Someant

创建飞书在线文档（Feishu/Lark Document），将 Markdown 内容输出为可分享的飞书文档链接

长桥（LongBridge）港股美股行情API，提供实时股票行情、财务数据、市场数据等

管理 OmniClaw 定时任务，支持创建、修改、删除定时推送任务

联网搜索与实时资讯查询。优先使用 Google 搜索引擎，支持并行多查询，以 Perplexity 风格（内联引用编号 + 原始来源链接）综合输出。适用于：最新新闻/资讯、实时信息（价格/事件/公告）、核实说法、话题背景调研。触发词：搜一下、查一下、最新消息、最近新闻、帮我查、有没有关于…的消息、搜索。

paper-read 流水线的 subagent 单篇精读 skill。由 paper-read 主 agent 启动，不直接由用户调用。读取单篇论文 PDF 全文，执行深度分析 + OpenJudge 五阶段审稿，将结果写入 /tmp/arxiv_results/{arxiv_id}.json。

自动爬取 arXiv cs.CV 当天所有新论文，两阶段筛选：先由 agent 读全量摘要快速初筛，再对 Top N 下载 PDF 全文精读；精读阶段从 paper-read-reviewer skill 注入完整 subagent prompt，结合 OpenJudge 五阶段审稿流水线综合打分，生成报告后清理本地 PDF。当用户要求"阅读论文"、"分析今日 arxiv 论文"、"paper-read"、"论文总结"时使用此 skill。

将单篇论文解读成高信息密度、适合人直接阅读的中文精读稿。默认输出“导语 → 问题 → 方法 → 图解 → 实验 → 评价 → 一句话总结”的结构，风格参考 2026-04-15 cs.CV Top10 报告中的单篇精读部分。当用户要求“精读这篇论文”“详细解读某篇 paper”“像日报里那样展开讲一篇论文”“单篇论文深度总结”时使用此 skill。

Use this skill when the user wants to write, rewrite, or template content for WeCom/Enterprise WeChat documents, especially news briefs, research digests, daily reports, weekly reports, or other information-dense content that must read well after pasting into a WeCom doc. This skill is for formatting around WeCom's lightweight Markdown support, stable heading/list structure, and scannable Chinese business writing.

Analyze images and videos using Z.AI GLM-4V vision model. Use when the user needs to analyze screenshots or UI images, extract text from images (OCR), diagnose error screenshots, understand technical diagrams or architecture charts, analyze data visualizations, compare two UI screenshots, or generate code from a UI image. Run scripts/analyze.py via Bash.

Fetch and extract content from any webpage by URL using Z.AI Web Reader. Use when the user provides a specific URL and wants its content read, summarized, or analyzed — API docs, articles, open-source project pages, blog posts. Run scripts/read.py via Bash.

Search the web for real-time information using Z.AI Web Search Prime. Use when the user needs current information not in your training data — latest news, recent events, current prices, live data (weather, stocks), or up-to-date documentation. Run scripts/search.py via Bash.

Explore GitHub repositories in depth using Z.AI ZRead — search documentation, browse directory structure, and read source files. Use when the user needs to understand an open-source project, explore a GitHub repo's codebase, read specific files, find issues or PRs, or analyze how a library works. Run scripts/zread.py via Bash.

Use MiniMax MCP to analyze images, extract text (OCR), identify objects, describe visual content, and answer questions about images. Use when the user wants to understand, describe, or extract information from image files or URLs via MiniMax API.

Use MiniMax MCP to perform web searches and retrieve up-to-date information from the internet. Use when the user asks to search the web, find current information, look up news, or research topics online via MiniMax API.