🎬 AI Storyboard + Comic Prompts Pack — 20 Prompts for Filmmakers, Creators, Authors

Your Storyboard/Sequence Session — Calmwork Founder Arc (6-Panel Instagram)

Format + Convention Check

Instagram carousel storytelling is a strong format choice for this emotional-arc content. 6 panels = good length (3-4 feels thin, 8-10 feels long for Instagram completion rates).

1:1 square is correct for Instagram feed. Reader swipes L→R, so panels must read in that order.

Panel count vs character consistency: 6 panels with ONE character = well within gpt-image-2's reliability zone. Character drift unlikely if prompted well.

Emotional arc (exhausted → contemplative → grounded → ships) is classic 'transformation story' structure that Instagram readers know. Reads well in 6 panels.

Caveat: Single-character emotional-arc stories are where AI can drift on subtle emotional expressions. Panel 3 ('oh.' moment) is the hardest to render — we'll address.

Character Lock Strategy

Primary lock at start of prompt:

'Character: Jenny, founder mid-30s, shoulder-length brown hair (styled loose, slightly messy), round black-framed glasses, navy hoodie. Slim build. Same face, same hair, same glasses, same hoodie across ALL 6 panels.'

Secondary lock at end of prompt:

'Maintain Jenny's consistent appearance (shoulder-length brown hair, round black-framed glasses, navy hoodie) across all 6 panels. Same face structure, same hair length. Only her expression + energy change between panels.'

Setting lock:

'Same home office setting in panels 1-3 (cluttered, late-night mood) and cleaner version in panels 4-6 (same space, tidied). Same desk, same laptop, same window placement, same plants — only lighting + clutter changes.'

Panel-by-Panel Breakdown

Panel 1: Late-Night Grind (wide-ish shot, 11:47 PM)

• Jenny hunched at her desk, laptop glow on face, visible fatigue
• Room dark except for laptop screen + small desk lamp
• Coffee mug with cold remnants, papers scattered
• Caption overlay: '11:47 PM' (small, bottom-left corner)
• Emotion: exhausted, depleted

Panel 2: Frozen Moment (medium shot, no text)

• Jenny staring at laptop screen blankly, hand stopped mid-type
• Same lighting as Panel 1 but more stillness
• Face partially visible in screen reflection
• No text
• Emotion: empty, frustrated

Panel 3: The Realization (close-up, caption 'oh.')

• Close-up of Jenny's face — eyes widening slightly, hand coming away from keyboard
• Subtle shift in expression — recognition
• Same lighting but FACE is the subject, not the screen
• Caption: 'oh.' (small, bottom-center)
• Emotion: realization, quiet

Panel 4: Morning Reset (medium shot, 7:15 AM)

• Jenny at same desk but morning light, warm, cleaner
• Wearing same hoodie but fresh face, coffee mug steaming
• Plant visible near window, morning sun
• Desk tidier — laptop closed, notebook open instead
• Caption: '7:15 AM' (small, bottom-left)
• Emotion: calm, present, contemplative

Panel 5: Flow State (medium-ish shot, no text)

• Jenny at same desk, laptop now open, focused work posture
• Different from Panel 1 — engaged not grinding, posture upright not hunched
• Natural morning light, coffee still steaming
• Same space but feels organized, breathing
• Emotion: flow, focused, calm energy

Panel 6: Ship (close-ish shot with environment, 'Ship.' + calmwork.co)

• Jenny mid-action pressing a button on laptop (finger on enter key or trackpad)
• Small visible smile, satisfied
• Browser tab or terminal window visible on screen showing action completed
• Caption: 'Ship.' (centered, readable)
• Small brand text 'calmwork.co' (very small, bottom corner)
• Emotion: accomplished, quiet satisfaction, grounded

The Prompt (paste into ChatGPT Plus with gpt-image-2)

Create a 6-panel Instagram carousel storytelling sequence, 1:1 square per panel.

CHARACTER LOCK (must be consistent across all 6 panels):
Jenny, founder mid-30s, shoulder-length loose brown hair (slightly messy), round black-framed glasses, navy blue hoodie, slim build. SAME face structure, SAME hair, SAME glasses, SAME hoodie in every panel. Only her expression and energy change between panels.

SETTING LOCK:
Home office with desk, laptop, coffee mug, plant by window. Panels 1-3: late-night, dark, desk cluttered. Panels 4-6: morning light, same space but tidied. Same desk, same laptop model, same window placement, same plant across all panels.

STYLE: Soft cinematic realism, warm natural lighting, slightly desaturated. Indie short-film aesthetic. NOT cartoony, NOT corporate, NOT overly polished.

COLOR PALETTE: Navy + sage green subtly present throughout (the 'Calmwork' brand feel). Warm amber light in morning panels. Cooler blue tones in night panels.

