aliyun-cli-skills

Pass

Audited by Gen Agent Trust Hub on May 15, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads Aliyun CLI installation packages and scripts from the official Alibaba Cloud CDN (aliyuncli.alicdn.com). This is standard behavior for tool setup and utilizes a well-known service domain.
  • [REMOTE_CODE_EXECUTION]: Installation instructions for Linux and macOS involve fetching a shell script via curl and executing it directly with bash. While this pattern is typically high-risk, it is used here to install official tooling from a trusted vendor.
  • [COMMAND_EXECUTION]: The skill is designed to construct and execute aliyun CLI commands on the host system. It correctly identifies the need for elevated privileges (sudo) when moving binaries to system paths like /usr/local/bin/.
  • [PROMPT_INJECTION]: As the skill translates user natural language into executable shell commands, there is an inherent risk of indirect prompt injection. However, the skill mitigates this by instructing the agent to use --dryrun for verification and to seek explicit user confirmation before performing any destructive or modification-based operations.
Audit Metadata
Risk Level
SAFE
Analyzed
May 15, 2026, 08:13 AM
Security Audit — agent-trust-hub — aliyun-cli-skills