📦 AI每日最新进展日报 — 技能工具
v1.0.0生成一份当日 LLM / 大模型领域的中文资讯简报。采集来源包括：**8 家头部厂商官方博客**（海外：OpenAI / Anthropic / Google DeepMind / Meta AI / xAI / Mistral；国内：DeepSeek / 通义千问 Qwen —— 最高优先级一手源）、GitHu...
0· 66·0 当前·0 累计
by @flowbywind 安全扫描
OpenClaw
安全
high confidenceThe skill's declared purpose (generate a daily Chinese LLM/large-model digest) matches its instructions and requirements — it only scrapes public sources, deduplicates, scores, and writes a Markdown file to the user's workspace without requesting extra credentials or installs.
评估建议
This skill is internally consistent with its goal: it scrapes public vendor blogs, GitHub, arXiv, Hugging Face, HN/X, and Chinese media, then writes a Markdown digest to ~/.openclaw/workspace/digests/. Consider these before enabling: (1) it needs outbound network access and the platform's browser/fetch tools to work — network restrictions will break it; (2) some sources (X/Twitter) may be behind login — the skill states it will skip rather than request credentials, but if you supply platform-lev...详细分析 ▾
✓ 用途与能力
Name/description match the runtime instructions: the SKILL.md explicitly lists the public sources to poll, the selection and formatting rules, and the output path. There are no unrelated required env vars, binaries, or config paths that would be disproportionate to producing a news digest.
✓ 指令范围
Instructions are narrowly scoped to fetching/rendering public pages (via the platform's browser/fetch tools), de-duplicating, scoring, formatting a Markdown report, saving it under ~/.openclaw/workspace/digests/, and returning the path + top 3 items. The skill does not instruct reading arbitrary local files or exfiltrating data to external endpoints.
✓ 安装机制
Instruction-only skill with no install spec and no code files — minimal risk from installation since nothing is written/configured by an installer.
✓ 凭证需求
The skill requires no environment variables or credentials. It may attempt to read public content on sites that sometimes require login (X/Twitter), but SKILL.md documents downgrade strategies and explicitly says to skip sources behind login rather than requesting credentials.
✓ 持久化与权限
always:false (default) and no requests to modify other skills or system-wide config. The skill writes its own output files under the user's OpenClaw workspace, which is within its stated purpose.
安全有层次，运行前请审查代码。
运行时依赖
无特殊依赖
安装命令
点击复制官方npx clawhub@latest install llm-daily-digest
镜像加速npx clawhub@latest install llm-daily-digest --registry https://cn.longxiaskill.com