Plan a product promo video, scene by scene

Extract the ZIP. Copy its lc-motion-product-promo folder into ~/.claude/skills/ for Claude Code or ~/.codex/skills/ for Codex, then refresh the native app. Ask for the skill by name and answer its questions. The package includes the same published instructions and result checks. Native host execution is not yet verified.

For Claude web, use the connector or copy the prompt. Keep private details in your assistant.

Copy, paste into your assistant, then answer its questions. No form or account is required here.

Open in ChatGPT ↗

Already connected? Copy the public starter below and paste into Claude. Keep private details in the chat.

Needed before the promised result: Your product. You can still copy the reusable prompt and provide these in your assistant.

Saved only on this device. Copy counts contain the skill ID, never your text.

## LazyCoder conversation guide (conversation-v2)
1. Start a conversation, not a form-filling task. Read my latest request and the facts already in this chat before asking anything.
2. Ask only the missing questions needed for the requested result, in a short numbered list (at most three at a time), then WAIT for my answers. Use a native question tool if available; plain chat works too. Never ask me to replace placeholders or fill in a template.
3. Skip questions already answered. Treat unresolved {{placeholders}} and [FIELD LABELS] as missing facts, never literal input. Do not treat an illustrative example as my facts.
4. Use the authored defaults only when relevant; mention material defaults. Ask conditional questions only when their condition applies; if unsure whether it applies, ask. Optional preferences must not block a safe first draft unless the authored instructions require confirmation.
5. Do not invent tasks, dates, names, claims, permissions or missing source material. When required facts are absent, ask and wait instead of producing the final artifact. If everything needed is already supplied, proceed without re-asking.
6. My latest scope correction overrides earlier examples and scope, within the recipe safety and source rules. Resolve conflicts before acting. The authored instructions and output contract remain authoritative; this conversation guide does not grant tool, account, payment or publishing permission.
7. Keep my private answers and generated result in this assistant chat. Do not send them to LazyCoder. Generate the requested format here and check it against the output contract before handing it over.

### Questions to ask only when missing
- What should I use for “Your product”? [required] What it is, what it does, who it helps, and any images you have. Spell the name exactly as it should appear.
- What should I use for “Who it is for and what they should do next”? [optional] Who watches, and the action or link viewers should use after watching.
- What should I use for “Length”? [optional] The final edited length, including the closing card. Authored default: 30 seconds
- What should I use for “Delivery frame”? [optional] Choose the publishing proportions. Authored default: Portrait 1080 by 1920 Choices: Portrait 1080 by 1920; Landscape 1920 by 1080; Square 1080 by 1080
- What should I use for “Your video or image tool”? [optional] Name the tool you will paste the prompts into, and whether it makes video or still images, if you know. Authored default: Not named
- What should I use for “Look, sound and permissions”? [optional] Colours, fonts, style, whether you want narration, and anything you already own or have permission to use. Authored default: Clean, product-led look; music and effects, no narration; rights still to check

## LazyCoder delivery guide (delivery-guide-v1)
1. Read the latest request, source material and authored requirements first. Ask only for missing facts. Check that this host can create the promised format; use only a fallback explicitly allowed by the authored recipe. If neither is available, explain the gap before starting. Do not promise a rendered film, editable deck or working file when you can only provide its script or outline.
2. Create the complete result in the authored output specification, preserving source facts and the latest agreed scope. Keep assumptions and missing information explicit. Do not silently omit required parts or leave unresolved ingredient markers. Intentional blanks in a reusable template are different from unanswered recipe inputs.
3. Check the actual result against every authored check, including source facts, counts, dates and required sections. Where tools allow it, inspect the actual saved file and compare it with the answer. Record which checks ran, which failed and which still need human or destination review. A filename, render, successful command or preview alone does not prove the promise was met.
4. Hand over the complete editable file, or the complete copyable fallback explicitly allowed by this recipe. Say which delivery you used and any unresolved limitations. Do not report a file as attached, imported, sent or published unless that action actually succeeded. Keep private inputs and output in this assistant; never send them to LazyCoder.
Use existing authorization for ordinary work. Before any additional paid generation, external write or publishing step, check that the current request authorizes that exact action and cost. If it does not, finish the useful authorized draft and ask at that boundary. Never silently retry or replace an ambiguous paid job.
The authored recipe and latest safe scope remain authoritative. This guide grants no new tools, permissions, provider access or claim of successful execution.

## Authored skill instructions
## 1. Outcome
Create a complete production package someone can execute in their own generator and editor. Choose one product-specific concept for the intended viewer, showing one supported benefit through a hook, a demonstration and a closing call to action.

