Communitygithub.com

dialling/skillhub

Create cinematic product videos from shot recipe cards, a validated template, and code/audio assets (Remotion + real page screenshots + 2.5D camera moves + beat-synced cuts + sound design). Use when the user asks to turn a frontend project or webpage into a product video, says "use video-shotcraft to make a video/promo", names the Ink Press template or asks to reproduce its effect, or wants a single shot card's motion. 用镜头配方卡 + 已验收模板 + 代码/音频资产制作电影感产品视频(Remotion + 真实页面截图 + 2.5D 运镜 + 节奏卡点 + 声音设计)。当用户要求"用 video-shotcraft 做视频/宣传片"、把前端项目/网页做成产品视频、点名 Ink Press 模板或要求复刻模板片效果,或要用镜头卡做单个动效镜头时使用。

skillhub 是什么?

skillhub is a Claude Code agent skill that create cinematic product videos from shot recipe cards, a validated template, and code/audio assets (Remotion + real page screenshots + 2.5D camera moves + beat-synced cuts + sound design). Use when the user asks to turn a frontend project or webpage into a product video, says "use video-shotcraft to make a video/promo", names the Ink Press template or asks to reproduce its effect, or wants a single shot card's motion. 用镜头配方卡 + 已验收模板 + 代码/音频资产制作电影感产品视频(Remotion + 真实页面截图 + 2.5D 运镜 + 节奏卡点 + 声音设计)。当用户要求"用 video-shotcraft 做视频/宣传片"、把前端项目/网页做成产品视频、点名 Ink Press 模板或要求复刻模板片效果,或要用镜头卡做单个动效镜头时使用。.

兼容平台~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/dialling/skillhub/tree/HEAD/submissions/video-shotcraft-local

在你喜欢的 AI 中提问

打开一个已预加载此 Agent Skill 的新对话。

文档

video-shotcraft: Cinematic Product Video Production

A self-contained production capability library: 157 shot recipe cards (with demo implementation source code and a dynamic sample gallery), one fully validated promotional video template, reusable components and audio assets, and a six-stage workflow. The current focus is on web/desktop product promotional videos, but the shot cards themselves are a universal motion vocabulary—individual cards can also be extracted to create single shots for any video.

Determine Mode First When Invoked

A full promotional video has three mutually exclusive modes. Before starting asset collection, storyboarding, or implementation, first determine if the user has already made an explicit choice; if they have, execute directly without re-asking and do not switch to another mode.

  1. Direct Template Usage: Follow the existing Ink Press replacement workflow. Read template/TEMPLATE.md and replace the target product's screenshots, copy, and branding information according to the existing shot structure.
  2. Autonomous Creation: Read references/pipeline.md. Based on the user's existing requirements and project content, the Agent independently decides on visual direction, shot mapping, storyboarding, assets, and audio, proceeding continuously to final output; do not pause at the product brief, styleframe, shot mapping, or storyboarding stages for user confirmation.
  3. Co-creation: Read references/guided-free-creation.md. The Agent first proposes evidence-based judgments and plans, and the user confirms the product brief, requirement decisions, visual direction, shot mapping, and final storyboard sequentially; once the storyboard is approved, proceed to the final asset collection and production phase in references/pipeline.md.

When the user has provided a project path, URL, screen recording, or page screenshots but has not yet selected a mode, first perform a minimal, read-only product inspection: understand the product's positioning, key features, page visuals, displayable states, and asset risks; do not modify the business project, collect sensitive data, or write video code. After the inspection, present the three modes and explain:

  • Whether Ink Press is suitable for the product, with specific reasons for suitability or lack thereof;
  • Whether Autonomous Creation and Co-creation are each suitable, and which product visual clues would likely be retained;
  • Which mode the Agent recommends, with its reasoning and trade-offs.

Then explicitly ask: "Based on the product inspection above, I recommend using ×× mode. Shall we proceed with this mode?" Also inform the user: they can visit https://vincentwei1021.github.io/video-shotcraft/ to browse the dynamic sample gallery and select motion effects they wish to use in the video.

