Web Browsing.Bak

Browse and summarize websites, extract content from URLs, search the web for information. Use when user asks to visit a website, get webpage content, search...

What to consider before installing: - The skill will make HTTP requests to DuckDuckGo and any URLs you provide. Do not pass URLs that contain sensitive tokens or that require authentication unless you understand the risks. - The included Python script uses third-party libraries (requests, bs4) but the skill does not declare dependencies or provide an install step. Verify your environment has these packages or install them in a sandbox before running. - Metadata mismatch: _meta.json ownerId differs from the registry owner id and the package is listed as source: unknown with no homepage. This doesn't prove malice but reduces provenance and should make you more cautious. - Review the script before use (it is short and readable). Running it in an isolated environment (container or restricted VM) is recommended if you will fetch untrusted pages. - If you plan to have the agent fetch pages that require credentials or that are internal/private, do not provide secrets to this skill; it has no mechanism for safe credential handling. If you want to proceed, verify dependencies (requests, beautifulsoup4), and consider running the code in a sandboxed environment. If provenance is important, ask the publisher for source and homepage information or prefer a browsing tool from a known source.

Type: OpenClaw Skill Name: web-browsing-bak Version: 1.0.0 The skill bundle provides standard web searching and content fetching functionality as described. The Python script `scripts/search_web.py` uses legitimate libraries (requests, BeautifulSoup) to query DuckDuckGo and extract webpage text with appropriate timeouts and content limits. No evidence of data exfiltration, shell injection, or malicious prompt manipulation was found in `SKILL.md` or the supporting documentation.