@renoise/video-maker 0.1.0

package/skills/director/references/narrative-pacing.md

@@ -0,0 +1,257 @@
+# Narrative Pacing Guide
+
+A comprehensive reference for structuring long videos (>15s) with cinematic narrative rhythm. The director MUST read this before generating multi-segment video prompts.
+
+## Core Principle
+
+**Every video tells a story, even a product ad.** Multi-segment videos fail when treated as a sequence of independent clips. Instead, design them as one continuous narrative with an emotional arc, energy variation, and intentional transitions.
+
+## Narrative Arc Templates
+
+### 30s — 2 Segments (Hook → Payoff)
+
+```
+Segment 1 (15s)                    Segment 2 (15s)
+├─ Hook [0-5s] ⚡ Energy: 7       ├─ Escalation [0-7s] ⚡ Energy: 8
+├─ Setup [5-10s] ⚡ Energy: 5     ├─ Climax [7-11s] ⚡ Energy: 10
+└─ Build [10-15s] ⚡ Energy: 6    └─ Resolution [11-15s] ⚡ Energy: 4
+```
+
+**Energy curve**: `7 → 5 → 6 │ 8 → 10 → 4`
+**Key**: S1 builds curiosity, S2 delivers and resolves.
+
+### 45s — 3 Segments (Setup → Confrontation → Resolution)
+
+```
+Segment 1 (15s) — ACT I            Segment 2 (15s) — ACT II           Segment 3 (15s) — ACT III
+├─ Hook [0-4s] ⚡ 8                ├─ Complication [0-5s] ⚡ 7        ├─ Climax [0-5s] ⚡ 10
+├─ World [4-10s] ⚡ 4              ├─ Rising [5-10s] ⚡ 8             ├─ Release [5-10s] ⚡ 6
+└─ Inciting [10-15s] ⚡ 6          └─ Tension peak [10-15s] ⚡ 9      └─ Coda [10-15s] ⚡ 3
+```
+
+**Energy curve**: `8 → 4 → 6 │ 7 → 8 → 9 │ 10 → 6 → 3`
+**Key**: Classic three-act. S2 ends on tension, S3 opens with climax.
+
+### 60s — 4 Segments (Hook → Development → Climax → Coda)
+
+```
+Segment 1 (15s) — HOOK             Segment 2 (15s) — DEVELOPMENT
+├─ Attention grab [0-5s] ⚡ 8      ├─ Deepen [0-7s] ⚡ 5
+├─ Context [5-10s] ⚡ 5            ├─ Complicate [7-12s] ⚡ 7
+└─ Promise [10-15s] ⚡ 6           └─ Raise stakes [12-15s] ⚡ 8
+
+Segment 3 (15s) — CLIMAX           Segment 4 (15s) — CODA
+├─ Build-up [0-5s] ⚡ 9            ├─ Aftermath [0-5s] ⚡ 5
+├─ Peak moment [5-10s] ⚡ 10       ├─ Reflection [5-10s] ⚡ 3
+└─ Turning point [10-15s] ⚡ 8     └─ Final image [10-15s] ⚡ 4
+```
+
+**Energy curve**: `8→5→6 │ 5→7→8 │ 9→10→8 │ 5→3→4`
+**Key**: S3 is the emotional peak. S4 gives the audience time to feel.
+
+### 90s-180s — 6-12 Segments (Extended Three-Act)
+
+For long content, follow this proportional structure:
+
+| Act | Duration | Segments | Energy Range | Purpose |
+|-----|----------|----------|-------------|---------|
+| **I — Setup** | 20% | 1-2 | 5-8 | Hook, establish world, introduce conflict |
+| **II-A — Rising** | 30% | 2-4 | 6-9 | Complications, deepening, subplot |
+| **Midpoint** | 10% | 1 | 8-10 | Major revelation or reversal |
+| **II-B — Escalation** | 20% | 2-3 | 7-10 | Stakes raise, momentum builds |
+| **III — Climax+Resolve** | 20% | 2-3 | 10→3 | Peak action, emotional resolution, coda |
+
+**Rule**: Place the energy peak at 70-80% of total duration, never at the very end.