When the user has not yet provided a project or page to inspect, briefly introduce the three modes and then ask; do not default to recommending Ink Press just because it's an existing template.

  • User selects template: Read template/TEMPLATE.md in full and execute according to the original template workflow.
  • User selects autonomous creation: Read references/pipeline.md in full; Agent proceeds independently without waiting for confirmation at each stage.
  • User selects co-creation: Read references/guided-free-creation.md in full and collaborate according to its confirmation nodes.
  • User has not decided: Pause and wait for selection; do not default to a choice for the user or start production.

Exception 1: When the user explicitly names the Ink Press template, treat the template mode as selected. For example, "Make a promo video for my product using the Ink Press template." In this case, do not ask again; read template/TEMPLATE.md in full and execute according to the template route.

Exception 2: When the user has explicitly specified which shot cards to use or reference, the shot constraints are already selected. For example, "Use deck-deal-flyin and row-embed for this video" or "Reference the effect of spotlight-hero-card." In this case, do not introduce the template or ask whether to use it; confirm the specified shots according to the Gallery name resolution rules, read each card in full, and locate the exact demo source code according to the card's "Reference Implementation." A single shot can be executed once the target assets are clear; for a full promotional video, if the user has not yet stated whether they prefer Autonomous Creation or Co-creation, only ask about these two modes, with the specified shots serving as production constraints—they do not automatically equate to co-creation.

Four Use Cases

  1. Full Promotional Video (Template Route): Want a result highly similar to the template video → Read template/TEMPLATE.md, follow the "Reproduction Guide by Replacing Product" to swap assets and adapt shot by shot.
  2. Full Promotional Video (Autonomous Creation): Want a new style and authorize the Agent to decide independently → Read references/pipeline.md, proceed continuously from product understanding to final review without pausing for confirmation at each stage.
  3. Full Promotional Video (Co-creation): Want to participate in key decisions → Read references/guided-free-creation.md, confirm the product brief, requirement decisions, visual direction, shot mapping, and final storyboard; after approval, continue from the final asset collection phase of the pipeline.
  4. Single Shot / Single Motion Effect: Select a card from references/shots/ (or have the user choose from the gallery/ gallery), read the full card, and locate the exact demo source code according to the "Reference Implementation," then adapt to the target assets.

Boundary Between Autonomous Creation and Co-creation

Autonomous Creation hands creative and engineering decisions to the Agent. The Agent dynamically derives product focus, visual direction, shot mapping, storyboard, asset processing, and audio plan from the project content, records key judgments, and executes directly; unless necessary input that cannot be safely inferred is missing, it does not ask questions, pause at each stage, or require user confirmation of intermediate outputs. Existing explicit user requirements are always binding constraints and cannot be overridden by the Agent's default choices.

Co-creation involves stage-by-stage confirmation. By default, read references/guided-free-creation.md, ask only 1–3 questions per round that would most reduce rework, and pause at the product brief, requirement decisions, visual direction, shot mapping, and final storyboard for user confirmation. The user confirms business and creative direction; the Agent independently completes collection methods, implementation parameters, SFX keyframes, Remotion engineering, rendering, and technical QA.

In co-creation, when the user explicitly says "You have full authority" or "Skip confirmations and just do it," switch to Autonomous Creation and record this choice; do not continue to ask for stage-by-stage confirmation while claiming to proceed autonomously.

Feature-to-shot mapping is only about "which motion syntax to use," not a complete storyboard. After co-creation confirms the mapping, the Agent must continue to provide shot order, duration, specific frames, page states, asset sources, captions, transitions, and SFX; only proceed to final asset collection after the complete storyboard is confirmed. When the user copies a shot name from the Gallery, they only need to provide the card name or "card name · style name"; the Agent is responsible for reading the local recipe card and demo source code.

