aliyun-kms-manage
Pass
Audited by Gen Agent Trust Hub on Oct 15, 2026
Risk Level: SAFEDATA_EXFILTRATIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [DATA_EXFILTRATION]: The skill is designed to interact with Alibaba Cloud services by accessing credentials from environment variables (ALIBABACLOUD_ACCESS_KEY_ID, ALIBABACLOUD_ACCESS_KEY_SECRET) and the standard shared credentials file located at
~/.alibabacloud/credentialsas part of its resource management workflow.\n- [COMMAND_EXECUTION]: The skill executes a local Python script,scripts/list_openapi_meta_apis.py, to dynamically discover and document available KMS APIs.\n- [EXTERNAL_DOWNLOADS]: The discovery script fetches OpenAPI specification data fromhttps://api.aliyun.com, which is the official domain for Alibaba Cloud services.\n- [INDIRECT_PROMPT_INJECTION]: The skill processes external metadata from Alibaba Cloud endpoints, which creates an attack surface for indirect prompt injection. Mandatory evidence chain:\n - Ingestion points: Data is retrieved from
https://api.aliyun.com/meta/v1/products/Kms/versions/2016-01-20/api-docs.jsonvia thelist_openapi_meta_apis.pyscript.\n - Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore potentially malicious content within the fetched JSON data.\n
- Capability inventory: The skill can execute local scripts, perform file writes to the output directory, and utilize cloud management SDKs.\n
- Sanitization: The script parses JSON and generates Markdown/JSON artifacts based on the content without performing validation or escaping on the retrieved API names.
Audit Metadata