alibabacloud-find-skills

Pass

Audited by Gen Agent Trust Hub on Mar 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes legitimate aliyun CLI commands to interact with the Alibaba Cloud AgentExplorer service, including list-categories, search-skills, and get-skill-content. It also utilizes npx to install selected skills.
  • [EXTERNAL_DOWNLOADS]: The skill directs users to download the Alibaba Cloud CLI and its plugins from official, vendor-controlled domains such as aliyuncli.alicdn.com and official GitHub repositories under the aliyun organization. These are trusted sources for the skill's intended functionality.
  • [PROMPT_INJECTION]: The skill has a surface for indirect prompt injection because it retrieves skill descriptions and content from an external repository that may include community-contributed content. However, the skill mitigates this risk by requiring explicit user confirmation for all parameters and search selections before execution.
  • Ingestion points: Data enters the agent context via aliyun agentexplorer search-skills and get-skill-content commands.
  • Boundary markers: While explicit delimiters are not mentioned for search result display, the skill workflow mandates user confirmation for every step.
  • Capability inventory: The skill can execute CLI commands and install new skills using npx.
  • Sanitization: No explicit sanitization of fetched markdown is performed, but the content is presented as human-readable documentation.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 30, 2026, 08:25 AM