Done means: the package contains a reconciled timeline, exact screen copy, per-scene animation and audio directions, an asset register with permissions, and one self-contained generation prompt per scene. All sections agree, with unresolved dependencies clearly identified. Deliver the written package in one response.

## 2. Context
Product facts and references:
"""
[YOUR PRODUCT]
"""
Audience, task and intended action: [VIEWER AND ACTION]
Final edited runtime: 30 seconds
Delivery frame: Portrait 1080 by 1920
Generator and limits: Not named
Brand guidance and permissions:
"""
Clean, product-led look; music and effects, no narration; rights still to check
"""

Write for the person producing the promo; use plain production instructions, in the language of the brief. Default to a clear, confident, benefit-led tone. Extract the exact product name, one customer task, what the product does, its real appearance, the channel and the action before choosing a treatment. Add narration only when the look-and-sound note asks for it.

Blank values or unchanged ingredient placeholders count as missing. Apply declared optional defaults. If audience is missing, infer a likely viewer from the supported task and label the inference; if the action is absent, use “Explore” followed by the exact product name without inventing a URL. Default a missing length to 30 seconds and a missing frame to portrait 1080 by 1920; with no tool named, treat it as unknown. Disclose these assumptions.

Never infer product identity or capabilities. If no exact name or usable product facts are available, return a concise missing-input notice identifying the essentials needed; do not fabricate a promo. Missing imagery is a production dependency, not permission to invent product appearance.

## 3. What to pull out
Use pasted facts and accessible attachments only; no browsing, repository access or external execution is required. Say when an attachment cannot be inspected and rely only on its supplied description.

Extract:
- Exact product spelling, what it does and a concrete customer success moment.
- Fact identifiers F1, F2 and onward. Treat the person's own description of what the product is and does as the fact source. Flag for verification only numbers, comparisons, superlatives ("fastest", "best"), third-party endorsements and health, money or safety claims; keep flagged claims out of the copy until confirmed.
- Official logos, authentic screenshots and user-owned images before considering newly created material.
- Palette, typography, distinctive brand or product behavior, and reference assets worth carrying through the animation.
- Provenance, permission status and attribution requirements for every visual, font and sound.

Merge duplicates. An explicit correction supersedes earlier material; otherwise list conflicting facts as unresolved and exclude the disputed claim. Review long briefs by topic without dropping restrictions. Phrase an unclear benefit plainly rather than dropping it. For missing evidence or permissions, record “Unresolved” and supply a concrete verification or replacement route; never assume ownership or clearance.

## 4. Steps
1. Build the fact and asset inventories. Select one supported task and benefit for the intended audience.
2. Consider contrasting concepts internally. Select one rooted in a recognizable product behavior or brand element; give a brief product-specific rationale, without presenting discarded concepts.
3. Establish one visual system: palette, typography, lighting, background, motion rhythm and product treatment.
4. Create one master scene ledger. Give each scene a communication purpose and connect attraction, demonstration and invitation. Derive every table, copy line, audio cue, asset reference and prompt from this ledger.
5. Allocate scene timing before writing narration or sound cues. Prefer hard cuts or transitions contained inside each scene’s allotted interval. Allocate a readable closing hold.
6. Plan composition, product motion, camera movement and entry/exit continuity. Fit generator clip lengths to the edit and specify trims.
7. Add concise copy, music and effects; add narration only when requested. Run part 7 and repair the package before replying.

## 5. Rules
- Preserve the product name character for character, including capitalization, punctuation and spacing, everywhere it appears. Translate surrounding copy only. Preserve supported features and supplied appearance; no invented claims, statistics, testimonials or implied endorsements.
- Trace each advertised capability to a fact identifier from the person's description. A metaphor must not suggest a nonexistent feature.
- Use supplied product imagery through references or editor compositing. Without suitable imagery, plan a labeled product insert with a sourcing dependency; do not synthesize a speculative product or interface.
- Describe placement in one plain line per scene ("text in the upper third, product centred"). Keep text and the product clear of the outer edges, where platforms crop or place their buttons.
- Keep text and captions off the product's key details, with strong contrast and a size readable on a phone. Shorten copy when reading time is tight.
- Put exact lettering, logos and captions in editor overlays whenever generation cannot preserve them reliably; default to editor overlays. Generation prompts must reserve the same space and prohibit generated lettering.
- Keep palette, lighting, typography and motion behavior consistent. Movement must direct attention or explain the task.
- Do not imitate a named artist, clone or imitate a real person’s voice, or copy lyrics or melodies. Describe original visual and audio qualities instead. Record asset rights without claiming legal clearance.
- No fabricated destinations, unfinished copy placeholders or invented licenses. State unresolved dependencies outside finished screen copy. Never claim footage was rendered, inspected or measured.

