Scenario to TikTok

Generate a production-ready TikTok script from a completed Shell scenario planning exercise, with meme integration and visual direction.

When to Use

• Completed scenario with 4+ narratives exists
• User wants short-form video content from the analysis
• Scenario has a clear thesis and counterintuitive findings

When NOT to Use

• Scenario is incomplete (no narratives yet)
• User wants long-form content (blog, report, documentary) — different format
• No comedic or irreverent angle exists

Required Inputs

Read these scenario files to build the script:

1. worldview_model.md — The user's original beliefs (this is your HOOK)
2. worldview_integration.md — Where beliefs were confirmed/challenged (this is your TWIST)
3. predetermined_elements.md — Hard facts and numbers (this is your EVIDENCE)
4. critical_uncertainties.md — The 2x2 matrix and axes (this is your STRUCTURE)
5. Scenario narratives — Character names and story beats (this is your COLOR)
6. strategy_analysis.md — The meta-strategy (this is your CLOSER)

Script Structure

TikTok scripts from scenarios follow a 5-section arc:

HOOK → COMMON BELIEF → COUNTERINTUITIVE FINDING → THE TWIST → CALLBACK + SEQUEL TEASE

Section Breakdown

Section	Source	Purpose	Duration
HOOK	worldview_model.md	State the question using the user's original framing	5-10s
COMMON BELIEF	predetermined_elements.md	Present what most people assume (and what's right about it)	15-25s
BODY GOES OFFLINE	context_packet.md, predetermined	The visceral details — numbers, timelines, consequences	20-30s
THE TWIST	critical_uncertainties.md	The single variable that changes everything	15-20s
SOCIAL CONSEQUENCES	worldview_integration.md	The human/social angle people don't think about	10-15s
CLOSER	strategy_analysis.md	The meta-strategy as a one-liner. Sequel tease with scenario names.	10-15s

Target total: 2:00-3:00

Writing the Hook

The hook comes from the GAP between what the user originally believed and what the scenario process revealed. Read worldview_model.md and worldview_integration.md together.

Pattern: "Important question. [State the question]. Would you [optimistic extreme] or [pessimistic extreme]?"

Writing the Twist

The twist is ALWAYS the scenario's primary axis — the single variable that matters most. In a good scenario, this is counterintuitive. Frame it as a conspiratorial lean-in: "BUT. If by [term] you mean [expanded definition]..."

Text-on-Screen Moments

Every good TikTok script needs 2-3 text-on-screen moments. These come from the scenario's most quotable findings. Look for:

• Phrases that are funny out of context ("STARVING, BUT WET")
• The thesis condensed to 3-4 words ("WEIRD, BUT SURVIVABLE")
• Specialist one-liners that hit hard

The Sequel Tease

Always end by naming the four scenarios. Scenario names are designed to be memorable and intriguing — they function as sequel bait.

Pattern: "Next I'll do the even more deranged version: [Scenario 1], [Scenario 2], [Scenario 3], and [Scenario 4]."

Meme Integration

After writing the base script, embellish with memes using the meme-rodeo MCP tools.

Sourcing Memes

1. Search meme-rodeo for topic-relevant terms (the scenario subject, key themes)
2. Search for emotional/tonal terms (chaos, conspiracy, confusion, existential)
3. If rate-limited on search, use random_memes or list_recent and filter manually
4. Aim for 12-18 memes total across a 2-3 minute script

Meme Placement Rules

Placement Type	Duration	When to Use
Flash (0.5s)	Subliminal comedy beat	Punctuate a punchline, no caption needed
Hold (1-1.5s)	Let audience read caption	When meme + caption overlay = the joke
Long Hold (2s)	Biggest laugh, most shareable frame	Max 2-3 per script, save for best moments
Callback	Same meme, bookend	Open and close with same meme for structural closure

Meme Mapping

Map each meme to a specific scenario or finding:

> **[MEME INSERT: Name]** — description of meme. Caption overlay: "text"
> ![](image_url)

Every meme insert needs:

• Which line it follows
• Duration (flash/hold/long hold)
• Caption overlay text (if any)
• Why it works for this moment

Meme Selection Criteria

Pick memes that work through JUXTAPOSITION with the script content, not literal illustration:

• A jogger past a burning house + "day 18: I feel amazing" (ironic contrast)
• Batman at a dinner table + "you at thanksgiving now" (absurd reframing)
• Tiny cocktail + "this is 15 calories" (scale visualization)

Avoid memes that just illustrate what's being said. The meme should add a layer.

Output Files

Write two files to the scenario's exports directory:

1. exports/tiktok_script.md — Base script without memes (for reading/editing)
2. exports/tiktok_script_with_memes.md — Full script with meme inserts, timing cues, and production notes

Include a Meme Placement Summary table at the end:

| Timestamp | Meme | Duration | Caption Overlay |
|-----------|------|----------|-----------------|
| Hook | [name] | 0.5s flash | — |
| ... | ... | ... | ... |

And Production Notes covering:

• Recommended total video length
• Flash vs hold guidance
• Which meme placement is most important
• Which frame is most shareable

Common Mistakes

Mistake	Fix
Script sounds like a report	Rewrite in spoken cadence. Short sentences. Fragments are fine.
No visual direction	Add [close-up], [pacing], [sit down], [lean in] cues
Memes are literal illustrations	Pick memes that contrast or reframe, not illustrate
Missing the twist	The twist is the primary scenario axis. Find it in critical_uncertainties.md
Too long	3 minutes max. Cut the least surprising section.
No text-on-screen moments	Find 2-3 quotable phrases from specialist consultations
Generic tone	Match the user's voice from the worldview elicitation. Mirror their humor.
No sequel hook	Always name the four scenarios at the end

Tone Calibration

Read the worldview_model.md elicitation notes for the user's communication style. The script should sound like THEM, not like a narrator. If they're irreverent, be irreverent. If they're dry, be dry. The scenario process is serious; the TikTok should be fun.