content-pillar-atomizer

What this skill does

# Content Pillar Atomizer

Take 1 blog post or article and generate 15-30 platform-native micro-content pieces. This is NOT reformatting — it's re-contextualizing each piece for the platform's culture, format, and audience expectations. A LinkedIn post reads nothing like a Reddit comment, even if they carry the same insight.

## Stage

S2: Content Creation — This IS content creation, just at 10x scale. One piece of deep work becomes a month of social content.

## When to Use

- User has a blog post, article, or long-form content and wants to maximize its reach
- User asks to "repurpose" or "atomize" content
- User says "turn this into social posts", "content multiplication", "pillar content"
- After `affiliate-blog-builder` (S3) produces an article — atomize it into social
- User wants to maintain consistent content output without creating from scratch daily

## Input Schema

```yaml
pillar_content: string        # REQUIRED — the full blog post/article text, or URL to fetch

platforms: string[]           # OPTIONAL — target platforms
                              # Options: "twitter", "linkedin", "reddit", "tiktok", "email", "threads"
                              # Default: ["twitter", "linkedin", "reddit"]

product: object               # OPTIONAL — affiliate product being promoted
  name: string
  url: string
  reward_value: string

mode: string                  # OPTIONAL — "quality" | "volume"
                              # Default: "quality"

tone: string                  # OPTIONAL — "professional" | "casual" | "edgy" | "educational"
                              # Default: inferred from pillar content
```

**Chaining from S3**: If `affiliate-blog-builder` was run, use its output article as `pillar_content`.

**Chaining from S1 monopoly-niche-finder**: Use `monopoly_niche` positioning to angle all micro-content.

## Workflow

### Step 1: Analyze Pillar Content

1. If URL provided, use `web_fetch` to retrieve content
2. Extract: key insights (5-8), data points, quotes, frameworks, stories, opinions
3. Identify the "atomic units" — self-contained ideas that work independently
4. Note the product/affiliate angle (if present)

### Step 1.5: Check Platform Performance for This Topic (data-driven)

Before atomizing equally across all platforms, understand which platforms are hot for this topic:

**If `trending-content-scout` ran:**
- Use platform-level engagement data from `pattern_analysis`
- Check `engagement_benchmark.platform_averages` — which platform has highest engagement for this keyword?
- Prioritize platforms where this topic has highest engagement
- Adjust platform allocation accordingly (see below)

**Quick check (no scout data):**
- `web_search "[topic] youtube vs tiktok vs linkedin"` → which platform dominates discussion?
- Check: is this topic more visual (→ TikTok/YouTube heavy) or professional (→ LinkedIn heavy)?
- Look for: which platform shows up most in search results for this topic?

**Apply to atomization allocation:**
- Default: equal split across platforms
- Data-driven: proportional to engagement potential
  - If TikTok engagement is 5x LinkedIn for this topic → generate 5 TikTok scripts, 1 LinkedIn post
  - If Reddit has high engagement → don't skip Reddit (often ignored by affiliates = opportunity)
  - If YouTube dominates → consider atomizing into YouTube Shorts scripts instead of just TikTok

**Platform allocation example:**
```
Default (no data):    Twitter: 5 | LinkedIn: 3 | Reddit: 3 | TikTok: 3 | Email: 2
Data-driven (TikTok hot): Twitter: 3 | LinkedIn: 1 | Reddit: 2 | TikTok: 6 | Email: 2
Data-driven (LinkedIn hot): Twitter: 3 | LinkedIn: 5 | Reddit: 2 | TikTok: 2 | Email: 2
```

### Step 2: Platform Mapping

Read `shared/references/platform-rules.md` for platform-specific rules.