+
+## Energy Curve Rules
+
+### Mandatory Constraints
+
+1. **No flat lines**: Never have 3+ consecutive segments at the same energy level. Audiences disengage from monotony.
+2. **Drop before climax**: Energy must dip at least 2 points before the climax segment. Silence makes the explosion louder.
+3. **Breathing room**: After any ⚡8+ segment, the next segment must open at ⚡6 or lower.
+4. **Distinct ending**: Final segment energy ≠ middle segment energy. End on resolution (low) or surprise (high), not the same plateau.
+5. **Opening hook**: First 3-5 seconds must be ⚡7+. You have 3 seconds to earn the viewer's attention.
+
+### Energy Level Reference
+
+| Level | Feeling | Camera | Pacing |
+|-------|---------|--------|--------|
+| 1-2 | Stillness, meditation | Static, slow drift | Long holds, silence |
+| 3-4 | Calm, reflective | Gentle push/pull | Unhurried, room to breathe |
+| 5-6 | Steady, engaging | Smooth tracking, orbit | Moderate, conversational |
+| 7-8 | Exciting, dynamic | Fast tracking, whip pans | Quick cuts, momentum |
+| 9-10 | Peak intensity | Snap zooms, vertigo, rapid cuts | Maximum density, beat-synced |
+
+## Camera Rhythm Patterns
+
+### Breathing Rhythm (fast → slow → fast)
+The most natural viewing rhythm. Alternate between tight, energetic sequences and open, calm moments.
+```
+[S1] Close-up montage, quick cuts ⚡8
+[S2] Wide establishing shot, slow drift ⚡4
+[S3] Medium tracking, building pace ⚡7
+[S4] Extreme close-up → snap to wide reveal ⚡10
+```
+
+### Tension Ramp (gradual build)
+Start wide and slow, progressively tighten shots and accelerate cuts.
+```
+[S1] Extreme wide, static ⚡3
+[S2] Medium shot, gentle tracking ⚡5
+[S3] Close-up, faster cuts ⚡7
+[S4] Extreme close-up, rapid montage ⚡10
+```
+
+### Pulse Pattern (rhythmic peaks)
+Regular energy spikes, like a heartbeat. Good for music-driven content.
+```
+[S1] Build ⚡6 → peak ⚡9
+[S2] Drop ⚡4 → peak ⚡8
+[S3] Drop ⚡3 → peak ⚡10
+[S4] Sustain ⚡7 → fade ⚡3
+```
+
+## Segment Transition Design
+
+Each transition between segments must be intentionally designed. Never default to a hard cut without purpose.
