菜单
Research emerging trends in a market or technology space. Returns signal vs noise breakdown with evidence. Use when the user asks what's changing in a market, what trends to watch, or what's coming next.
默认使用会先检查来源的 Prompt；你也可以切换为直接命令，或下载本地副本。
决定是否安装前，请先阅读 SKILL.md，以及 SkillsMP 当前展示的配套文件。
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt；运行前请先检查来源。
npx skills add https://github.com/hamzafarooq/multi-agent-course --skill trend-spotter命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地？可下载 SkillsMP 当前能够提供的文件。
Generate a Product Evaluation report (.md, optionally PDF) for FDE Assignment 3 — Moment Search at Scale. Runs the automated rubric + benchmark (incl. --resilience), then does a real cross-source test (ingest a video + a research paper + a deck the student doesn't control and verify one query cites all three), and writes PRODUCT_EVAL.md. Use when the student wants to evaluate their Moment Search product, produce their submission report, or "eval my product".
Generate a Product Evaluation report (.md, optionally PDF) for the FDE Assignment 1 Live Translate product. Runs the automated rubric + benchmark, then does a real live-website test (e.g. homedepot.com), and writes PRODUCT_EVAL.md. Use when the student wants to evaluate their translation product, produce their submission report, or "eval my product".
Turn an Anthropic Agent Skill into a runnable standalone Python app built on the Anthropic SDK. Use this whenever the user wants to "publish", "export", "package", or "turn into an app" a skill, or asks for the "SDK version" / "API version" / "standalone version" of a skill, or wants a script that loads a SKILL.md and runs it via the Claude API outside of Claude Code. Trigger whether the user provides an actual SKILL.md file OR just names/describes a skill in text — in the latter case, generate the SKILL.md first, then wrap it. Use this even if the user only says something like "make my X skill into a program" without saying "SDK" explicitly.
基于 SOC 职业分类
正在显示 SKILL.md
| name | trend-spotter |
| description | Research emerging trends in a market or technology space. Returns signal vs noise breakdown with evidence. Use when the user asks what's changing in a market, what trends to watch, or what's coming next. |
Ask the user: "What market or technology space do you want to scan? Any specific angle (e.g. enterprise adoption, pricing models, new entrants)?"
Then research and return:
3–4 trends with evidence: funding rounds, product launches, job posting spikes, pricing shifts. One paragraph each.
2–3 things that are nascent but directional. Explain what would need to be true for each to become a strong signal.
1–2 things getting hype that are unlikely to materially change the market in the next 18 months. Say why.
Two sentences on what a product team should do differently based on these trends.
Cite specific evidence for every strong signal (company names, dates, data points). Avoid vague statements like "AI is changing everything."