PANEL 1 (wide-ish shot): Jenny hunched at desk, laptop glow on tired face, late night. Coffee mug cold, papers scattered. Caption overlay bottom-left: '11:47 PM'

PANEL 2 (medium shot): Jenny frozen, hand stopped mid-type, staring blankly at laptop screen. Same dark lighting. NO text.

PANEL 3 (close-up): Jenny's face in close-up, eyes widening slightly in recognition, hand coming away from keyboard. Caption bottom-center: 'oh.'

PANEL 4 (medium shot): Same desk, morning light now, warm amber. Jenny fresh-faced but still in navy hoodie, steaming coffee, laptop closed, notebook open. Desk tidier. Caption bottom-left: '7:15 AM'

PANEL 5 (medium shot): Jenny at desk working with focused calm energy, upright posture (contrast with Panel 1's hunch). Laptop open, working. Natural morning light. NO text.

PANEL 6 (close-ish): Jenny pressing enter/trackpad button, small visible smile, satisfaction. Browser/terminal on screen shows action completed. Caption centered: 'Ship.' Small brand text bottom corner: 'calmwork.co'

Maintain Jenny's consistent appearance (shoulder-length brown hair, round black-framed glasses, navy hoodie) across all 6 panels. Same face structure across panels. Only expression + energy + lighting change.

Thinking Mode enabled. Generate all 6 panels in sequence.

Expected Output + Quality Control

gpt-image-2 will likely produce:

• Strong character consistency across 6 panels (within reliability zone)
• Clear emotional progression (exhausted → contemplative → calm)
• Lighting shift correctly showing night → morning transition
• Text rendering: '11:47 PM', '7:15 AM', 'Ship.', 'calmwork.co' should render correctly (these are short captions, within gpt-image-2 text strength)
• 'oh.' may render with period or without — verify

Verify on first generation:

• Jenny's hair length and glasses consistent across all 6 panels (#1 drift risk)
• Panel 3 emotional expression actually communicates 'recognition' (subtlest emotion to render)
• Morning panels 4-6 feel like same space as night panels 1-3 (not a completely different room)
• Caption text spellings correct

Post-Generation Workflow (Figma)

1. Composite final captions in Figma. Even if AI rendered them, redo in your brand font for crispness and exact positioning.

2. Verify Instagram preview. Upload as carousel draft to Instagram, scroll through. Does the arc read? Does Panel 3 land the emotional turn? Are transitions smooth?

3. Consider sound/accessibility. Instagram carousels benefit from: alt text per image (accessibility), first-frame hook that makes people swipe.

4. Add caption story below carousel. The six images tell the story visually; your Instagram caption expands in text: 'I hit my wall at 11:47 PM last Tuesday. Then I realized the real problem was how I was working, not what I was working on. Here's what changed. [swipe →]'

Common Failure Modes for This Sequence

Character drift (most likely):

• Panel 4+ Jenny may have slightly different face, longer hair, or different glasses
• Fix: re-prompt with ONLY panels 4-6, explicitly lock character from panel 1 as reference

Panel 3 emotion-unclear:

• 'oh.' moment is subtle recognition — AI may render confusion or surprise instead
• Fix: re-prompt Panel 3 with 'subtle realization, eyes softly widening, mouth relaxed — not shock, not surprise, quiet recognition'

Lighting discontinuity:

• Morning panels may look like a completely different room
• Fix: re-prompt Panel 4-6 referencing 'same desk, same plant, same window — only the lighting and clutter change'

Caption text errors:

• 'Ship.' may render as 'Ship' (no period) — minor, fix in Figma
• 'calmwork.co' brand text may drift to 'calmworkco' or similar — acceptable, composite real text in post

Never iterate on same generation more than 2x (noise bug) — if quality degrades, start fresh session.

Ethical Considerations

• No real-person likeness issues — Jenny is a fictional founder
• 'Calmwork' brand is your own — no trademark concerns
• Content is honest storytelling (burnout → recovery), not misleading
• No deceptive 'before/after' claims — this is clearly an illustrative sequence, not a product testimonial
• If this goes in paid advertising, some jurisdictions may require 'illustrative' disclosure since the character is AI-generated

Key Takeaways

• 6 panels with ONE character is within gpt-image-2 reliability zone — character drift risk is low if prompted with dual-lock (start + end).
• Lock character traits twice in prompt — once at start, once at end. Single declaration drifts.
• Vary shot types across panels — wide / medium / close-up / medium / medium / close-ish. Not all the same.
• Panel 3 is the hardest — subtle recognition emotion. Budget re-iteration for this panel specifically.
• Captions '11:47 PM' / '7:15 AM' / 'oh.' / 'Ship.' are within gpt-image-2 text strength — short, clear.
• Morning/night lighting shift is the visual storytelling device — make it explicit in prompt.
• Post-generation: redo captions in Figma for crispness, add accessibility alt-text, write Instagram caption that expands the story.
• Publishing: first-frame hook matters — Panel 1 should make people want to swipe.