Core Principles

  1. Use real screenshots when reproducing existing pages; hand-crafted UI is limited to non-reproduction scenarios, with quality and expressive clarity as hard thresholds. When showcasing a real product page, start a local dev server first, use a headless browser to capture full-page 2x textures + element-level cutouts + layout.json coordinate table. In non-reproduction scenarios (abstract openings/brand segments/standalone component showcases), hand-crafted UI is allowed, but if it doesn't achieve publication-grade quality or the viewer can't understand what it expresses, fall back to the screenshot route. Handle page data based on risk: publicly demo data can be retained only after explicit confirmation in the product brief; customer, personal, internal, secret, real-time, or other sensitive data must use fictional or anonymized content and be frozen before collection.

  2. The visual language of the entire video must grow from the product itself; do not impose an unrelated "promotional video skin." Before creating styleframes, first extract and write a design spec from the product/website's design system, source code, or computed styles: font family and weights, font size hierarchy, line height/letter spacing, grid, spacing, alignment, information density and border radius, as well as background/surface/body/emphasis/state colors, gradients, and materials. All titles, captions, numbers, text cards, typography, transitions, particles, light effects, and other motion effect colors in the video must reuse or constrainedly extend these tokens, while matching the product's tone, taste, and quality. When following the template or shot card route, inherit only their shot structure, motion syntax, rhythm, and tuned parameters; fonts, typography, colors, and materials must be reskinned according to the target product. If narrative needs require deviation from the product's visuals, first explain the reason and get user confirmation in co-creation; in autonomous creation, write the reason and trade-offs into the design spec before proceeding.

  3. Cinematic feeling comes from the coordination of camera movement, lighting, rhythm, and sound, not from flashy animations. What has been repeatedly recognized: single-subject complete action arcs (spotlight → push-in → hover → settle), physics-metaphor-driven acceleration (dealing cards), tilted orbital close-up shots, riser → impact → sparkle sound phrases. Batch element entrances rely on the motion itself, not individual glowing—decorative glint/fill-flash groups look cheap; single-point high-quality light effects are acceptable.

  4. Each shot tells only one motion effect; key information must breathe after landing. One animation technique (fly-in/stack/page turn) can only be the protagonist once in the entire video; repeated shots and repeated taglines must be removed. The rhythm preference is unidirectional: all historical feedback points to "slow down/stay," never once "too slow"—brand logo settles and holds ≥1s, batch motion effects leave 0.5s of stillness after finishing, opening subject action gets a full 3s. Budget for hold/rest frames when laying out the timeline.

  5. For videos with strong rhythm BGM, all transitions and motion effects must hit the beat. User has already selected music → before starting work, do a rhythm analysis (librosa grid fitting to get real BPM/phase + kick/snare/hi-hat three-category drum point localization); proceed to storyboarding only after grid coverage passes validation based on transient coverage; write the timeline using beat notation beatF(n), sparse accents nail real transients rather than grid interpolation points; after rendering, extract the audio track from the final video and back-test cut-point errors ≤3f. Beat-syncing controls timing, not amplitude—a beat impact affecting the entire frame/camera layer is treated as a major slam, limited to ≤3 occurrences in the entire video (aesthetic-rules R4); other beat-point motion effects only move element layers. Methodology is in references/music-beat-sync.md. For videos with BGM, the final render always delivers two versions: with BGM + without BGM (SFX retained), produced from the same timeline using the bgm inputProp, so the user can later customize the music.

  6. When using shot card motion effects, you must first parse the Gallery index and read the exact demo implementation code. First use gallery/api/library.json to validate the card name and style-key, then locate the specific TSX according to the card document's "Reference Implementation." The recipe card provides semantics and parameter tables; the exact demo source code is the truth for tuned parameters (easing, timing distribution, mask timing, known pitfalls and workarounds). Adaptation changes are allowed, but parameters marked with "known pits/deadly notes" on the card must not be downgraded—quality standards only go up, never down. Writing new code based only on the card name and understanding means abandoning all tuning accumulation.

