- CVSS
- MEDIUM · 4.7v3.1CVSS:3.1/AV:L/AC:H/PR:N/UI:R/S:U/C:N/I:H/A:N
- Published
- 2026-10-10
- Weakness
- CWE-59
- Source
- nvd.nist.gov/vuln/detail/CVE-2026-108600
Description
open-multi-agent (@open-multi-agent/core) 1.5.0 through 1.21.2 contains a link following vulnerability in the file_write tool sandbox that allows attackers to create files outside the workspace root by using dangling symlinks. Attackers can plant a dangling symlink in the workspace and steer the agent via prompt injection to write attacker-influenced content anywhere the agent process can write.
References
- https://github.com/open-multi-agent/open-multi-agent
- https://github.com/open-multi-agent/open-multi-agent/blob/6ec5498643c7e431aa66751f17f47a529e7f927c/packages/core/src/tool/built-in/file-write.ts#L67
- https://github.com/open-multi-agent/open-multi-agent/blob/6ec5498643c7e431aa66751f17f47a529e7f927c/packages/core/src/tool/built-in/path-safety.ts#L104-L113
- https://hackmd.io/@haind/open-multi-agent-file-write-dangling-symlink-escape
- https://www.vulncheck.com/advisories/open-multi-agent-1.5.0-through-1.21.2-sandbox-escape-via-file-write-dangling-symlink
- https://hackmd.io/@haind/open-multi-agent-file-write-dangling-symlink-escape
How GTK Cyber trains on this
AI security training at GTK Cyber covers the LLM and ML-pipeline vulnerability classes that vulnerabilities like CVE-2026-108600 fall into. Our hands-on courses are taught by Charles Givre and other practitioners who break and defend production AI systems.