+
+### 7 Transition Types
+
+| Type | Technique | When to Use | Prompt Keywords |
+|------|-----------|-------------|-----------------|
+| **Action Bridge** | End S1 mid-action → S2 continues the motion | Physical movement, dance, sports | `Continuing the motion from the previous shot:` |
+| **Gaze Lead** | Character looks toward something → S2 reveals it | Mystery, discovery, reaction | `Revealing what they're looking at:` |
+| **Sound Bridge** | S1 fades with S2's ambient sound bleeding in | Scene changes, time shifts | `The sound of [next scene] begins before the cut.` |
+| **Match Cut** | Similar shape/color/motion links two different shots | Thematic connections, metaphor | `Matching the circular shape:` / `The spinning motion carries into:` |
+| **Emotional Shift** | Abrupt mood change (quiet→loud or loud→quiet) | Surprises, twists, contrast | `Suddenly:` / `The silence breaks with:` |
+| **Time Jump** | Visual time indicators (light change, clock, seasons) | Montage, passage of time | `Hours later:` / `As night falls:` |
+| **Spatial Flow** | Camera moves through a door/window/portal into new space | Exploration, journey, reveal | `The camera pushes through the doorway into:` |
+
+### Transition Selection Guide
+
+| From Energy → To Energy | Recommended Transitions |
+|------------------------|------------------------|
+| High → High | Action Bridge, Match Cut |
+| High → Low | Emotional Shift, Time Jump |
+| Low → High | Emotional Shift, Sound Bridge |
+| Low → Low | Spatial Flow, Gaze Lead |
+| Any → Climax | Sound Bridge (build), then Action Bridge (peak) |
+| Climax → Resolution | Emotional Shift, Time Jump |
+
+### How Transitions Work with ref_video Chaining
+
+When using serial chain generation (each segment's video passed as `ref_video` to the next), the model **physically continues from the previous segment's ending frame**. This changes how you design transitions:
+
+- **You don't need to describe the previous ending** in the next segment's prompt — the ref_video already provides that context
+- **Focus the transition design on the ENDING of each segment** — what the last 2-3 seconds look like determines how the next segment begins
+- **The `Continuing from the previous shot:` prefix** tells the model to seamlessly extend rather than start fresh
+- **Transition keywords in the prompt** (Gaze Lead, Spatial Flow, etc.) still help because they tell the model the *intent* of the transition, even though the visual anchor comes from ref_video
+
+**Example**: If S1 ends with a character looking right toward a door, and S2's prompt says `Continuing from the previous shot: the door opens to reveal a sunlit garden`, the model will generate S2 starting from that exact visual context.
+
+## Video Type Pacing Templates
+
+### Product Advertisement (30-60s)
+
+```
+ACT I — INTRIGUE (20%)
+  Hook: Mystery/problem/desire. Don't show the product yet.
+  Energy: 7→5
+
+ACT II — SHOWCASE (50%)
+  Reveal: Product enters frame dramatically
+  Function: Show key features in action (not listing specs)
+  Emotion: Connect product to a feeling or lifestyle
+  Energy: 6→8→7
+
+ACT III — CLOSE (30%)
+  Payoff: The "aha" moment or emotional climax
+  Brand: Logo/tagline, but make it feel earned
+  Energy: 9→4
+```
+
+**Common mistake**: Showing the product in frame 1. Build anticipation first.
+
+### Short Drama (60-180s)
+
+```
+ACT I — ORDINARY WORLD (15-20%)
+  Establish the character and their normal life
+  Energy: 5→4
+
+INCITING INCIDENT (5-10%)
+  Something disrupts the normal → story begins
+  Energy: 4→8
+
+ACT II — RISING ACTION (40-50%)
+  Obstacles, complications, deepening conflict
+  Include at least one "false victory" or "false defeat"
+  Energy: 6→7→8→6→9
+
+CLIMAX (10-15%)
+  The decisive moment — highest emotional intensity
+  Energy: 10
+
+RESOLUTION (10-15%)
+  New equilibrium — changed character, new understanding
+  Energy: 5→3
+```
+
+**Common mistake**: Flat Act II. Add reversals — things seem to get better, then get worse.
+
+### Brand Story (45-90s)
+
+```
+PROBLEM (20%)
+  Show the world before — what's missing, what's broken
+  Energy: 6→4
+
+JOURNEY (40%)
+  Exploration, experimentation, the search for something better
+  Energy: 5→7→6→8
+
+DISCOVERY (25%)
+  The brand/product as the answer — but show it through story, not claims
+  Energy: 9→10
+
+INSPIRATION (15%)
+  Leave the viewer with a feeling, not a sales pitch
+  Energy: 5→3
+```
+
+**Common mistake**: Making the "discovery" about features. Make it about transformation.
+
+## Rhythm Blueprint Format
+
+Before writing any multi-segment prompt, output a rhythm blueprint for user confirmation:
+
+```
+🎬 RHYTHM BLUEPRINT — [Title] ([Duration])
+
+Segments: [N] × 15s
+Narrative arc: [Template name]
+Overall energy curve: [number sequence]
+
+S1 (0-15s) — [Act label]
+  Goal: [What this segment achieves narratively]
+  Energy: [start]→[mid]→[end]
+  Camera: [Primary movement]
+  → Transition to S2: [Type] — [Brief description]
+
+S2 (15-30s) — [Act label]
+  Goal: ...
