Skip to content
Back to search

replycueai_public

AI-powered Claude Skills for YouTube comment analysis — brand relevance scoring, smart replies, KOL ROI reports, threat detection, and keyword extraction. Built by ReplyCue AI.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/molly554--replycueai-public/badge.svg)](/listings/molly554--replycueai-public)

HTML

<a href="/listings/molly554--replycueai-public"><img src="/api/listings/molly554--replycueai-public/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

1. Prerequisites

You need two API keys:

# Required for all skills
export ANTHROPIC_API_KEY="sk-ant-..."

# Required for YouTube-related skills (1, 3, 5)
export YOUTUBE_API_KEY="AIza..."

- Get your Anthropic API key at [console.anthropic.com](https://console.anthropic.com) - Get your YouTube API key at [Google Cloud Console](https://console.cloud.google.com/apis/credentials) (enable YouTube Data API v3)

2. Install Skills

Clone this repo and add the skills directory to your Claude Code configuration:

git clone https://github.com/YOUR_USERNAME/replycue-ai-skills.git

# Add to your project's .claude/settings.json
# or copy skills/ into your .claude/skills/ directory

3. Use in Claude Code

Just describe what you want in natural language:

> Analyze the comments on this YouTube video for my brand ShieldVPN:
  https://www.youtube.com/watch?v=VIDEO_ID

> Generate a professional reply to this comment:
  "Your product crashed 3 times today, worst purchase ever"

> Give me a KOL ROI report for this sponsored video:
  https://www.youtube.com/watch?v=VIDEO_ID
  Brand: Notion, Collaboration type: sponsored

> Extract brand keywords from https://notion.so

> Scan this video for brand safety threats:
  https://www.youtube.com/watch?v=VIDEO_ID
  Brand: Notion

