Fluencer Predict Track
投资大V合订本:通过可配置的本机 Chrome CDP endpoint 抓取知乎回答页、类似公开回答列表,或根据用户名搜索出的投资/财经账号,筛选不同时间跨度的预测性内容,检查最后编辑时间是否早于预测事件,用外部公开数据验证结果,并输出带来源的 Markdown 报告。适用于用户要求分析投资大V、财经博主、知乎用户或公开发言者的预测准确性、历史预测记录、forecast verification、prediction tracking 或合订本式复盘;如果用户只给用户名而非网址,按搜索结果最靠前且确实发表过投资观点的账号为准。
Install
npx skills add wzj998/fluencer-predict-track --skill fluencer-predict-trackgit clone --depth 1 https://github.com/wzj998/fluencer-predict-track.git /tmp/fluencer-predict-track
mkdir -p ~/.claude/skills
cp -r /tmp/fluencer-predict-track/.claude/skills/fluencer-predict-track ~/.claude/skills/fluencer-predict-trackPersonal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.
From the SKILL.md
# 投资大V合订本 ## 核心约定 使用本机 Chrome/Chromium CDP endpoint 访问公开页面和交互式登录会话。 默认 endpoint: ```text http://127.0.0.1:15166 ``` 可覆盖方式: - 环境变量:`FLUENCER_CDP_ENDPOINT` - 脚本参数:`-Endpoint` - 用户在任务中明确提供 endpoint 开始浏览前先运行辅助脚本: ```powershell .\skills\fluencer-predict-track\scripts\ensure-chrome-cdp.ps1 ``` 不要读取、导出、打印或上传 cookie、localStorage、浏览器 profile、密码或 storage-state JSON。如果目标站点要求登录,让用户在浏览器窗口中完成登录,然后继续。 ## 工作流程 1. 解析目标和验证截止日。除非用户另有指定,验证截止日使用任务当天。 - 如果用户给的是 URL,直接使用该 URL。 - 如果用户只给用户名、昵称或中文名,先用网页搜索或站内搜索查找候选账号。 - 搜索时优先组合投资语境关键词,例如 `<用户名> 投资`、`<用户名> 股票`、`<用户名> A股`、`<用户名> 知乎 回答`。 - 按搜索结果顺序检查候选账号,选择最靠前且公开内容中确实发表过投资/财经观点的账号。 - 如果多个候选都符合,仍以搜索排序更靠前者为准;如果没有候选发表过投资观点,说明无法确定目标,不要强行分析同名无关账号。 - 在报告中记录搜索词、候选判断和最终选择的目标 URL。 2. 使用 `scripts/ensure-chrome-cdp.ps1` 确认 CDP 可用。 3. 通过浏览器自动化连接该 CDP endpoint。 4. 打开目标用户的公开回答页或帖子列表。 5. 优先在浏览器上下文中调用页面自己的接口,不要一开始就模拟滚动。知乎常见接口示例: ```text https://www.zhihu.com/api/v4/members/<url-token>/answers?offset=0&limit=20&sort_by=created&include=data[*].content,voteup_count,comment_count,created_time,updated_time,question ``` 6. 采集回答或帖子的链接、标题、创建时间、最后编辑时间和纯文本正文。 7. 建立预测候选池。优先选择具备下列特征的内容: - 有明确未来窗口:日期、年份、季度、明年、年底前、未来几个月等。 - 有可验证结果:价格、指数点位、政策事件、宏观数据、公司或行业事件。 - 有清晰方向或阈值:上涨/下跌、突破/不突破、发生/不发生、跑赢/跑输。 8. 排除下列内容: - 最后编辑时间晚于预测事件或验证窗口。 - 预测过于模糊,必须发明口径才能验证。 - 关键判断主要依赖无法可靠读取的图片。 - 预测目标日在验证截止日之后。除非用户要求列出,否则这类内容只标记为 `尚未到期`,不计入准确率。 9. 除非可验证候选不足,否则每个目标至少选择 20 条。多目标任务按目标分别计数,不要把不同账号的样本合并后凑满 20 条。每个目标的样本都要尽量覆盖不同年份或时间段,不要只选最近内容。
Details
- Skill name
fluencer-predict-track- Repository
- wzj998/fluencer-predict-track
- Path
.claude/skills/fluencer-predict-track/SKILL.md- License
- MIT
- Repo updated
- 29 June 2026
FAQ
What does the Fluencer Predict Track skill do?
投资大V合订本:通过可配置的本机 Chrome CDP endpoint 抓取知乎回答页、类似公开回答列表,或根据用户名搜索出的投资/财经账号,筛选不同时间跨度的预测性内容,检查最后编辑时间是否早于预测事件,用外部公开数据验证结果,并输出带来源的 Markdown 报告。适用于用户要求分析投资大V、财经博主、知乎用户或公开发言者的预测准确性、历史预测记录、forecast verification、prediction tracking 或合订本式复盘;如果用户只给用户名而非网址,按搜索结果最靠前且确实发表过投资观点的账号为准。
How do I install the Fluencer Predict Track skill?
skills CLI (Claude Code, Cursor, Codex and more): npx skills add wzj998/fluencer-predict-track --skill fluencer-predict-track. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.
Is Fluencer Predict Track popular?
It lives in wzj998/fluencer-predict-track, which has 43 GitHub stars and 12 forks.
Related skills
Zhihu Report Polisher
43wzj998/fluencer-predict-track
Convert an internal influencer prediction report Markdown into a Zhihu-ready public article draft, preview it in Zhihu's editor, or prepare a publish flow. Use when the user wants to publish or prepare a consolidated report for Zhihu, especially when the source report contains local paths, Chrome CDP details, report file names, iteration wording, or engineering notes that should not appear in a public article.
Using Datahub
46datahub-project/datahub-skills
This skill provides routing guidance for all DataHub interaction skills. It is injected at session start and helps map user intent to the correct skill. Do not invoke this skill directly — it is loaded automatically.
Datahub Quality
46datahub-project/datahub-skills
Use this skill when the user wants to manage data quality in DataHub: create or run assertions, check assertion outcomes, raise or resolve incidents, create notification subscriptions, or diagnose health problems across their estate. Triggers on: "create assertion", "run assertion", "check quality", "data quality", "health check", "raise incident", "resolve incident", "subscribe to", "failing assertions", "active incidents", or any request involving data quality, assertions, incidents, or quality notifications.
My Phone Skill
49function1st/PhoneZero
Replace this. Say what the skill does and when to use it (the router). Copy this folder to ~/.phonezero/skills/<name>/ or ~/.cursor/skills/<name>/. Do not put keys or real phone numbers here.
Datahub Enrich
46datahub-project/datahub-skills
Use this skill when the user wants to add or update metadata in DataHub: descriptions, tags, glossary terms, ownership, deprecation, domains, data products, structured properties, documents, or field-level metadata. Triggers on: "add tag to X", "update description for X", "set owner of X", "add glossary term", "deprecate X", "create a domain", "create a glossary term", "add a document", or any request to modify DataHub metadata.
Phonezero
49function1st/PhoneZero
Deprecated alias. Use phonezero-runtime for setup and dial, book-restaurant for table reservations. Invoke only when an old chat still names this skill.
Browse the most popular Claude skills, all other skills, or write your own with the skills generator.