alibabacloud-dataworks-metadata
Pass
Audited by Gen Agent Trust Hub on May 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands via the Aliyun CLI. It enforces safety through mandatory timeouts and explicit user confirmation for all write operations, reducing the risk of unauthorized or accidental changes.\n- [EXTERNAL_DOWNLOADS]: The skill installs the dataworks-public plugin using the official Aliyun CLI. It provides safe installation paths through official documentation and specifically prohibits insecure practices like piping remote scripts into the shell.\n- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface. It ingests metadata (such as table descriptions and READMEs) from DataWorks into the agent context via commands like get-table and list-tables (found in SKILL.md) and possesses the capability to perform write operations via the CLI. While no specific boundary markers are used for this data, the risk is mitigated by the requirement for explicit user confirmation before any modifications are executed.
Audit Metadata