Token导航 LogoToken导航TokenDH.com
研究检索权限需确认github未标认证来源可访问许可证需确认审计通过

video-analytics-interpreter视频分析解释器

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

489

周安装

21

GitHub Stars

175

下载量

171
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:video-analytics-interpreter(视频分析解释器)
来源仓库:https://github.com/nicepkg/ai-workflow
仓库路径:skills/video-analytics-interpreter
安装命令:
npx skills add https://github.com/nicepkg/ai-workflow --skill video-analytics-interpreter
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/nicepkg/ai-workflow --skill video-analytics-interpreter

简介

video-analytics-interpreter 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。

  • 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。
  • 通过 npx skills add 命令从指定仓库安装,需结合原始 README 核验具体用法。
  • 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。
  • 适用宿主包括 Codex、Claude、Cursor、Gemini CLI,接入前应确认版本、权限和运行环境要求。

SKILL.md

Video Analytics Interpreter

Transform raw video metrics into actionable growth strategies.

Key Metrics Explained

YouTube Analytics

📊 CORE METRICS:

VIEWS
- What: Total video plays (30+ seconds or full if shorter)
- Good: Trending upward week-over-week
- Warning: Sudden drops may indicate algorithm changes

WATCH TIME
- What: Total minutes watched
- Why it matters: #1 factor for YouTube algorithm
- Good: Higher than channel average

AVERAGE VIEW DURATION (AVD)
- What: Average time viewers watch
- Benchmark: 50%+ of video length is good
- Tip: Longer videos = lower % is acceptable

CLICK-THROUGH RATE (CTR)
- What: Impressions → Clicks percentage
- Good: 4-10% (varies by content type)
- Excellent: 10%+
- Warning: <2% needs thumbnail/title work

IMPRESSIONS
- What: Times thumbnail shown to users
- Note: Higher impressions = YouTube promoting you
- Tip: CTR × Impressions = Views potential

AUDIENCE RETENTION
- What: Graph showing when viewers leave
- Key: Look for drop-off points
- Goal: Flat line is ideal, gradual decline acceptable

ENGAGEMENT RATE
- What: (Likes + Comments) / Views
- Good: 4-8%
- Excellent: 8%+

TikTok Analytics

📊 TIKTOK METRICS:

VIDEO VIEWS
- Includes replays and loops
- Higher than YouTube due to autoplay

AVERAGE WATCH TIME
- Critical for algorithm
- Goal: Above 100% (indicates replays)

WATCH FULL VIDEO RATE
- % who watched entire video
- Good: 30%+ for 15-30 sec videos
- Excellent: 50%+

ENGAGEMENT RATE
- (Likes + Comments + Shares) / Views
- Good: 5-10%
- Viral potential: 15%+

SHARES
- Most important engagement type
- Strong shares = algorithm boost
- Indicates "save for later" or "send to friend"

PROFILE VIEWS
- Viewers who clicked your profile
- Indicates content sparked curiosity
- Goal: 1-3% of views

FOLLOWER CONVERSION
- New followers / Profile views
- Good: 10-20%
- Tip: Pin best content, optimize bio

Analytics Interpretation Framework

Step 1: Identify the Pattern

PERFORMANCE CATEGORIES:

🚀 BREAKOUT SUCCESS (Top 10% of your content)
- Views: 3x+ your average
- CTR: Above your channel average
- Retention: Higher than similar videos
- Action: Double down, create more like this

✅ SOLID PERFORMER (Above average)
- Views: 1.5-3x your average
- CTR: At or above average
- Retention: Consistent with similar content
- Action: Note what worked, iterate

😐 AVERAGE
- Views: Near your typical numbers
- CTR: Around channel average
- Retention: Normal patterns
- Action: Test new elements

⚠️ UNDERPERFORMER (Below average)
- Views: Below your average
- CTR: Lower than normal
- Retention: Early drop-offs
- Action: Analyze what went wrong

❌ FLOP (Bottom 10%)
- Views: Significantly below average
- CTR: Much lower than normal
- Retention: Severe early drop-off
- Action: Don't delete - learn from it

Step 2: Diagnose the Problem

LOW VIEWS DIAGNOSIS:

Q1: Is CTR low?
YES → Thumbnail/title problem
NO → Algorithm not promoting (impressions issue)

Q2: Is retention low?
YES → Content/hook problem
NO → Distribution/timing issue

Q3: Low impressions but good CTR/retention?
YES → Algorithm testing phase, be patient
     → Or topic has limited search volume

DIAGNOSIS FLOWCHART:
┌─────────────────────────────────────────────────────────┐
│ Low Views?                                              │
│     ↓                                                   │
│ Check CTR → Low? → Fix thumbnail/title                  │
│     ↓                                                   │
│ CTR OK? → Check Retention → Low? → Fix hook/content     │
│     ↓                                                   │
│ Both OK? → Check Impressions → Low? → Topic/timing issue│
│     ↓                                                   │
│ All OK but low views? → Give it time, algorithm testing │
└─────────────────────────────────────────────────────────┘

Step 3: Read the Retention Graph

RETENTION PATTERNS:

📉 EARLY DROP (0-30 seconds)
Problem: Weak hook, misleading thumbnail/title
Fix: Stronger hook, match expectations

