Clawbrowser Skill
Security checks across malware telemetry and agentic risk
Overview
The skill's code, runtime instructions, and required binaries are coherent with a headless browser automation / web-scraping tool, but it writes cache/output to the user's home directory and includes optional external upload behavior (Feishu) that you should review before use.
This skill appears to do what it says: it uses the agent-browser CLI to automate a headless browser and save scraped pages. Before installing, consider the following: - agent-browser will download Chromium (large network download) when you run 'agent-browser install' — verify the agent-browser package source and trust it. - The skill creates caches and saves scraped output under your home directory (~/.openclaw/cache/browser and ./scraped_content). Expect persistent files; review and clear them if needed. - The tool accepts file:// URLs and can snapshot pages; do not allow it to open local-sensitive paths (e.g., file:///etc/passwd, private files) unless you trust the running agent and environment. - web_scraper has an optional Feishu upload path (uses an imported feishu_doc helper). If you enable saving to Feishu, ensure you know what credentials will be used and that you trust the Feishu integration; the skill metadata does not declare or restrict those credentials. - Because the skill executes a CLI (subprocess.call to agent-browser), run it in an isolated environment (container or VM) if you are unsure, and inspect the included Python files yourself. If you need higher assurance, ask the author for a signed release or a vetted agent-browser package source. Overall: internally coherent for browser automation, but treat external uploads and file:// access with care.
SkillSpector
SkillSpector findings are pending for this release.
VirusTotal
No VirusTotal findings