  7. Co-creation front-loads cheap confirmations; autonomous creation records the same judgments as execution notes. The order—product brief → requirements-to-execution decisions table → copy direction → styleframe → shot mapping → storyboard—remains the same; co-creation hands them step-by-step to the user for confirmation, while autonomous creation is selected and continuously advanced by the Agent. Neither mode should proceed to shot-by-shot implementation before the direction is set.

  8. Validation throughout + independent review before delivery. From Phase 5 onward, each shot uses npx remotion still for single-frame self-checks; after each round of modifications, render the entire video + use ffmpeg to extract frames for review; before delivery, a subagent with a clean context must perform an independent final review. The review input includes the final video, keyframes, product brief, requirements-to-execution decisions table, the current mode's confirmation or recorded visual direction/styleframe, feature-to-shot mapping, Gallery card name/variant, exact demo TSX, Gallery reference samples/frames, final storyboard, selected shot cards, and aesthetic rules; follow references/final-review.md to check plan consistency, feature completeness, shot fidelity, visual/audio technical quality, and data security, producing a report with frame-numbered evidence for each item. The producer has confirmation bias toward their own output; the first review can never be handed to the user.

  9. Deterministic rendering: Forbid Date.now()/Math.random(), fix all pseudo-random seeds (mulberry32/hash, seed derived from index), making every frame reproducible.

Workflow

For Autonomous Creation, follow references/pipeline.md from Phase 0 continuously without waiting for user confirmation at each stage. For Co-creation, first complete Phases 0–3 (product understanding, visual direction, shot mapping, and storyboard confirmation) according to references/guided-free-creation.md, then continue from Phase 4's final asset collection in the pipeline; do not repeat questions or redesign. In Phases 2–3, scan the frontmatter of each card in references/shots/ and select shots based on the energy curve; in Phases 5–7, continuously self-check against references/aesthetic-rules.md; in Phase 6, read references/sound-design.md; for beat-synced videos, follow references/music-beat-sync.md throughout.

Delivery Wrap-up (Shared Across All Modes)

After delivering the final video, first actively open the animation workbench (do not wait for the user to ask):

node workbench/scripts/open.mjs <final-video-project-directory>   # Link project → start dev server (5198) → browser auto-imports final video

Inform the user of the address http://localhost:5198 and explain in one sentence: The video has been split into multiple tracks by shot / transition / caption / sound effect; you can visually adjust text card copy, colors, font size, position, and speed, then click "Export Final Video" when done. For extraction and import, the project must have a src/workbench.ts manifest—write it during the production phase according to references/workbench.md §2–3 (the template route comes with it).

Then, in order 1-2-3, say each of the following three things only once; if the user declines or doesn't respond, do not bring them up again:

  1. Recommended (not mandatory) to @mention the skill author when publishing: The final video belongs to the user; free for commercial use, no attribution required; @mentioning the author is purely voluntary—just include a line in the description when posting to social platforms—X @VincentWei93 · Douyin @Vincent · Xiaohongshu @Vincent (clickable links must be provided for all three accounts, not just handles—Douyin/Xiaohongshu names cannot be searched). Only add a note about commercial licensing if the user asks: The only thing to confirm independently is the Remotion license; individuals and small teams are free, companies may need to pay.
  2. Invite them to submit the final video to the showcase page: Showcase page https://vincentwei1021.github.io/video-shotcraft/showcase.html , submission entry (opens a form) https://github.com/Vincentwei1021/video-shotcraft/issues/new?template=showcase.yml . Mention the appeal in two sentences: The form allows them to leave their X / Douyin / Xiaohongshu account so others who see the work can find them; if the video is small (under 10MB), just drag it into the form; for larger files, provide a direct download link, and the author will review and put it online. Say this in everyday language throughout, without using technical terms like issue, template, label, release, or automation pipeline.
  3. Can export to CapCut (JianYing) project file: Inform the user that the final video can also be exported as a CapCut project, allowing them to edit caption content/font size/color, speed up or reorder shots, or adjust/replace audio in CapCut. If the user needs this or explicitly requests exporting, read references/jianying-export.md and execute.