+  Energy: ...
+  Camera: ...
+  → Transition to S3: ...
+
+[... repeat for all segments ...]
+
+Key moments:
+  - Hook (0-3s): [Description]
+  - Midpoint ([time]): [Description]
+  - Climax ([time]): [Description]
+  - Final image ([time]): [Description]
+```
+
+Present this blueprint BEFORE writing prompts. Iterate on structure before investing in prompt details.

package/skills/director/references/style-library.md

@@ -0,0 +1,179 @@
+# Style Library
+
+A reference of visual style directions for the director to draw from and adapt per project.
+Each style includes camera, lighting, BGM, dialogue tone, rhythm, and transition guidance.
+
+## Product / E-Commerce
+
+### Minimal Showcase
+- **Camera**: Smooth orbit, slow dolly in, locked-off hero shot at close
+- **Lighting**: Clean studio, soft diffused light, white or light gray backdrop
+- **BGM**: Subtle electronic, minimal ambient
+- **Dialogue**: Brief, confident. "This is the one." / "Finally found it."
+- **Rhythm**:
+  - Energy pattern: `4→3→5→8→4` (slow build to single reveal peak)
+  - Cut rhythm: Long holds (4-6s), minimal cuts, let product breathe
+  - Breathing points: Between every angle change — hold for 2s before next move
+  - Transition style: Spatial Flow (orbit around product), Match Cut (product detail → wide)
+  - Tempo: 80 BPM ambient, no beat drops
+- **Best for**: Premium products, tech gadgets, jewelry, watches
+
+### Lifestyle Vlog
+- **Camera**: Handheld tracking, subtle slider drift, natural framing
+- **Lighting**: Natural daylight, window light, golden hour outdoors
+- **BGM**: Acoustic guitar, indie pop, lo-fi beats
+- **Dialogue**: Casual first-person. "Okay so I've been using this for a week..." / "Not gonna lie, I was skeptical."
+- **Rhythm**:
+  - Energy pattern: `6→5→7→6→8→5` (conversational ups and downs)
+  - Cut rhythm: Medium (2-4s per shot), natural rhythm like actual speech
+  - Breathing points: After each dialogue beat — let the reaction land
+  - Transition style: Action Bridge (hands picking up product), Gaze Lead (looking at product → detail)
+  - Tempo: 100-110 BPM indie, gentle momentum
+- **Best for**: Daily use products, fitness gear, kitchen tools, food
+
+### Premium Commercial
+- **Camera**: 2.35:1 widescreen, slow-motion details, cinematic dolly, rack focus
+- **Lighting**: Dramatic side light, rim light, volumetric haze
+- **BGM**: Orchestral swell, ambient synth, or silence with SFX
+- **Dialogue**: Minimal, measured. "Designed for moments that matter." Or no dialogue — let visuals speak.
+- **Rhythm**:
+  - Energy pattern: `3→2→4→6→9→10→5→3` (slow burn to grand crescendo)
+  - Cut rhythm: Very long holds (4-8s) in setup, accelerating to 1-2s at climax
+  - Breathing points: Long silence before the reveal moment (2-3s of near-stillness)
+  - Transition style: Match Cut (texture close-up → product), Sound Bridge (SFX crescendo into reveal)
+  - Tempo: Starts 60 BPM, builds to 120 BPM at climax, drops to 70 BPM at close
+- **Best for**: Luxury goods, skincare, fashion, high-end tech
+
+### UGC / Authentic Review
+- **Camera**: Front-facing selfie angle, close-up unboxing, handheld vertical
+- **Lighting**: Ring light or bedroom natural light
+- **BGM**: Trending TikTok audio, upbeat pop
+- **Dialogue**: Excited, authentic. "Okay this just arrived and I'm OBSESSED." / "Be honest — does this live up to the hype?"
