vmware-vcenter-attack
Warn
Audited by Gen Agent Trust Hub on Oct 28, 2026
Risk Level: MEDIUMCOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill contains a set of pre-defined shell commands using
curl,openssl, andldapsearchto fingerprint VMware services and perform enumeration of SSO and LDAP boundaries. - [DATA_EXFILTRATION]: Step 8 includes instructions and command patterns for listing and downloading virtual machine disks (VMDKs) and snapshots from datastores, which represent a significant data exfiltration capability.
- [INDIRECT_PROMPT_INJECTION]: The skill implements an attack surface for indirect injection by interpolating the user-provided
TARGETvariable into shell commands. - Ingestion points: The
TARGETvariable is defined withinSKILL.mdand relies on external input provided by the user. - Boundary markers: The skill utilizes a 'Mandatory confirmation gate' requiring target verification and explicit authorization before running any active probes.
- Capability inventory: The skill uses
curl,openssl, andldapsearchfor network interactions and references multiple exploitation tools such as Metasploit and Nuclei. - Sanitization: There is no evidence of sanitization or character escaping for the
TARGETvariable before it is interpolated into executable shell commands.
Audit Metadata