agentleFS
nikhilbhansalinikhilbhansali/youtube-data-skills/skills/youtube-comment-miner/SKILL.md

youtube-comment-miner

Skill4 starsChanged 3 months agoMITView on GitHub

Mine YouTube comments for content ideas, audience questions, pain points, and monetization signals using YouTube Data API v3. Analyze comments from specific videos, top videos of a channel, or search results for a topic. Use when users want to (1) Find what their audience is asking for, (2) Mine content ideas from comments, (3) Discover audience pain points, (4) Find FAQ patterns in comments, (5) Detect monetization signals, (6) Understand audience language and sentiment. Requires user's YouTube Data API v3 key.

  • Reads credentials

The file

We split this skill into the rules it gives an agent

3 rules. About 0 are common, 2 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈2

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Check the user's Claude memory for a YouTube Data API v3 key.

    line 34
  2. Very specific:

    Run scripts/mine_comments.py — resolve the path relative to this skill's own directory:

    line 39
  3. Distinctive:

    Write to the path the script printed:

    line 72

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

YouTube Comment Miner

Mine YouTube comments to extract content ideas, audience questions, pain points, and monetization signals.

Usage

/youtube-comment-miner https://youtube.com/watch?v=VIDEO_ID
/youtube-comment-miner @ChannelHandle --top 5
/youtube-comment-miner --topic "meditation for beginners" --top 10
/youtube-comment-miner VIDEO_ID1 VIDEO_ID2 VIDEO_ID3

Instructions

Step 1: Parse Arguments

Input mode (one of):

  • Video URL(s) or ID(s): specific videos to mine
  • Channel (@handle, URL, or ID) + --top N: mine the channel's top N videos by views (default: 5)
  • Topic (--topic "keyword") + --top N: search for videos on the topic, mine the top N (default: 10)

Also:

  • --max-comments N (optional): max comments per video (default: 100, max: 500)
  • --scan-limit N (optional, channel mode): cap how many uploads get scanned (default: whole channel)

Step 2: Get the API Key

Check the user's Claude memory for a YouTube Data API v3 key. If not found, ask:

"I need a YouTube Data API v3 key to mine comments. You can get one from the Google Cloud Console. Please paste your key."

Step 3: Run the Bundled Script

Run scripts/mine_comments.py — resolve the path relative to this skill's own directory:

# Specific videos
YT_API_KEY=API_KEY python3 <skill-dir>/scripts/mine_comments.py --videos VIDEO_ID1 VIDEO_ID2 --max-comments 100

# Channel top videos
YT_API_KEY=API_KEY python3 <skill-dir>/scripts/mine_comments.py --channel "@handle" --top 5 --max-comments 100

# Topic search
YT_API_KEY=API_KEY python3 <skill-dir>/scripts/mine_comments.py --topic "topic keyword" --top 10 --max-comments 100

Dependency: pip3 install google-api-python-client.

The script fetches relevance-ordered top-level comments, tags each one into categories (question, content_request, pain_point, praise, criticism, suggestion, monetization_signal, personal_story), flags "gold nuggets" (5+ likes on a question or request), and builds an audience-language word frequency list.

Channel mode note: the script finds top videos by paging the uploads playlist and ranking by actual view count, which is both cheaper and more accurate than search.list(order=viewCount) — that endpoint only ever sees a truncated slice of a channel.

Step 4: Read the Data

reports/data/comment-mine-<slug>-<YYYY-MM-DD>.json

Step 5: Write the Report

Write to the path the script printed:

reports/comment-mine-<slug>-<YYYY-MM-DD>.md
# Comment Mining Report: [Source]
*Analyzed [date] | [N] comments across [N] videos*

## Executive Summary
- 3-4 bullet points: Key findings from comment analysis
- What the audience wants, worries about, and loves

## Videos Analyzed
| # | Title | Views | Comments Mined |
|---|-------|-------|----------------|

## Content Requests (What Your Audience Wants)
Top requests ranked by likes. Include exact quotes.
- "Can you make a video about..." patterns
- Specific topics requested multiple times

## Frequently Asked Questions
Questions ranked by frequency/likes.
Group similar questions together.
These are potential video topics.

## Pain Points & Struggles
What viewers are struggling with.
Each pain point = potential video solving that problem.

## Audience Language Patterns
- Words and phrases viewers use repeatedly
- This is the language to mirror in titles, descriptions, thumbnails
- Common vocabulary table

## Praise Patterns (What Works)
What viewers love -- tells you what to do MORE of.

## Criticism Patterns (What to Fix)
What viewers complain about -- tells you what to avoid/improve.

## Gold Nugget Comments
High-engagement comments with questions or requests.
Each one is a validated content idea.

## Monetization Signals
Comments asking about courses, products, tools, etc.
Revenue diversification opportunities.

## Suggestions & Tips from Viewers
Viewer-suggested improvements and ideas.

## Actionable Content Ideas
5-8 specific video ideas derived from comment data:
- Idea title
- Source (which comments inspired it)
- Why it would work
- Priority (based on frequency/engagement)

## Quota Usage
| Operation | Calls | Units |
|-----------|-------|-------|
Use the `quota_used.breakdown` block from the JSON.

Step 6: Report Completion

Tell the user the report path, total comments mined, the top finding, the number of content ideas generated, and the quota consumed.

Quota Estimate

commentThreads.list costs 1 unit per 100 comments.

ModeOperationsTotal
Direct videos (5 videos, 100 comments each)1 videos.list + 5 commentThreads.list~6
Channel (500-video channel, top 5)1 channels.list + ~10 playlistItems.list + ~10 videos.list + 5 commentThreads.list~26
Topic (top 10)1 search.list + 1 videos.list + 10 commentThreads.list~111

Channel mode previously cost ~107 units because it used search.list; the uploads-playlist route cuts that to ~26 and scans the whole channel instead of a partial slice. Daily default allowance is 10,000 units, resetting at midnight Pacific Time.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.