alibabacloud-mse-nacos-inspection

Pass

Audited by Gen Agent Trust Hub on Jul 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill guides the user to download the Alibaba Cloud CLI binary from the official vendor distribution domain (aliyuncli.alicdn.com). These references are consistent with the skill's stated purpose and originate from the author's infrastructure.
  • [COMMAND_EXECUTION]: The skill utilizes the aliyun CLI and curl to perform read-only queries against cloud services. These operations are restricted to obtaining instance metadata and monitoring metrics.
  • [INDIRECT_PROMPT_INJECTION]: The skill has a surface for indirect prompt injection as it processes data from external cloud APIs.
  • Ingestion points: Monitoring metrics from Prometheus HTTP APIs and instance metadata from MSE cluster listing commands.
  • Boundary markers: Not explicitly defined in the prompt templates.
  • Capability inventory: Executes cloud API commands and HTTP requests via shell.
  • Sanitization: Uses Python's urllib.parse module to safely encode PromQL query parameters before interpolation, mitigating command and query injection risks.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 13, 2026, 05:40 AM
Security Audit — agent-trust-hub — alibabacloud-mse-nacos-inspection