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 206 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # ReplyCue AI — Claude Skills for YouTube Brand Intelligence
  2. [![Claude Code](https://img.shields.io/badge/Claude_Code-Skills-blue?logo=anthropic)](https://claude.ai/code)
  3. [![License: MIT](https://img.shields.io/badge/License-MIT-green.svg)](LICENSE)
  4. **AI-powered YouTube comment analysis for brands, marketers, and creators.** Analyze comments, detect threats, generate smart replies, and measure KOL collaboration ROI — all from your terminal with Claude Code.
  5. > Built by the team behind [ReplyCue AI](https://replycueai.com) — the AI comment moderation platform for YouTube.
  6. >
  7. <div align="center">
  8. <img src="./replycueai.gif" width="9600">
  9. </div>
  10. ---
  11. ## What's Inside
  12. | Skill | Description | Use Case |
  13. |-------|-------------|----------|
  14. | [YouTube Comment Analyzer](skills/youtube-comment-analyzer/) | Analyze video comments with 5-layer brand relevance scoring | "Is anyone talking about my brand in this KOL video?" |
  15. | [Smart Reply Generator](skills/smart-reply-generator/) | Generate brand-appropriate replies in 3 tones | "Help me write a professional reply to this complaint" |
  16. | [KOL ROI Report](skills/kol-roi-report/) | Comprehensive collaboration effectiveness report | "Was this $5K sponsorship worth it?" |
  17. | [Brand Keyword Extractor](skills/brand-keyword-extractor/) | Extract keywords from any website for SEO and monitoring | "What keywords should I track for my brand?" |
  18. | [Comment Threat Scanner](skills/comment-threat-scanner/) | Detect scams, fake claims, and brand safety risks | "Are there any dangerous comments I need to handle?" |
  19. ## Quick Start
  20. ### 1. Prerequisites
  21. You need two API keys:
  22. ```bash
  23. # Required for all skills
  24. export ANTHROPIC_API_KEY="sk-ant-..."
  25. # Required for YouTube-related skills (1, 3, 5)
  26. export YOUTUBE_API_KEY="AIza..."
  27. ```
  28. - Get your Anthropic API key at [console.anthropic.com](https://console.anthropic.com)
  29. - Get your YouTube API key at [Google Cloud Console](https://console.cloud.google.com/apis/credentials) (enable YouTube Data API v3)
  30. ### 2. Install Skills
  31. Clone this repo and add the skills directory to your Claude Code configuration:
  32. ```bash
  33. git clone https://github.com/YOUR_USERNAME/replycue-ai-skills.git
  34. # Add to your project's .claude/settings.json
  35. # or copy skills/ into your .claude/skills/ directory
  36. ```
  37. ### 3. Use in Claude Code
  38. Just describe what you want in natural language:
  39. ```
  40. > Analyze the comments on this YouTube video for my brand ShieldVPN:
  41. https://www.youtube.com/watch?v=VIDEO_ID
  42. > Generate a professional reply to this comment:
  43. "Your product crashed 3 times today, worst purchase ever"
  44. > Give me a KOL ROI report for this sponsored video:
  45. https://www.youtube.com/watch?v=VIDEO_ID
  46. Brand: Notion, Collaboration type: sponsored
  47. > Extract brand keywords from https://notion.so
  48. > Scan this video for brand safety threats:
  49. https://www.youtube.com/watch?v=VIDEO_ID
  50. Brand: Notion
  51. ```
  52. ## How the Scoring Works
  53. Our **5-layer brand relevance model** goes beyond simple keyword matching:
  54. ```
  55. Final Score = 50% Semantic Score ← Does the comment text relate to the brand?
  56. + 15% Video Relevance ← Is the video itself about the brand?
  57. + 15% Entity Match ← Is the brand name explicitly mentioned?
  58. + 10% Intent Score ← Is this a purchase intent, question, or risk?
  59. + 10% Context Score ← How deep is the brand integration in the video?
  60. ```
  61. Different collaboration types have different relevance thresholds:
  62. | Type | Threshold | Why |
  63. |------|-----------|-----|
  64. | Dedicated (full review) | 50% | Most comments on a dedicated video are relevant |
  65. | Sponsored (segment) | 60% | Filter off-topic discussion |
  66. | Mention (brief) | 85% | Only clearly brand-related comments pass |
  67. | Placement (background) | 85% | Product is barely discussed |
  68. [Read the full scoring model documentation →](docs/scoring-model.md)
  69. ## Output Formats
  70. All analysis skills support multiple output formats:
  71. - **Markdown** (default) — Human-readable reports
  72. - **CSV** — Import into spreadsheets or databases
  73. - **JSON** — Integrate into your own tools and pipelines
  74. [See example outputs →](docs/examples/)
  75. ## Multi-Language Support
  76. All skills support comments in any language:
  77. - English, Chinese (中文), Japanese (日本語), Korean (한국어)
  78. - Hindi (हिन्दी), Arabic (العربية), Spanish, Portuguese
  79. - And 20+ more languages
  80. Output language automatically matches your preference or the comment's language.
  81. ## Skill Details
  82. ### YouTube Comment Analyzer
  83. The flagship skill. Fetches comments from any YouTube video and classifies each one into 7 business categories:
  84. | Category | What it means | Priority |
  85. |----------|--------------|----------|
  86. | `purchase_intent` | User wants to buy | Medium |
  87. | `product_question` | Questions about features, pricing, bugs | Medium |
  88. | `brand_risk` | False claims, scams, privacy concerns | **High** |
  89. | `competitor_mention` | Compares with competitors | Medium |
  90. | `positive_feedback` | Praises the product | Low |
  91. | `negative_feedback` | Complaints, disappointment | Medium |
  92. | `irrelevant` | Not about the brand | — (filtered) |
  93. ### Smart Reply Generator
  94. Generates 3 reply options per comment with strategy explanations:
  95. - **Professional** — Clear, direct, respectful
  96. - **Friendly** — Warm, approachable, personable
  97. - **Casual** — Conversational, like texting a friend
  98. ### KOL ROI Report
  99. Full collaboration analysis including:
  100. - Engagement quality score
  101. - Purchase signal strength
  102. - Brand risk level
  103. - Audience-brand alignment
  104. - Actionable recommendations
  105. ### Brand Keyword Extractor
  106. Extracts 5 categories of keywords:
  107. - Primary keywords (high-value, core brand terms)
  108. - Secondary keywords (related, broader reach)
  109. - Long-tail phrases (with search intent classification)
  110. - Competitor keywords
  111. - Negative keywords (terms to exclude)
  112. ### Comment Threat Scanner
  113. Focused brand safety scan with 3 severity tiers:
  114. - **HIGH** — Scams, impersonation, false claims, doxxing
  115. - **MEDIUM** — KOL misrepresentation, privacy concerns, competitor attacks
  116. - **LOW** — Negative experiences, general toxicity, spam
  117. ## Cost Estimate
  118. All skills use `claude-haiku-4-5-20251001` for cost efficiency:
  119. | Skill | Typical API Cost | Comments Analyzed |
  120. |-------|-----------------|-------------------|
  121. | Comment Analyzer | ~$0.02-0.05 | 100 comments |
  122. | Smart Reply | ~$0.001 | 1 comment |
  123. | KOL ROI Report | ~$0.03-0.08 | 100 comments + summary |
  124. | Keyword Extractor | ~$0.002 | 1 website |
  125. | Threat Scanner | ~$0.02-0.05 | 100 comments |
  126. ## Want Continuous Monitoring?
  127. These skills are great for one-time analysis. For **ongoing, automated** YouTube comment monitoring:
  128. **[ReplyCue AI](https://replycueai.com)** provides:
  129. - Real-time comment monitoring across multiple videos
  130. - Automatic AI analysis as new comments arrive
  131. - Team collaboration with role-based access
  132. - One-click AI reply generation
  133. - Dashboard with business intelligence
  134. - Alert rules for high-priority comments
  135. - Multi-video KOL performance comparison
  136. [Try ReplyCue AI →](https://replycueai.com)
  137. ## License
  138. MIT License — use these skills freely in your projects.
  139. ## Contributing
  140. Found a bug or have an improvement? PRs welcome!
  141. 1. Fork this repo
  142. 2. Create your feature branch (`git checkout -b feature/improvement`)
  143. 3. Commit your changes
  144. 4. Push and open a PR

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.