skills/aliyun/alibabacloud-aiops-skills/alibabacloud-elasticsearch-instance-manage/Gen Agent Trust Hub
alibabacloud-elasticsearch-instance-manage
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill references downloads for the Aliyun CLI setup script and binaries from
https://aliyuncli.alicdn.com/. As these resources belong to the official vendor domain for this skill, they are considered legitimate installation components. - [REMOTE_CODE_EXECUTION]: Instructions in
SKILL.mdandreferences/cli-installation-guide.mdsuggest installing the Aliyun CLI usingcurl -fsSL https://aliyuncli.alicdn.com/setup.sh | bash. This pattern involves executing remote code, but it is targeted at a trusted vendor-owned CDN. - [COMMAND_EXECUTION]: The skill uses local system utilities like
openssl,python3,uuidgen, andheadto generate session identifiers and idempotency tokens. These are standard operations for ensuring session observability and API call uniqueness. - [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided inputs (Region, InstanceId, Password) for interpolation into Alibaba Cloud CLI commands. While it implements clear boundary markers and validation steps, this data ingestion represents a potential surface for indirect prompt injection.
- Ingestion points: User-provided parameters for ES instance creation and configuration in
SKILL.mdandreferences/instance-manage.md. - Boundary markers: The skill uses explicit checklists and status validation calls (
describe-instance) to ensure instance readiness before executing write operations. - Capability inventory: Execution of
aliyunCLI commands for instance lifecycle, plugin, and configuration management. - Sanitization: The skill provides strong warnings and rejection logic for users attempting to provide AccessKeys directly in the conversation.
Audit Metadata