Token导航 LogoToken导航TokenDH.com
前端设计敏感数据github未标认证来源可访问许可证需确认审计通过

analytics分析

Agent Skill

用于辅助数据整理、表格处理、CSV/Excel 分析、指标计算和图表准备。它适合让 Agent 清洗字段、汇总数据、发现异常、生成统计口径或把分析结果转成可读说明。使用时需要确认数据来源、字段含义和时间范围,避免把样本数据当全量事实;涉及敏感数据、导出文件或批量写回时,应先确认权限和脱敏边界。

总安装

269

周安装

11

GitHub Stars

168

下载量

87
CodexClaudeCursorGemini CLI

安装说明

本站只整理中文说明和来源信息,不托管安装包,也不代用户安装。

GitHub

来源数

2

许可证

unknown

最后核验

2026-05-01

来源状态

来源可访问

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:analytics(分析)
来源仓库:https://github.com/whawkinsiv/claude-code-skills
仓库路径:skills/analytics
安装命令:
npx skills add https://github.com/whawkinsiv/claude-code-skills --skill analytics
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/whawkinsiv/claude-code-skills --skill analytics

简介

analytics 用于辅助数据整理、表格处理和指标计算。

  • 适合在 Codex、Claude、Cursor、Gemini CLI 中清洗字段、汇总数据或发现异常。
  • 通过 npx skills add 命令从指定仓库安装并使用。
  • 使用时需确认数据来源和时间范围,避免把样本当全量事实。
  • 涉及敏感数据导出时应先确认脱敏边界和权限。

SKILL.md

Analytics & Instrumentation

Track decisions, not dashboards. This skill helps you set up analytics that answer the questions that matter — without drowning in data.

Core Principles

  • Every metric should connect to an action you'd take differently based on the result.
  • Instrument events at build time, not as an afterthought. It's 10x easier.
  • More data ≠ better decisions. Focus on 5-10 key metrics per stage.
  • Data quality matters more than data quantity. One clean funnel beats fifty broken ones.

Choosing an Analytics Tool

You don't need to write code to track events. Pick a tool, add the snippet, and configure in the UI.

Recommended by Stage

StageToolWhyCost
Pre-revenuePostHogProduct analytics, funnels, session replay. Built for SaaS.Free (1M events/mo)
Pre-revenuePlausibleSimple, privacy-first traffic analytics.$9/mo
$0-5K MRRPostHog or MixpanelDeeper funnel analysis, cohort retentionFree tier or ~$200/mo
$5K+ MRRAmplitude or HeapAdvanced analytics, experimentation$500+/mo

Recommendation for solo founders: Start with PostHog (free tier). It covers product analytics, funnels, session recordings, and feature flags — all things you'll need.

For basic traffic analytics (where visitors come from), add Plausible or Google Analytics 4 (free).

Setup

Tell AI:

Set up PostHog analytics:
1. Create a PostHog project and get the API key
2. Add the PostHog JavaScript snippet to our app (in the <head> or layout component)
3. Verify it's tracking page views by checking the PostHog dashboard
4. Configure: exclude admin/test accounts from tracking

What to Track (By Stage)

Pre-Launch (Just These 3)

  1. Landing page visits (traffic source)
  2. Waitlist/signup conversions
  3. Where visitors drop off

0-$1K MRR (Add These)

  1. Signup → activation rate (completed key action)
  2. Time to activate (hours from signup to key action)
  3. D7 retention (% returning after 1 week)
  4. Free-to-paid conversion

$1K-$10K MRR (Add These)

  1. Feature adoption (% of users using each feature weekly)
  2. Cohort retention curves (track each month's signups over time)
  3. Expansion revenue triggers (what behavior precedes upgrades)
  4. Churn rate and reasons

Event Naming Conventions

Use a consistent schema so events make sense to anyone reading them.

Format: [Object] [Action] in past tense, Title Case.

