The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the official Alibaba Cloud CLI (aliyun) via the subprocess.run function in _cli.py. The implementation uses list-based arguments rather than shell strings (shell=False), which is a secure practice that prevents command injection. This execution is central to the skill's purpose of gathering diagnostic information.\n- [CREDENTIALS_UNSAFE]: The module _cli.py accesses the local configuration file ~/.aliyun/config.json to automatically resolve the region_id for API calls. While this file typically contains sensitive credentials, the skill only parses it for region settings and does not expose or exfiltrate any secrets. As this is the official configuration path for the tool being wrapped, and the skill is authored by the vendor (aliyun), this access is considered legitimate and safe functionality.\n- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data in the form of cloud metadata (e.g., resource names, descriptions) returned by the Alibaba Cloud APIs. This creates a potential surface for indirect prompt injection if the metadata contains malicious instructions. However, the risk is mitigated by the skill's instructions which mandate structured Markdown table output and restrict the agent to a read-only diagnostic role, preventing the execution of embedded instructions.

alibabacloud-lb-healthcheck