+- **Rhythm**:
+  - Energy pattern: `8→7→9→8→10→6` (high throughout, brief dips for reaction)
+  - Cut rhythm: Fast (1-2s per shot), jump cuts, unboxing reveals
+  - Breathing points: Brief pause on genuine reaction face (1-2s hold)
+  - Transition style: Action Bridge (ripping open package), Emotional Shift (surprise face)
+  - Tempo: 120-130 BPM pop, constant energy
+- **Best for**: Beauty products, snacks, gadgets under $50, impulse buys
+
+### Dynamic Sports
+- **Camera**: Fast tracking, whip pans, low angle hero shots, GoPro POV
+- **Lighting**: Harsh natural sun, gym fluorescent, dramatic shadows
+- **BGM**: Trap, hip-hop, high-BPM electronic
+- **Dialogue**: Motivational, punchy. "No excuses." / "This changed my entire routine."
+- **Rhythm**:
+  - Energy pattern: `3→7→10→8→10→5` (explosive bursts with brief recovery)
+  - Cut rhythm: Fast (0.5-2s), beat-synced, whip pans as transitions
+  - Breathing points: One slow-motion hero shot per segment (2-3s) before next burst
+  - Transition style: Action Bridge (movement carries across cut), Match Cut (ball spin → wheel spin)
+  - Tempo: 140+ BPM trap/electronic, cuts land on every 2nd or 4th beat
+- **Best for**: Fitness equipment, sportswear, supplements, outdoor gear
+
+### Calm & Aesthetic
+- **Camera**: Slow drift, extreme macro details, gentle push-in
+- **Lighting**: Golden hour warmth, candle glow, soft backlight
+- **BGM**: Acoustic, ambient piano, ASMR-quality silence with subtle SFX
+- **Dialogue**: Whispered or minimal. "My evening ritual." / No dialogue — pure atmosphere.
+- **Rhythm**:
+  - Energy pattern: `3→2→4→3→5→2` (gentle waves, never above 5)
+  - Cut rhythm: Very slow (4-8s holds), dissolve-like transitions
+  - Breathing points: Everywhere — this style IS the breathing room
+  - Transition style: Spatial Flow (drifting through space), Time Jump (candle lit → candle low)
+  - Tempo: 60-70 BPM ambient, or no rhythm — pure texture
+- **Best for**: Candles, tea, home decor, skincare rituals, stationery
+
+## Narrative / Story
+
+### Dramatic Conflict
+- **Camera**: Shot/reverse-shot, fast cuts on tension beats, close-up on faces
+- **Lighting**: High contrast, motivated light sources, shadows
+- **BGM**: Tension strings, bass drops, silence before reveals
+- **Dialogue**: Character-driven, emotional stakes. Conflicting perspectives.
+- **Rhythm**:
+  - Energy pattern: `5→4→6→8→9→10→6→3` (classic dramatic arc)
+  - Cut rhythm: Slow in setup (3-5s), accelerating through conflict (1-2s), long hold at resolution
+  - Breathing points: After each confrontation beat — 2s silence before next escalation
+  - Transition style: Gaze Lead (character looks → reaction), Emotional Shift (calm → explosion)
+  - Tempo: Mirrors energy — 80 BPM setup, 130 BPM confrontation, 70 BPM resolution
+- **Best for**: Short drama, brand storytelling, episodic content
+
+### Suspense & Twist
+- **Camera**: Slow push-in, Dutch angle hints, locked-off wide with movement in frame
+- **Lighting**: Cold blue/green tones, vignette, underlit faces
+- **BGM**: Low drone, heartbeat, sudden silence
+- **Dialogue**: Sparse, cryptic. "You weren't supposed to see this." / Voiceover narration.
