alibabacloud-lingjun-node-diagnose

Pass

Audited by Gen Agent Trust Hub on Sep 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads an installation script from the vendor's official content delivery network (https://aliyuncli.alicdn.com/setup.sh). This is standard procedure for setting up the Alibaba Cloud CLI environment required for the skill's operation.
  • [COMMAND_EXECUTION]: The skill executes the downloaded setup script and uses the aliyun command-line interface to perform resource lookups, diagnostic tasks, and node repairs. All operations are within the stated scope of Alibaba Cloud infrastructure management.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data such as system logs (list-syslogs) and user-provided job logs (AiJobLogs) to generate diagnostic reports. This surface is mitigated by strict human-in-the-loop (HITL) requirements: all mutating actions (reboot, reimage, stop) require a two-phase commit process and explicit manual confirmation by the user replying with a specific word ("confirm" or "确认") after reviewing the proposed parameters.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 7, 2026, 02:12 AM
Security Audit — agent-trust-hub — alibabacloud-lingjun-node-diagnose