## 6. Output
Deliver one complete Markdown document, ready to save as `product-promo-package.md`. Use these sections in order; write out every scene and prompt in full.

1. **Brief and concept**: about 150 words covering viewer, task, benefit, chosen concept, hook-to-action progression, rationale, dimensions, runtime, tool mode and assumptions. Include the exact product name and the fact list, with any claim flagged for verification.
2. **Visual and composition guide**: shared palette, fonts, lighting and motion rules, and where text and captions sit in the chosen frame, in plain words.
3. **Master scene table**: columns `Scene | Start–end | Edit duration | Purpose | Visual and placement | Product movement and camera | Entry, exit and transition | Exact screen copy (editor overlay) | Audio cue | Trim and assembly note | Fact and asset IDs`. Number S01 onward. Mark intentional text-free scenes “Text-free.” Include the exact product name in visible copy and a specific final action. State line breaks. Include music/effect cues or explicit silence for every scene.

Use consecutive, nonoverlapping timeline intervals by default; durations must sum to the target. State whether transitions are inside these intervals. If overlapping clips are necessary, give clip placements, overlap lengths and the equation: sum of retained clip lengths minus overlaps equals final runtime. Distinguish source generation lengths from edited scene durations; never double-count transitions.

4. **Audio plan**: music mood, approximate tempo, instrumentation, energy changes, scene-linked effects and a deliberate final accent, fade or silence. Specify narration ducking when applicable. For requested narration, provide `Scene | Spoken words | Delivery | Verbatim captions`; keep speech comfortably within each scene, roughly two words per second with breathing space. Captions must reproduce the spoken words exactly. Otherwise state “No narration.”
5. **Asset register**: columns `Asset ID | Item and use | Source or creation route | Permission status and evidence | Attribution | Unresolved dependency or replacement`. Cover every product image, logo, background, generated element, font, music track, effect and optional voice. Mark proposed assets as unsourced. Unknown attribution obligations remain unresolved, never “None” by assumption.
6. **Scene prompt blocks**: one independently pasteable prompt per scene, labeled with its scene ID and mode. Repeat dimensions, requested generation duration, exact product identity and appearance constraints, reference asset IDs with descriptive filenames or descriptions, composition, action, camera, style, starting/ending state, continuity and exclusions. Do not refer to “the style above.” State required reference uploads before each block; asset IDs alone are insufficient generator instructions.

For video mode, describe timed action and camera behavior. For image-only mode, request a still composition and provide separate editor instructions for movement and its scene duration; do not ask an image model to render motion. If the tool is not named, give one video prompt per scene, then one shared paragraph explaining how to use any of them as a still-image prompt and add the movement in the editor. Unknown duration support means requested duration, unverified; explain once how to generate the nearest supported longer clip and trim it. Overlays, sound cues and trims live in the master table, not repeated per prompt. Keep overlay text outside visual-generation instructions.
7. **Assembly and checks**: numbered editing instructions following the timeline, then a compact acceptance report. Separate checks verified by rereading or arithmetic from production checks still pending, including visual fidelity, actual reading comfort and licensing confirmation. List blockers and their replacement routes; no claim of inspected footage.

## 7. Self-check
Before replying, run these named checks and fix failures:
- **Coverage**: every required section and scene field exists; no unfinished copy placeholders remain.
- **Truth and identity**: exact product spelling agrees throughout; claims come from the person's description; flagged claims stay out of the copy; no invented endorsements or appearance.
- **Timing**: recalculate intervals, durations, trims and overlaps; the edited runtime equals the target, with the closing action included.
- **Composition and reading**: frame dimensions and reserved areas agree; copy is brief enough for planned holds. Treat readability as a planning judgment pending production.
- **Sound and narration**: every scene has matching cues; narration allows pauses, captions match verbatim and the ending is intentional.
- **Rights**: every referenced asset exists in the register with provenance, permission and attribution status or a replacement route.
- **Consistency**: compare scene IDs, product treatment, dimensions, timings, references, overlays and sound between the table and the prompts. Correct discrepancies at their source in the ledger.

Report unresolved factual or permission dependencies honestly. Do not present proposed directions as completed production checks.

## 8. Setup
Paste this recipe and fill-ins into a chat app. Attach product references if supported; otherwise paste facts and visual descriptions. No tools or installs are needed for the package. If the app offers file tools, save the document as Markdown; otherwise return its complete paste-ready text.

Upload the specified references to your own generator and paste each matching scene prompt. Generate stills or clips, then assemble them in your editor using the supplied timing, trims, overlays and sound cues. Longer supported clips can be trimmed to the planned intervals. Rendering, editing, inspection and final licensing checks happen in your production tools.

Paste the result after your assistant creates it. This optional check stays in this tab and is not sent to LazyCoder.