alicloud-ai-misc-crawl-and-skill

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill utilizes npx to fetch and run the @just-every/crawl package for the purpose of retrieving documentation from official Alibaba Cloud URLs. This is consistent with the skill's primary function of data discovery.
  • [COMMAND_EXECUTION]: Local Python scripts (refresh_models_summary.py and refresh_alicloud_skills.py) are executed to parse crawled content. A review of these scripts shows they only use standard libraries (json, pathlib, re, urllib) for data transformation and file writing within the project scope.
  • [PROMPT_INJECTION]: No attempts to override agent instructions, bypass safety filters, or extract system prompts were detected in the skill's markdown or configuration files.
  • [DATA_EXFILTRATION]: There is no access to sensitive local files (e.g., SSH keys, environment variables) and no network calls to unknown or suspicious external endpoints.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 10:14 AM