anolisa-guide
Pass
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructions in
SKILL.mdrequire the agent to execute a documentation management script (check_docs.py). This script coordinates the checking and selection of reference documents. - [DYNAMIC_EXECUTION]: The
setup_env.pyscript manages a local Python virtual environment within the user's cache directory (~/.cache/anolisa/.venv/). It usessubprocessto callpipfor installing standard dependencies includingrequests,beautifulsoup4, andmarkdownify. - [EXTERNAL_DOWNLOADS]: The
crawl_docs.pyscript fetches documentation updates fromhelp.aliyun.com. As Alibaba Cloud is a well-known service, these downloads are considered legitimate functionality for a documentation assistant. - [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data by crawling external web pages. It implements security best practices by using
markdownifyto strip potentially dangerous HTML tags (such as<script>,<style>, and<nav>) and prepending metadata headers to the content before it is processed by the AI agent. The capability inventory includes shell execution and file system management, but these are scoped to the agent's primary purpose as an OS assistant.
Audit Metadata