+- **Rhythm**:
+  - Energy pattern: `4→3→5→4→6→5→7→3→10` (long slow burn, sudden spike at twist)
+  - Cut rhythm: Deliberately slow (4-8s), getting slightly faster as tension builds, then SNAP
+  - Breathing points: Extended stillness before the reveal (3-5s of near-nothing)
+  - Transition style: Sound Bridge (creeping sound), Gaze Lead (what's behind the door?)
+  - Tempo: 60 BPM drone throughout, silence at twist, then 140 BPM burst
+- **Best for**: Thriller shorts, teasers, mystery content
+
+### Warm & Heartfelt
+- **Camera**: Medium shots, gentle tracking follows, long takes
+- **Lighting**: Warm palette, soft natural light, sunrise/sunset
+- **BGM**: Piano, acoustic, gentle strings
+- **Dialogue**: Honest, vulnerable. "I didn't realize how much I needed this."
+- **Rhythm**:
+  - Energy pattern: `4→3→5→6→8→7→5→3` (gentle rise to emotional peak, soft landing)
+  - Cut rhythm: Long takes (4-6s), few cuts, let moments breathe
+  - Breathing points: After every emotional line — hold the reaction for 2-3s
+  - Transition style: Time Jump (morning → evening), Spatial Flow (walking through home)
+  - Tempo: 75 BPM piano, slight swell at emotional peak, gentle fade
+- **Best for**: Slice-of-life, family content, feel-good stories, brand purpose
+
+### Comedy / Exaggerated
+- **Camera**: Wide-angle distortion, snap zooms on reactions, breaking the fourth wall
+- **Lighting**: Bright, flat, sitcom-style
+- **BGM**: Comedic stings, sound effects, trending audio
+- **Dialogue**: Deadpan, sarcastic, or over-the-top. Timing is everything.
+- **Rhythm**:
+  - Energy pattern: `6→4→8→3→10→5` (setup-punchline rhythm, contrast is key)
+  - Cut rhythm: Variable — long holds for deadpan (3-5s), snap cuts for punchlines (0.5s)
+  - Breathing points: The "beat" before the punchline — 1-2s of pregnant pause
+  - Transition style: Emotional Shift (serious → absurd), Match Cut (for visual gags)
+  - Tempo: Irregular — follows joke timing, not music
+- **Best for**: Skit content, viral comedy, brand humor
+
+## Visual / Artistic
+
+### Cyberpunk / Sci-Fi
+- **Camera**: Tracking through environments, HUD overlays implied by composition
+- **Lighting**: Neon (pink/cyan/purple), high contrast, rain reflections
+- **BGM**: Synthwave, dark electronic, glitch
+- **Dialogue**: Dystopian narrator or none. "In 2087, they said it was impossible."
+- **Rhythm**:
+  - Energy pattern: `6→8→7→9→10→8→6` (pulsing, electronic rhythm)
+  - Cut rhythm: Beat-synced (1-2s on beat), glitch cuts, strobe moments
+  - Breathing points: One wide establishing shot per 30s (3-4s hold) for scale
+  - Transition style: Match Cut (neon reflections), Spatial Flow (camera through neon corridors)
+  - Tempo: 128 BPM synthwave, cuts on every 4th beat, glitch on drops
+- **Best for**: Tech products, gaming, edgy brands, futuristic concepts
+
+### Comic / Manga Animation
+- **Camera**: Panel-by-panel reveal, speed lines, impact frames
+- **Lighting**: Cel-shaded, bold colors, ink outlines
+- **BGM**: Action anime soundtrack, dramatic SFX on impacts
+- **Dialogue**: Exaggerated character voices, text overlays as speech bubbles
+- **Rhythm**:
+  - Energy pattern: `5→7→4→9→10→3→8→5` (manga page-turn rhythm with impact spikes)
+  - Cut rhythm: Panel holds (2-3s), then rapid impact frames (0.5s), then hold for reaction
+  - Breathing points: After each impact frame — 2s hold on reaction panel
+  - Transition style: Match Cut (speed lines carry across panels), Action Bridge (punch carries through)
+  - Tempo: Variable — slow dramatic reads, explosive action bursts
+- **Best for**: Animated comics, IP character content, manga adaptations
+
+### Abstract / Art Film
