alibabacloud-bailian-managed-agent
Pass
Audited by Gen Agent Trust Hub on Sep 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs users on how to install and update the official Bailian CLI using npm or a shell script hosted on official vendor domains (e.g.,
https://bailian.aliyun.com/cli/install.sh). These downloads originate from the software vendor's own infrastructure. - [REMOTE_CODE_EXECUTION]: Documentation includes standard installation patterns such as piping a remote shell script into bash (
curl -fsSL ... | bash). While this is a high-privilege execution pattern, it is the vendor's official distribution method and targets their own verified domain. - [INDIRECT_PROMPT_INJECTION]: The
scripts/cookbook_smoke.pyscript provides a testing interface that creates cloud resources and sends messages to AI models. This creates an attack surface for indirect prompt injection if the script were used to process untrusted data, but the script is designed for opt-in administrative testing and requires explicit user activation via command-line flags. - [COMMAND_EXECUTION]: The skill provides numerous examples of shell commands and scripts used to manage cloud resources and process data. These commands are intended for use by the operator or the agent within the managed cloud sandbox and are consistent with the skill's purpose as a development and operations tool.
Audit Metadata