YouTube Content Ideator

Generate 10 ranked video ideas for the AI Automation for Business YouTube channel by scanning trends, analyzing competitors, and finding content gaps.

Quick Actions

Need	Action
Full ideation run	-> Run all 3 phases below
Quick trends only	-> Phase 1 only
Competitor check	-> Phase 2 only
Refresh ideas	-> Phase 3 with cached data
Video brief	-> `assets/templates/video-brief.md`
Competitor list	-> `references/competitors.md`
X influencers	-> `references/x-influencers.md`

Cost Estimates

Run Type	Apify Cost	Notes
Full run (all phases)	~$0.10-0.15	1 YouTube scraper run + Claude API tokens
Quick trends only (Phase 1)	$0.00	WebSearch only, no Apify
Competitor check only (Phase 2)	~$0.08	Apify YouTube scraper only
Budget full run (Tier 1 only)	~$0.05	6 channels instead of 15

Content Pillars

Pillar	%	Topics
Tool Reviews & Demos	30%	New AI tools, side-by-side comparisons, "I tried X for 7 days"
Tutorials & How-To	25%	Build automations, integrate tools, step-by-step walkthroughs
Strategy & Business	20%	AI agency models, pricing, client acquisition, scaling
News & Trends	15%	Weekly AI news, product launches, industry shifts
Behind the Scenes	10%	Client projects, revenue reports, workflow reveals

Phase 1: Trending Tools Scan

Goal: Identify top 15 AI trends from the last 7 days ranked by buzz + business relevance.

Step 1A: Web Search (3 parallel calls)

Run these WebSearch queries in parallel:

"new AI tool launch" OR "AI product launch" site:producthunt.com OR site:news.ycombinator.com (last 7 days)
"AI automation tool" OR "AI agent" new launch 2026 (last 7 days)
"best new AI tools" OR "AI tools this week" 2026 (last 7 days)

Step 1B: AI News Search (2 parallel calls)

"AI news this week" OR "AI update" site:theverge.com OR site:techcrunch.com (last 7 days)
"artificial intelligence" announcement OR launch OR release (last 7 days)

Step 1C: X/Twitter Influencer Scan (WebSearch)

Use WebSearch to scan X/Twitter for AI influencer discussions. Run 3 parallel WebSearch calls using handles from references/x-influencers.md:

site:x.com karpathy OR AndrewYNg OR svpino AI tool [current month year]
site:x.com "AI agent" OR "AI tool" trending [current month year]
[top influencer name] AI opinion latest [current year] (pick 3-4 key influencers)

Extract from results:

Tool/product mentions (look for links, @mentions, product names)
Trending topics and recurring themes
Hot takes and debates (potential contrarian content)

Note: WebSearch with site:x.com is the primary method. Apify Twitter scrapers (e.g., apidojo/tweet-scraper) require auth cookies and often return empty results. Only use them if the user has confirmed working auth cookies are configured.

Step 1D: Optional — Research Papers (only if AI research is trending)

If web searches reveal trending research (e.g., new model architectures, benchmarks):

mcp__plugin_bio-research_pubmed__search_articles — search for trending AI/ML papers
mcp__plugin_bio-research_biorxiv__search_preprints — search for AI preprints

Step 1 Output

Compile a ranked list:

#	Trend/Tool	Source	Buzz Score (1-5)	Business Relevance (1-5)	Total
1	[Tool/Trend]	[Web/X/News]	X	X	X
...	...	...	...	...	...
15	...	...	...	...	...

Buzz Score: Social mentions, news coverage, influencer attention Business Relevance: Applicability to AI automation businesses and their clients

Phase 2: Competitor Analysis

Goal: Build a content map of what competitors posted recently and identify gaps.

Step 2A: Fetch Competitor Videos

Use mcp__apify__call-actor with streamers/youtube-scraper to fetch recent videos from ALL channels listed in references/competitors.md in a single run.

