Rank content angles by engagement data, competition level, and platform fit. Data-driven angle selection instead of guesswork. Use this skill when the user has a keyword or product and needs to decide WHAT to create, which angle to take, which format to use, or which platform to target. Triggers on: "what angle should I use", "rank content ideas for [keyword]", "best angle for [product]", "which content idea will perform best", "help me pick an angle", "what should I write about", "content angle for [topic]", "rank my content ideas", "which approach will get the most views", "data-driven content planning", "angle ranker", "content scoring", "which hook should I use", "compare these content ideas", "prioritize my content angles", "what video should I make".
Installer avec Codex ou Claude Copiez ce prompt, collez-le dans Codex, Claude ou un autre assistant, puis laissez-le vérifier la page du skill et l'installer pour vous.
Une commande directe contourne le prompt de vérification. Examinez la source avant de l'exécuter.
Rank content angles by engagement data, competition level, and platform fit. Data-driven angle selection instead of guesswork. Use this skill when the user has a keyword or product and needs to decide WHAT to create, which angle to take, which format to use, or which platform to target. Triggers on: "what angle should I use", "rank content ideas for [keyword]", "best angle for [product]", "which content idea will perform best", "help me pick an angle", "what should I write about", "content angle for [topic]", "rank my content ideas", "which approach will get the most views", "data-driven content planning", "angle ranker", "content scoring", "which hook should I use", "compare these content ideas", "prioritize my content angles", "what video should I make".
You have a keyword. You know the niche. But what specific content should you create?
Which angle, format, and hook will actually perform? This skill answers that question
with data — not gut feeling.
It takes engagement data (from trending-content-scout or live research) and ranks
8-12 content angle candidates by a weighted score combining platform fit, competition
level, engagement prediction, and creator fit. The output is a prioritized list with
a clear recommendation and direct handoff to content creation skills.
Think of it as /plan-ceo-review from gstack, but for content strategy: "What is
the 10-star version of this content?" — except the answer is backed by engagement data.
Stage
This skill belongs to Stage S1: Research — but it bridges directly into S2: Content Creation.
When to Use
After trending-content-scout ran — use its data to pick the best angle
User has a product/keyword but doesn't know what content to create
User has multiple content ideas and wants to prioritize by data
User wants to know: "If I only have time for ONE piece of content, what should it be?"
Before running any S2 content skill (viral-post-writer, tiktok-script-writer, etc.)
Input Schema
keyword:string# (required if no scout_data) "AI video tools"product:object# (optional) Affiliate product being promotedname:string# "HeyGen"description:string# What it doesurl:string# Product URL or affiliate linkreward_value:string# Commission info — never shown in contentplatform:string# (required) Target platform for content creation# "youtube" | "tiktok" | "linkedin" | "x" | "reddit" | "blog"
# (optional) User's own angle ideas to include in ranking
Auto-detection: If trending-content-scout ran earlier in the conversation,
its output is automatically used as the data foundation. No need to pass it explicitly.
Workflow
Step 1: Gather Engagement Data
If trending-content-scout output exists in context:
Use pattern_analysis (winning formats, hooks, engagement benchmarks)
Use content_gaps as angle candidates
Use top_content for competition assessment
Skip to Step 2
If no scout data:
Run a quick scout internally:
web_search "[keyword] site:youtube.com" → top 10 videos, note formats and view counts
web_search "[keyword] site:tiktok.com" OR web_search "[keyword] tiktok viral" → top TikTok content
web_search "[keyword] site:reddit.com top" → top Reddit discussions
web_search "[keyword] [platform] best performing" → meta-analysis of what works
Extract: dominant formats, popular hooks, view count ranges, gaps
This takes 30-60 seconds and provides enough signal for angle scoring.
Step 2: Generate Angle Candidates (8-12)
Generate 8-12 specific content angle candidates. Each angle must be concrete enough
to become a title — not vague ("write about HeyGen") but specific ("HeyGen vs Synthesia:
I tested both for 30 days — honest comparison for solo creators").
Sources for angles:
Gap-based angles (from scout data or web_search):
Content gaps: topics nobody has covered well
Format gaps: popular topic but missing in a specific format (e.g., comparison exists on YouTube but not TikTok)
Audience gaps: existing content targets general audience, specific audience underserved
Recency gaps: existing content is outdated, fresh version needed
Pattern-based angles (from winning formats):
Take the winning format and apply it to the keyword
Combine the best hook type with the topic
Replicate the structure of the highest-engagement content with a fresh perspective
Contrarian angles:
If all content is positive → honest cons angle
If all content targets beginners → advanced user angle
If all content is listicles → deep single-product dive
User-provided angles (from custom_angles):
Include any angles the user suggested
Score them alongside generated candidates — no bias
For each angle, define:
Angle:title:string# Specific, could be an actual content titleangle:string# Brief description of the angleformat:string# "comparison" | "review" | "tutorial" | "listicle" | "demo" | "story" | "reaction" | "explainer"hook:string# The actual hook/opening linehook_type:string# "question" | "shock" | "bold_claim" | "demo_first" | "relatable" | "contrarian"source:string# "gap" | "pattern" | "contrarian" | "user_provided"
Step 3: Score Each Angle
Score every angle on 4 dimensions (1-10 each), then calculate a weighted total:
How well does this format/hook work on the target platform?
Format
YouTube
TikTok
LinkedIn
X
Reddit
Blog
comparison
9
8
7
5
8
9
review
8
6
5
4
9
9
tutorial
9
7
6
3
7
10
listicle
7
8
9
8
6
8
demo
8
10
5
4
3
5
story
6
9
10
8
7
7
reaction
7
10
4
6
5
3
explainer
8
5
8
6
8
9
Adjust based on actual scout data if available (if comparisons outperform on a platform
where they usually don't, use the real data instead of the default table).
