doca-rdma
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFECOMMAND_EXECUTIONPRIVILEGE_ESCALATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the user to execute various shell commands for software development and hardware diagnostics. These include build system invocations (
meson,ninja) and network utility commands (pkg-config,ibdev2netdev,ethtool). These operations are standard for the development and testing of RDMA applications. - [PRIVILEGE_ESCALATION]: Several diagnostic commands in the
Command appendixofTASKS.mdare marked as requiringsudo(e.g.,dmesg,mlxconfig,ibv_devinfo). These commands are used to access kernel logs and hardware-level configuration, which is typical for debugging InfiniBand and RoCE network devices. - [INDIRECT_PROMPT_INJECTION]: The skill demonstrates a vulnerability surface by ingesting data from external system tools like
dmesganddoca_caps. This data is then used in the agent's context for debugging purposes. However, this ingestion is necessary for the skill's primary function, and there are no signs of malicious patterns designed to exploit this surface.
Audit Metadata