Single consolidated run with all 15 channel URLs:

{
  "startUrls": [
    {"url": "https://www.youtube.com/@nateherk"},
    {"url": "https://www.youtube.com/@NateBJones"},
    {"url": "https://www.youtube.com/@briancasel"},
    {"url": "https://www.youtube.com/@simonscrapes"},
    {"url": "https://www.youtube.com/@BenAI92"},
    {"url": "https://www.youtube.com/@leonvanzyl"},
    {"url": "https://www.youtube.com/@dylandavisAI"},
    {"url": "https://www.youtube.com/@rasmic"},
    {"url": "https://www.youtube.com/@TaylorAHaren"},
    {"url": "https://www.youtube.com/@theboringmarketer"},
    {"url": "https://www.youtube.com/@DanielPriestley"},
    {"url": "https://www.youtube.com/@peterdiamandis"},
    {"url": "https://www.youtube.com/@LennysPodcast"},
    {"url": "https://www.youtube.com/@GregIsenberg"},
    {"url": "https://www.youtube.com/@DwarkeshPatel"}
  ],
  "maxResults": 5,
  "sortVideosBy": "NEWEST"
}

Retrieve results with mcp__apify__get-actor-output, using the fields parameter to retrieve only needed columns: title,viewCount,date,channelName,url,likes,duration.

Post-processing: Categorize results by tier based on channel name matching the tier lists in references/competitors.md.

For quick/budget runs: Scrape Tier 1 only (6 channels, ~$0.05) by using only the first 6 URLs.

Step 2B: Build Content Map

For each competitor video, extract:

Title
Topic category (map to content pillars)
Upload date
View count (if available)

Step 2 Output

Competitor Content Map:

Topic	Covered By	Times Covered (7d)	Avg Views	Saturation
[Topic]	[Channels]	X	X	High/Med/Low

Gap Analysis:

Topics trending (Phase 1) but NOT covered by competitors
Topics covered poorly (low views relative to channel size)
Emerging tools no one has reviewed yet

Phase 3: Cross-Reference & Generate Ideas

Goal: Match Phase 1 trends against Phase 2 gaps. Score and rank 10 video ideas.

Scoring Criteria

Each idea scored out of /40:

Criterion	Weight	Description
Trend Momentum	/10	How hot is this trend right now? Rising or peaked?
Gap Score	/10	How underserved is this topic by competitors?
Business Relevance	/10	How relevant to AI automation business audience?
Evergreen Potential	/10	Will this still get views in 6 months?

Idea Categories

Categorize each idea:

Quick Wins (score 30+, easy to produce, ride a trend wave)
Deep Dives (score 25+, longer format, high evergreen potential)
Contrarian Takes (score 20+, counter-narrative to popular opinion)

Output Format

For each of the 10 ideas, use the template from assets/templates/video-brief.md:

Title (3 options: curiosity, direct, contrarian)
Hook (first 5 seconds script)
Trend Driver (which trend from Phase 1)
Gap Analysis (why competitors missed this)
Format (tutorial / review / strategy / reaction / listicle)
Thumbnail Concept (visual description)
Target Keywords (3-5 for SEO)
Score Breakdown (Trend/Gap/Biz/Evergreen = Total)

Final Report

Generate the full report content, then use the pdf skill to create a styled PDF:

Output path: tmp/youtube-ideas-[date].pdf
Layout:
- Title page: "TOP 10 VIDEO IDEAS - [Date] - AI Automation for Business"
- Table of contents
- Phase 1: Trends table with scores
- Phase 2: Competitor content map + saturation analysis + gap analysis
- Phase 3: All 10 ranked ideas with full briefs, grouped by category:

QUICK WINS (Film This Week)
----------------------------
#1. [Title] — Score: XX/40
#2. ...

DEEP DIVES (Schedule for Next 2 Weeks)
----------------------------------------
#3. [Title] — Score: XX/40
#4. ...

CONTRARIAN TAKES (High Risk / High Reward)
-------------------------------------------
#5. [Title] — Score: XX/40
#6. ...

Sources section (all URLs referenced)

Error Handling

Issue	Fallback
Apify YouTube scraper fails	Use `WebSearch` for "[channel name] youtube recent videos"
WebSearch X/Twitter returns few results	Try broader queries without `site:x.com`, or add more influencer names
Too few trends found	Expand date range to 14 days
Competitor channel unavailable	Skip and note in output
PubMed/bioRxiv timeout	Skip research phase entirely
PDF generation fails	Fall back to markdown at `tmp/youtube-ideas-[date].md`

Tips

Run weekly (ideally Monday/Tuesday) for freshest ideas
Compare output against your own recent uploads to avoid repeats
Quick Wins should be filmed within 48 hours of trend peak
Deep Dives can be batched and scheduled
Save unused ideas to a backlog for slow news weeks

youtube-content-ideator