For each platform, map the culture:

| Platform | Format | Tone | Length | CTA Style |
|---|---|---|---|---|
| Twitter/X | Thread or single tweet | Punchy, opinionated | 280 chars or 5-10 tweet thread | Last tweet |
| LinkedIn | Story or insight post | Professional, first-person | 1300 chars | Soft CTA in comments |
| Reddit | Value-first post/comment | Helpful, honest, skeptical-aware | Variable | Disclosure + subtle |
| TikTok | Script with hook | Casual, energetic | 30-60s script | Verbal + bio link |
| Email | Newsletter section | Conversational | 200-400 words | Direct link |
| Threads | Conversational take | Casual, authentic | 500 chars | Bio link |

### Step 3: Generate Micro-Content

For each platform, generate pieces from different atomic units:

- **Twitter**: 3-5 pieces (1 thread, 2-3 standalone tweets, 1 hot take)
- **LinkedIn**: 2-3 pieces (1 story post, 1 insight post, 1 question post)
- **Reddit**: 2-3 pieces (1 detailed post, 1-2 comment-ready responses)
- **TikTok**: 2-3 scripts (1 educational, 1 hot take, 1 tutorial)
- **Email**: 1-2 pieces (newsletter section, dedicated email)
- **Threads**: 2-3 pieces (conversational takes)

Each piece must:
- Stand alone (makes sense without reading the pillar)
- Feel native to the platform (not a copy-paste resize)
- Carry one clear insight or value point
- Include appropriate FTC disclosure for affiliate content

### Step 4: Tag for Tracking

Tag each piece with:
- Source pillar reference
- Platform
- Content type (thread, single, story, script)
- Affiliate product (if applicable)
- Suggested posting time/day

### Step 5: Self-Validation

- [ ] Each piece feels native to its platform (not copy-pasted)
- [ ] Each piece stands alone without needing the pillar
- [ ] FTC disclosure included where affiliate links present
- [ ] No two pieces on the same platform say the same thing
- [ ] Platform rules followed (Reddit skepticism, LinkedIn professionalism, etc.)

## Output Schema

```yaml
output_schema_version: "1.0.0"
atomized_content:
  pillar_title: string
  total_pieces: number
  platforms_covered: string[]

  pieces:
    - platform: string
      type: string              # "thread" | "single" | "story" | "script" | "email" | "comment"
      content: string           # The actual content, ready to post
      insight_source: string    # Which atomic unit from the pillar
      has_affiliate_link: boolean
      suggested_timing: string  # e.g., "Tuesday 9am"
      variant_id: string        # For volume mode A/B tracking

  content_pillars: string[]    # Atomic units extracted (for chaining)

chain_metadata:
  skill_slug: "content-pillar-atomizer"
  stage: "content"
  timestamp: string
  suggested_next:
    - "social-media-scheduler"
    - "email-drip-sequence"
    - "ab-test-generator"
```

## Output Format

```
## Content Atomizer: [Pillar Title]

### Pillar Analysis
- **Atomic units extracted:** X insights
- **Platforms:** [list]
- **Total pieces generated:** XX

---

### Twitter/X (X pieces)

**Thread: [Title]**
🧵 1/ [first tweet]
2/ [second tweet]
...
[last tweet with CTA]

**Standalone Tweet:**
[tweet text]

---

### LinkedIn (X pieces)

**Story Post:**
[full LinkedIn post]

---

### Reddit (X pieces)

**Post: r/[subreddit]**
Title: [title]
[body with disclosure]

---

[Continue for each platform]

### Posting Schedule
| Day | Platform | Piece | Time |
|---|---|---|---|
| Mon | Twitter | Thread | 9am |
| Tue | LinkedIn | Story | 8am |
| Wed | Reddit | Post | 12pm |
```

## Error Handling

- **No pillar content provided**: "Paste your blog post or article, or give me the URL and I'll fetch it."
- **Content too short**: "This is quite short for atomization. I'll extract what I can, but consider writing a longer pillar first with `affiliate-blog-builder`."
- **No affiliate angle**: Generate content without affiliate links. Pure value content builds audience for future promotions.
- **Platform not supported**: "I don't have specific rules for [platform]. I'll format it generically — review before posting."

## Examples

**Example 1:** "Atomize my HeyGen review blog post into social content"
→ Extract 6 key insights, generate 15 pieces across Twitter (thread + 3 tweets), LinkedIn (2