期货早报生成工具。用户输入网址或留空默认使用西本资讯，自动抓取当天最新行业要闻早餐，按固定HTML格式输出，包含：整体综述（置顶）、重点关注（3条）、信息速览（行情表格）、宏观要闻（10条+详情）、产业要闻（6条+详情）。触发词：生成早报、期货早报、早间新闻、行业要闻早餐、早班车、卓钢早报。

What to consider before installing: - Confirm you are comfortable with the skill saving files to your Desktop and with the agent (runtime) having filesystem write permission. If you run the agent on a shared machine, generated files may be visible to others. - The skill explicitly hides source attribution in every output. That can create copyright, audit, or trust issues — if you need provenance or must keep attribution, do not use this behavior or modify the template/instructions. - The provided HTML template contains external image/link URLs (wework.qpic.cn, open.work.weixin.qq.com, etc.). Opening the generated HTML will cause your browser or agent to contact those servers, which can expose your IP and that you opened the file. Remove or inline external assets if you need privacy, or configure the agent to sanitize/strip remote resources. - Avoid pointing the tool at private/internal URLs or pages requiring authentication unless you understand how the agent handles fetched content and whether outputs will be transmitted off-host. Recommended mitigations: - Inspect and edit assets/template.html to remove or localize external resources before use. - Change the output-format to include source attribution if you require provenance. - Test the skill on a public, non-sensitive URL first to observe behavior. - Ask the maintainer (or provide additional docs) about where the agent runs (local vs. cloud), whether outputs are uploaded automatically, and whether screenshots or generated files are transmitted to external services — this information would increase confidence and could change the verdict to benign.

Type: OpenClaw Skill Name: futures-morning-brief Version: 1.0.0 The skill bundle is a specialized tool for generating financial reports (futures morning briefs) by scraping industry news from specified or default websites (e.g., steelx2.com). While the workflow involves high-privilege agent actions such as writing HTML files to the user's desktop and taking screenshots for verification (SKILL.md), these behaviors are transparently documented and directly support the stated functionality. No evidence of data exfiltration, malicious payloads, or harmful prompt injection was found in the code or instructions.