huawei-cloud-icp-rule-consult
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains specific instructions to guide the agent's behavior towards accuracy and adherence to official rules. There are no attempts to bypass safety filters, extract system prompts, or override core behavioral guidelines.
- [DATA_EXFILTRATION]: No sensitive file access or credential harvesting patterns were found. Network activity is limited to fetching public documentation and performing searches on the vendor's official domain (support.huaweicloud.com).
- [REMOTE_CODE_EXECUTION]: The skill invokes local Python scripts (
web_fetcher.pyandweb_searcher.py) to retrieve information. There is no evidence of downloading and executing untrusted remote scripts or piping network content to a shell. - [COMMAND_EXECUTION]: The skill uses predefined shell commands to run its own scripts. These commands are strictly for documentation retrieval and do not accept arbitrary user input in a way that would allow for command injection.
- [EXTERNAL_DOWNLOADS]: Dependencies listed in
requirements.txt(requests, beautifulsoup4, playwright, aiohttp) are standard, well-known libraries used for the skill's intended web-scraping and search functionality.
Audit Metadata