Event NameWhen It Fires
Signup Form SubmittedUser completes registration
Project CreatedUser creates their first/any project
Dashboard ViewedUser visits the main dashboard
Report ExportedUser downloads or shares a report
Subscription UpgradedUser moves to a higher plan
Invite SentUser invites a teammate

Rules:

  • Object first, then action (makes grouping in tools easier)
  • Past tense (the event already happened)
  • Be specific: "Signup Form Submitted" not "Form Submitted"
  • Never include personal data in the event name

Event Properties

Every event should include context:

"Project Created"
  template_used: "blank" | "marketing" | "engineering"
  source: "dashboard" | "onboarding" | "api"

"Subscription Upgraded"
  from_plan: "free"
  to_plan: "pro"
  billing_period: "annual"
  trigger: "usage_limit" | "feature_gate" | "self_serve"

Tell AI:

Set up event tracking for our core funnel:
- Track these events: [list your key events from the table above]
- Include these properties with each event: [relevant properties]
- Set user-level properties on identify: plan, signup_date, role
- Verify events are firing correctly in PostHog/Mixpanel

Core Metrics Framework

The 5 Metrics That Matter

MetricWhat It Tells YouHow to Calculate
Signup rateIs your marketing working?Visitors → Signups
Activation rateIs your onboarding working?Signups → Completed key action
D7 retentionDoes your product deliver ongoing value?% of users returning 7 days later
Free-to-paidIs your product worth paying for?Free users → Paying users
MRRIs your business growing?Sum of all monthly subscription revenue

Setting Up Your Dashboard

Tell AI:

Create an analytics dashboard with these 5 charts:
1. Daily signups (line chart, last 30 days, by source)
2. Activation rate (% of signups who [completed key action], last 30 days)
3. D7 retention (% of each week's signups who returned 7 days later)
4. Free-to-paid conversion rate (last 30 days)
5. MRR (line chart, all time)

This should be the first thing I see when I open analytics.

Funnel Instrumentation

Define your core funnel and track every step:

Typical SaaS funnel:
  1. Landing Page Viewed
  2. Signup Form Viewed
  3. Signup Form Submitted
  4. Onboarding Started
  5. [Key Activation Action] Completed
  6. Second Session (returned next day/week)
  7. Upgrade Page Viewed
  8. Subscription Created

The biggest drop-off between steps = your biggest opportunity. Fix that first.

Tell AI:

Instrument our conversion funnel:
- Track these events in order: [list your funnel steps]
- Build a funnel visualization showing drop-off between each step
- Set up an alert if any step's conversion drops below [X]% week-over-week

Privacy and Compliance

Add these before you launch:

  • Add analytics tool reference to your Privacy Policy
  • Add cookie consent banner if required (depends on tool and geography)
  • Enable "anonymize IP" in your analytics tool
  • Exclude admin/test accounts from tracking
  • Never track PII (email, name) in event properties without consent
  • Allow users to opt out of tracking
  • Respect Do Not Track (DNT) browser settings

Tell AI:

Add privacy compliance for analytics:
- Add a cookie consent banner that loads analytics only after consent
- Exclude users with DNT enabled
- Add an "opt out of tracking" toggle in account settings
- Ensure no PII is sent in event properties

Common Mistakes

MistakeFix
Tracking everything from day oneStart with 5 metrics. Add more when you have questions they can't answer.
No analytics until post-launchAdd the snippet during build. Track from first user.
Dashboard but no actionsEvery metric should have an "if this drops, I'll do X" response
Ignoring data qualityCheck that events are firing correctly once a week for the first month
Paying for expensive tools too earlyPostHog free tier handles most needs until $5K MRR

If You Can't Set Up Analytics Right Now

Track these manually in a spreadsheet until you can:

WeekSignupsActivatedRetained (D7)PaidMRR
Week 1
Week 2

Even rough manual tracking beats flying blind. Graduate to a real tool as soon as you can.

适合场景

01

用户想查找某类 Agent Skill 时

02

需要根据任务场景推荐可安装能力包时

03

需要对比不同来源的安装命令和来源信息时

能力概览

能力 1

按任务关键词查找相关 Skills

能力 2

展示可复制的安装命令

能力 3

保留来源站点、仓库和原始说明,方便继续核验

能力 4

展示第三方安全扫描或审计结果

安装后应在对应宿主中按原始 README 的触发条件使用;具体调用方式请以来源页面和 README 为准。

平台分布

Codex

37.83%
按下载量换算33

Claude

29.7%
按下载量换算26

Cursor

17.92%
按下载量换算16

Gemini CLI

10.29%
按下载量换算9

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

敏感数据

该 Skill 可能接触密钥、Token、环境变量或敏感配置,应进入高风险复核队列,默认不自动发布。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。当前只有一个来源,正式发布前建议补源仓库或其他目录站核验。

来源信息

继续浏览同类 Skills