restricted sandbox when isolated Docker execution is unavailable.
The record
- CVE
- CVE-2026-37008
- Published
- Sep 13, 2026
- Updated
- Sep 13, 2026
- Vendor
- CrewAI
- Product
- CrewAI
- Classifications
- T1059.006
- Attack vector
- Unavailable
- Privileges
- Unavailable
Timeline
How it unfolded
- Sep 13, 2026CVE publishedPublication date reported by the CVE source.
- Sep 13, 2026Record updatedLatest update available in the CVE record.
Exploitability
Present is not the same as exploitable
Compare your product and version with the public record. A matching version still requires validation against your environment.
Is a vulnerable build present?
Compare these published version ranges with your installed build and any vendor patches.
- Affected versionversion=0 <fb2323b3deb3ec62b3965526857e77a2264e4cd0
What conditions does exploitation require?
What is affected?
Published CVSS scores
No CVSS assessment is available in this record.
CVSS describes severity. EPSS estimates exploitation probability.
Attacks
What attackers are doing with it
Daily unique IPs observed by Shadowserver honeypots for known exploited vulnerabilities (KEVs). Missing observations do not establish an absence of attacks.
Weakness, pattern, technique
Public exploit references
- Upstream SandboxPython escape demonstrationproof of concept · validated
Labels summarize the accepted research assessment. They do not indicate a test against your environment.
Technologies
Your stack
See the directory against your own environment.
Your stack
Check the software in your environment
Book a demo to see how Hinoki identifies affected software and validates exploitability in your environment.
Book a demo