+- **Camera**: Experimental — overhead, extreme angles, split screen, mirrored
+- **Lighting**: Moody, single source, chiaroscuro, color washes
+- **BGM**: Ambient, experimental, or classical reinterpretation
+- **Dialogue**: Poetic voiceover or complete silence
+- **Rhythm**:
+  - Energy pattern: `3→5→2→7→4→8→2→6→3` (non-linear, emotion-driven waves)
+  - Cut rhythm: Unpredictable — 8s holds next to 1s flashes, rhythm follows feeling
+  - Breathing points: Extended stillness (4-6s) between visual movements
+  - Transition style: Match Cut (abstract shape connections), Time Jump (surreal time flow)
+  - Tempo: No fixed BPM — rhythm emerges from visual texture and sound design
+- **Best for**: Brand films, art projects, mood pieces, festival content

package/skills/product-sheet-generate/SKILL.md

@@ -0,0 +1,75 @@
+---
+name: product-sheet-generate
+description: >
+  Generates multi-angle Product Design Sheet from product photos using Renoise
+  nano-banana-2. Shows front/back/side views, color palette, materials, and
+  construction details. Use when user says "product design sheet", "product
+  reference image", "multi-angle product view", or needs a visual reference
+  for video production. Do NOT use for scene backgrounds or video generation.
+allowed-tools: Bash, Read
+metadata:
+  author: renoise
+  version: 0.2.0
+  category: video-production
+  tags: [product, design, renoise]
+---
+
+# Product Design Sheet Generation
+
+Generate a comprehensive Product Design Sheet image from product photos using Renoise `nano-banana-2` model.
+
+## Arguments
+
+- First argument — Product images directory or single image path (required)
+- Second argument — Output directory (required, e.g. `output/foam-roller-20260313-1420/analysis`)
+
+## What is a Product Design Sheet?
+
+A single image containing:
+- Multiple angle views (front, back, side, three-quarter, bottom)
+- Color palette with Pantone or hex references
+- Material and texture callouts
+- Construction details and hardware close-ups
+- Proportions and measurements
+- Interior views (if applicable)
+
+## Instructions
+
+1. List images in the input path to confirm they exist.
+
+2. Upload each product image as a material:
+
+```bash
+node ${CLAUDE_SKILL_DIR}/../renoise-gen/renoise-cli.mjs material upload "<image_path>"
+```
+
+Note the material ID from each upload output (e.g. `#42`).
+
+3. Write an English prompt for the design sheet. Example:
+
+> Based on the provided product reference photos, generate a professional Product Design Sheet on a single clean white background. Include: multiple angle views (front, back, side, three-quarter, top), color palette swatches, material and texture close-ups, construction detail callouts, and relative proportions. Layout: clean organized grid. Photorealistic rendering. No cartoons or illustrations.
+
+Add `Do NOT include any text, labels, titles, annotations, or typography.` if user wants no-text version.
+
+4. Generate the design sheet, referencing the uploaded materials:
+
+```bash
+node ${CLAUDE_SKILL_DIR}/../renoise-gen/renoise-cli.mjs task generate \
+  --model nano-banana-2 \
+  --prompt "<english_prompt>" \
+  --materials "<id1>:ref_image,<id2>:image1,<id3>:image2" \
+  --resolution 2k \
+  --ratio 1:1
+```
+
+Assign material roles: first image as `ref_image`, additional images as `image1`, `image2`, etc.
+
+5. The command will output an `imageUrl` when complete. Download the image:
+
+```bash
+curl -sL "<imageUrl>" -o "<output_dir>/product_design_sheet.png"
+```
+
+6. Verify the output image was created and print file size.
+
+7. Show the generated image to the user for approval. If not satisfactory, adjust the prompt and regenerate.