📉 GRADUAL DECLINE (Steady downward slope)
Normal: Expected for most videos
Optimize: Add pattern breaks every 30-60 sec

📉 CLIFF DROP (Sudden sharp decline)
Problem: Boring section, promise unfulfilled
Fix: Cut/improve that section, better pacing

📈 SPIKE (Retention increases)
Meaning: Replay/rewind point
Opportunity: Create clip from this moment

─── FLAT LINE (Stable)
Ideal: Viewers engaged throughout
Indicates: Strong content-hook match

Analytics Report Template

═══════════════════════════════════════════════════════════════
VIDEO ANALYTICS REPORT
Video: [Title]
Published: [Date]
Analysis Period: [X days since publish]
═══════════════════════════════════════════════════════════════

📊 PERFORMANCE SUMMARY:
─────────────────────────────────────────────────────────────
Views: [X] ([+/-X%] vs channel avg)
Watch Time: [X hours] ([+/-X%] vs channel avg)
CTR: [X%] ([+/-X%] vs channel avg)
Avg View Duration: [X:XX] ([X%] of video)
Engagement: [X%] (Likes: X, Comments: X)

Performance Tier: [🚀 Breakout / ✅ Solid / 😐 Average / ⚠️ Under / ❌ Flop]

📈 RETENTION ANALYSIS:
─────────────────────────────────────────────────────────────
Hook Effectiveness (0-30s): [Strong/Moderate/Weak]
- [X%] still watching at 30 seconds

Key Drop-off Points:
- [Timestamp]: [X%] dropped - Likely cause: [reason]
- [Timestamp]: [X%] dropped - Likely cause: [reason]

Rewatch Spikes:
- [Timestamp]: Viewers replayed - Content: [what happened]

Overall Shape: [Early drop / Gradual decline / Cliff / Flat]

🎯 TRAFFIC SOURCES:
─────────────────────────────────────────────────────────────
1. [Source]: [X%] - [Insight]
2. [Source]: [X%] - [Insight]
3. [Source]: [X%] - [Insight]

Best Performing Source: [Source] - Why: [explanation]
Underperforming Source: [Source] - Recommendation: [action]

👥 AUDIENCE INSIGHTS:
─────────────────────────────────────────────────────────────
New vs Returning: [X% new / X% returning]
Geographic: Top countries [list]
Demographics: Primary age/gender [if available]
Device: [X% mobile / X% desktop / X% TV]

Subscriber Impact: [+X subscribers] ([X%] conversion)

🔍 DIAGNOSIS:
─────────────────────────────────────────────────────────────
Primary Issue: [Identify main problem if any]
Root Cause: [Why this happened]
Secondary Issues: [Other concerns]

💡 RECOMMENDATIONS:
─────────────────────────────────────────────────────────────
IMMEDIATE (This Video):
1. [Action] - Expected impact: [result]
2. [Action] - Expected impact: [result]

FUTURE VIDEOS:
1. [Lesson learned] - Apply to: [future content]
2. [Lesson learned] - Apply to: [future content]

A/B TEST SUGGESTION:
- Test: [Element to test]
- Hypothesis: [What you expect]
- Measure: [Metric to track]

🎬 CLIP OPPORTUNITY:
─────────────────────────────────────────────────────────────
High-engagement moment at [Timestamp]: "[Description]"
Recommended for: [TikTok/Shorts/Reels]
Hook angle: "[Suggested hook]"

📅 NEXT STEPS:
─────────────────────────────────────────────────────────────
[ ] [Action item 1]
[ ] [Action item 2]
[ ] [Action item 3]
═══════════════════════════════════════════════════════════════

How to Use

Full Analytics Review

Analyze this video's performance:
[Paste analytics data or describe metrics]

Comparisons:
- Channel average CTR: [X%]
- Channel average retention: [X%]
- Similar video performance: [description]

Quick Diagnosis

My video has [X views] but my CTR is [X%].
Why is it underperforming and how do I fix it?

Trend Analysis

Analyze these videos' performance trends:
Video 1: [Title] - [Views, CTR, Retention]
Video 2: [Title] - [Views, CTR, Retention]
Video 3: [Title] - [Views, CTR, Retention]

What patterns do you see?

Channel Health Check

Here are my last 10 videos' metrics:
[List videos with key metrics]

How is my channel performing overall?
What should I focus on?

Benchmarks by Content Type

                    CTR      Retention    Engagement
Educational         3-6%     40-60%       3-6%
Entertainment       5-10%    30-50%       5-10%
Gaming              4-8%     25-45%       4-8%
Vlog                3-7%     35-55%       4-8%
Product Review      4-8%     40-60%       3-6%
Tutorial            3-6%     45-65%       3-5%
Short-form          8-15%    70-100%+     8-15%

Warning Signs to Watch

⚠️ IMMEDIATE ATTENTION:
- CTR dropped 50%+ from average
- Retention cliff in first 30 seconds
- Impressions declining week-over-week
- Subscriber loss instead of gain

📉 CONCERNING TRENDS:
- AVD decreasing over time
- Engagement rate declining
- More dislikes than usual
- Comments increasingly negative

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

35.53%
按下载量换算61

Claude

30.98%
按下载量换算53

Cursor

17.23%
按下载量换算29

Gemini CLI

9.84%
按下载量换算17

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

权限需确认

当前来源未能明确判断权限范围,默认进入异常复核队列。

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