Dimension 2: Competition Level (weight: 30%)
How many similar content pieces already exist? Higher score = LESS competition.
IF scout data available:
Count how many top_content pieces match this angle's format + similar topic
10 = zero similar content found (blue ocean)
7-9 = 1-3 similar pieces (low competition)
4-6 = 4-10 similar pieces (moderate competition)
1-3 = 10+ similar pieces (saturated)
IF no scout data:
web_search for the exact angle title → count results
Fewer results with exact match = higher score
Dimension 3: Engagement Prediction (weight: 30%)
How likely is this angle to get high engagement based on data?
IF scout data available:
Look at engagement scores of similar formats and hooks in top_content
If this angle's format has avg_engagement > median → higher score
If this angle's hook_type has avg_engagement > median → higher score
Combine: angle uses top format + top hook → 9-10
Angle uses average format + average hook → 5-6
Angle uses underperforming format → 3-4
IF no scout data:
Use platform defaults and general engagement patterns
Comparisons generally outperform reviews → 8 vs 6
Bold claim hooks generally outperform questions → 8 vs 6
Dimension 4: Creator Fit (weight: 15%)
How well does this angle match the creator's strengths?
IF creator_strengths provided:
"storytelling" → story format, relatable hooks → high fit
"technical" → tutorial format, demo hooks → high fit
"humor" → reaction format, relatable hooks → high fit
"authority" → review format, bold claim hooks → high fit
"visual" → demo format, demo_first hooks → high fit
"data" → comparison format, explainer → high fit
"personal_experience" → story format, reaction → high fit
Match count: 2+ matches → 9-10, 1 match → 6-7, 0 matches → 4-5
IF no creator_strengths:
Default all angles to 7 (neutral)
**Alternative paths:**
- `tiktok-script-writer` — for short-form video version of Angle #1
- `content-pillar-atomizer` — create a blog post, then atomize across all platforms
- `comparison-post-writer` — if the winning angle is a comparison
Error Handling
No scout data and no keyword: Ask user: "What topic or product are you creating content for? And which platform?"
Only 1 platform specified + limited data: Generate angles anyway using platform-specific defaults. Note: "Limited data available. Scores are based on general platform patterns. Run trending-content-scout first for data-backed scoring."
All angles score similarly (within 0.5 points): Spread them out by double-weighting the most differentiating dimension. Present as: "These angles are closely matched. The tiebreaker is [competition/creator fit/etc.]."
User's custom angles score low: Still include them but be honest: "Your angle '[X]' scored [X.X/10] — competition is high and the format doesn't match platform trends. Consider the #1 angle instead, or combine your angle with a [winning format]."
Time budget too short for any good angle: Recommend the easiest angle regardless of score, and flag: "With [30 min], your best option is [quick angle]. For higher impact, allocate [2 hours] for [best angle]."
Examples
Example 1:
User: "I want to promote HeyGen on TikTok. What angle should I use?"
→ keyword: "HeyGen", platform: "tiktok"
→ Quick scout: web_search "HeyGen tiktok" → mostly demo-first content, 30-45s
→ Generate 10 angles: "HeyGen vs Synthesia comparison", "I made a $2000 video for free with HeyGen",
"POV: your boss asks for a video and you use AI", "HeyGen for real estate agents", etc.
→ Top angle: "I replaced a $2000 video production with HeyGen" — Score: 8.7
(bold_claim hook, demo format, low competition on TikTok, high engagement predicted)
→ Next: tiktok-script-writer with hook_style: bold_claim, duration: 45s
Example 2:
User: "Rank these content ideas for my YouTube channel about email marketing:
ConvertKit vs Mailchimp comparison
How I grew my list to 10K subscribers
Top 5 email marketing mistakes"
→ platform: "youtube", custom_angles provided
→ Scout YouTube for email marketing content
→ Score all 3 + generate 5 additional angles
→ Result: "How I grew to 10K" scores highest (story format + bold_claim = 8.5)
because competition for comparisons is saturated (score: 6.2) and listicles are average (7.1)
→ The user's story angle wins — with a suggested hook: "I went from 0 to 10K subscribers in 6 months. Here's what nobody tells you."
Example 3:
User: "I'm good at storytelling and humor. What should I create about AI writing tools on LinkedIn?"
→ creator_strengths: ["storytelling", "humor"], platform: "linkedin"
→ Generate angles weighted toward story format
→ Top: "I let AI write my LinkedIn posts for a week. My boss noticed." — Score: 9.1
(story format, relatable hook, low competition on LinkedIn for this angle, perfect creator fit)
Feedback & Issue Reporting
When this skill produces unexpected, incomplete, or incorrect output, generate a
skill_feedback block (see shared/references/feedback-protocol.md for full schema).
Skill-specific failure modes:
All angles score within 0.5 points: Scoring not differentiated enough. Report as wrong_output with the scores.
Top angle is generic: "Write a review of X" instead of a specific, titled angle. Report as data_quality.
Downstream skill can't use recommended_skill_params: Schema mismatch. Report as chain_break.
Auto-detect triggers:
Score range (max - min) < 1.0 across all angles
<8 angles generated
recommended_skill_params missing required fields for the suggested next skill
niche-opportunity-finder (S1) — niche context for angle generation
competitor-spy (S1) — competitor gaps to exploit
performance-report (S6) — historical angle performance data
Feedback Loop
S6 performance-report shows which angles actually performed → update scoring weights and format preferences for next run → content strategy improves with every cycle
