alibabacloud-emr-cluster-manage
Pass
Audited by Gen Agent Trust Hub on Mar 31, 2026
Risk Level: SAFE
Full Analysis
- [COMMAND_EXECUTION]: The skill manages Alibaba Cloud EMR clusters via the
aliyunCLI. It includes a dedicated 'Input Validation' section that instructs the agent to validate user-supplied strings (such as ClusterName and Description) against specific allow-lists and to reject input containing shell meta-characters (e.g., backticks, pipe, semicolons) to prevent command injection. - [CREDENTIALS_UNSAFE]: No hardcoded secrets or credentials were identified. The documentation correctly uses placeholders (e.g., 'YourRdsPassword123') and recommends standard, secure authentication practices such as using
aliyun configureand STS tokens. - [DATA_EXFILTRATION]: All network operations and API calls are directed to official Alibaba Cloud endpoints (aliyun.com and aliyuncs.com). The skill's 'Runtime Security' policy explicitly prohibits downloading external scripts or executing code from remote URLs.
- [PROMPT_INJECTION]: The instructions use natural language to define operational constraints and safety checkpoints. There are no attempts to bypass model safety filters, extract system prompts, or override core agent behavior.
- [DATA_EXPOSURE]: The skill provides documentation on creating least-privilege RAM (Resource Access Management) policies, encouraging users to limit the agent's permissions to only the necessary API actions.
Audit Metadata