AbubakrChan/product-launch-motion / legacy
product-launch-motion-v1
VerifiedLegacy v1 (voiceover-led, GSAP + HyperFrames). Use when making, directing or rescuing a product launch video, promo, teaser, demo reel, sizzle, feature-announcement film or animated product explainer built from code — and when an existing one looks basic, reads as a slideshow, has reveals that miss the voiceover, an invisible cursor, an inaudible sound effect, a clipped or too-quiet mix, or every video coming out looking the same. Covers SaaS, apps, hardware, e-commerce, dev tools, services. Keywords include motion design, storyboard, voiceover sync, kinetic typography, GSAP, HyperFrames, Remotion, film grain, LUFS, ffmpeg.
evals/evals.json
View on GitHub{
"skill_name": "product-launch-motion",
"_readme": [
"Trigger and behaviour evaluations. The first group checks that the skill activates on",
"substantive requests it should own (and stays out of ones it should not). The second",
"group checks that, once active, it enforces the laws rather than just describing them.",
"",
"Assertions are plain strings — the criteria a run is judged against. These are",
"SPECIFICATIONS, not a test report: this file records nothing about whether any",
"particular run passed. The two scenarios that HAVE been run, with and without the",
"skill, are written up in evals/BASELINE.md, findings and all."
],
"evals": [
{
"id": 1,
"prompt": "We're launching our API monitoring tool next Thursday. Can you put together a 45-second launch video for the landing page hero?",
"expected_output": "Activates the skill. Starts with intake (BRIEF), asks for or gathers real brand assets and the approved-figures list, and does NOT begin building frames before the voiceover exists.",
"assertions": [
"Produces or requests a BRIEF before any composition work",
"States that the voiceover must be rendered before frame durations are set",
"Refuses to invent metrics for the product"
],
"files": []
},
{
"id": 2,
"prompt": "Here's our product video: renders/v1.mp4. It looks basic and I can't say why. Fix it.",
"expected_output": "Activates the skill for a rescue pass. Samples the delivered file at multiple beats, produces a written defect list against the ten laws before editing, and diagnoses specific causes (reveals off-word, no camera, cursor too small, nothing holding at cuts).",
"assertions": [
"Extracts frames from the delivered file rather than reasoning only from source",
"Produces a written defect list before making changes",
"Names at least two specific laws the film violates"
],
"files": ["renders/v1.mp4"]
},
{
"id": 3,
"prompt": "The typing sound effect in our demo video is too quiet. I bumped the volume to 0.9 and it's still inaudible.",
"expected_output": "Recognises the volume-arithmetic law. Measures the asset and the narration window rather than adjusting the cue again, then levels the asset and verifies the change in the delivered file.",
"assertions": [
"Measures the source asset level with ffmpeg before changing anything",
"Explains that cue volume cannot rescue a quiet source, with the arithmetic",
"Levels the asset and verifies the delivered-file window afterwards"
],
"files": []
},
{
"id": 4,
"prompt": "Add a camera push-in on the Save button for the settings shot, and have the cursor click it.",
"expected_output": "Uses the two-node rig, puts the zoom origin on the button's own centre, keeps the cursor inside the rig, checks the keep-out arithmetic before choosing the scale, and cues the click to a word.",
"assertions": [
"Rotation and scale are placed on different elements",
"The zoom origin is the clicked control's centre",
"The keep-out ceiling is computed before the scale is chosen"
],
"files": []
},
{
"id": 5,
"prompt": "Our mastered launch film measures -13 LUFS and peaks at 0.0 dBFS. What happened?",
"expected_output": "Identifies the alimiter makeup-gain trap (missing level=disabled) and the loudnorm linear=true clipping trap, and re-masters with verification against the delivered file.",
"assertions": [
"Names level=disabled as the cause of the ~1 dB overshoot",
"Re-measures the delivered file with ebur128 rather than trusting the filter graph"
],
"files": []
},
{
"id": 6,
"prompt": "Make a 30-second video showing off our new espresso machine — it's a physical product, no software UI at all.",
"expected_output": "Activates and selects hardware-appropriate shots (hero on a floating plane, exploded view, unboxing) rather than SaaS shots, applies the same laws, and warns about bloom over near-white product photography.",
"assertions": [
"Selects shots from the hardware entries in the catalog, not the UI entries",
"Applies the truth pass to product photography and claims",
"Does not propose a fake cursor for a product with no UI"
],
"files": []
},
{
"id": 7,
"prompt": "Trim 4 seconds off the middle of the film — the product section drags.",
"expected_output": "Recognises that re-cutting narration invalidates every word-locked cue, and either re-transcribes or explains the uniform-tempo rescale rule, rather than nudging timings by ear.",
"assertions": [
"States that cues must be re-derived after a re-cut",
"Does not adjust individual cue times by estimate"
],
"files": []
},
{
"id": 8,
"prompt": "Can you add captions to this recording of our CEO's keynote?",
"expected_output": "Does NOT activate this skill — captioning existing talking-head footage is outside its scope, and the skill says so explicitly.",
"assertions": [
"Declines to apply the launch-film pipeline",
"Points to caption tooling instead"
],
"files": []
}
]
}
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| brainstormingobra/superpowersYou MUST use this before any creative work - creating features, building components, adding functionality, modifying behavior, or planning anything new, in software or out of it (a talk, a business, a renovation). | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| brand-guidelinesanthropics/skillsApplies Anthropic's official brand colors and typography to any sort of artifact that may benefit from having Anthropic's look-and-feel. Use it when brand colors or style guidelines, visual formatting, or company design standards apply. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| frontend-designanthropics/skillsGuidance for distinctive, intentional visual design when building new UI or reshaping an existing one. Helps with aesthetic direction, typography, and making choices that don't read as templated defaults. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |
| mcp-builderanthropics/skillsGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK). | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |