用户反馈收集，问卷 + 分析 + 改进。

This skill appears to be what it says, but the SKILL.md tells you to clone and run code from GitHub without verification. Before installing or running it: (1) Inspect the repository contents (especially requirements.txt and app.py) to understand what the code does and what network endpoints it contacts; (2) Run pip install and python app.py only inside an isolated environment (container/VM) or virtualenv, not as root and not on a machine with sensitive data or credentials; (3) Check dependency versions and consider pinning or auditing requirements for malicious packages; (4) Verify the repository owner and review license and recent commits; (5) If you need to trust this in production, request a packaged install spec with checksums/pinning or ask the author for clarifying documentation about data flows and required secrets. If you cannot review the repo, treat this as untrusted code and avoid running it on sensitive systems.

Type: OpenClaw Skill Name: ai-intelligent-feedback-collection Version: 1.0.0 The skill bundle contains only metadata and documentation for an AI-driven feedback collection system. No executable code, suspicious network calls, or malicious instructions were found in the provided files (_meta.json and SKILL.md). The content is purely descriptive and aligns with its stated purpose.