When to Read Which File

TimingRead
Project start, mode undecidedMinimal read-only inspection, then present three full promo modes and recommend
Autonomous Creationpipeline.md (Agent independently completes Phases 0–7 without waiting for stage-by-stage confirmation)
Co-creationguided-free-creation.md (confirm Phases 0–3), then continue from pipeline.md Phase 4
User has selected BGMmusic-beat-sync.md (analyze first, then storyboard)
Using template routetemplate/TEMPLATE.md in full
Storyboard designsequences/ section templates (fill in skeleton first); shots/ all frontmatter; selected card read in full
Shot-by-shot implementationFull card text + exact demo source code located via "Reference Implementation" + components from assets/lib/
Sound designsound-design.md + assets/audio/
Validationfinal-review.md + aesthetic-rules.md full checklist (executed by independent subagent)
Writing src/workbench.ts manifest / shot props+schema during shot implementationworkbench.md §2–3 (example template/src/workbench.ts)
After final video delivery (open animation workbench)workbench.md (node workbench/scripts/open.mjs <project>; workbench own docs workbench/README.md)
After final video delivery (CapCut project export)jianying-export.md + jianying-export/ platform module

Asset Usage

  • assets/lib/ components: Copy into new project and modify freely (do not import from this repo). List: PageCam (2.5D page camera—the foundation of all "real page" shots), ClipCard (wraps external mp4 video footage as a "card protagonist" that can be driven by shot card motion skeletons, includes OffthreadVideo crossfade looping—seamlessly continues playback when footage is shorter than the shot), DigitRoll, FlashCut, Caption, FlatPanel, VerticalTicker (3D infinite scrolling wall), helpers(rand/shake/camera/motion). FlatPanel and helpers/camera require three + @react-three/fiber + @remotion/three dependencies; the rest only need remotion.
  • assets/scripts/capture-template.mjs: Copy and modify the CONFIG at the top (BASE/routes/selectors).
  • assets/audio/ sound effects: Copy and use directly (free commercial license, see audio/ATTRIBUTION.md): audio/bgm/ contains strongly rhythmic BGM options; audio/sfx/<category>/ 149 sound effects organized into 16 categories by scenario (transition, impact, riser, camera, ui, text, paper, film, light, data, scifi, mech, glass, fluid, crowd, counter). When searching for sounds, first enter the category directory; the inventory is listed in sound-design.md. The sparkle vocabulary is in the light/ directory (no sparkle/ directory); S1 prohibits sound color, not action—if the frame actually shows a click or switch, foley it, but half of ui/ is synthetic feedback sounds (tone/bleep/ notification) that must be auditioned individually and cannot be batch-approved (list in sound-design 3.3); long samples and light audio assets each have lists requiring special handling (sound-design 4.1).
  • demos/ implementation source code for each card: Mostly self-contained grayscale demos (some import demos/_fixtures/Fixtures.tsx for mock UI scene components, a few import demos/_textures/ for real page textures); copy into a Remotion project and run; a few demos use @remotion/motion-blur (CameraMotionBlur), requiring npm i @remotion/motion-blur; list in demos/README.md.
  • template/ complete renderable project: npm install && npx remotion render src/index.ts AiflPromo out/promo.mp4.
  • Testing (automated validation in repo, run after adding demos/components):
    • Pure function unit tests: npm test (repo root vitest, covers assets/lib/helpers mulberry32 / velocityAt / lagged / dampedSettle / handheld, deterministic assertions).
    • Demo render smoke test: python3 assets/scripts/smoke-render-demos.py renders the first frame of each duration-exported demo, asserting no crash (requires cd template && npm ci + motion-blur; see demos/README.md Testing and Verification section).
    • CI (pr-checks.yml) already runs: tsc strict compilation of all demos + vitest + smoke render.
  • jianying-export/ CapCut project export module: mac_draft.py (tested on Mac CapCut 11.2), windows_draft.py (implemented per upstream support path, not validated on actual hardware), smoke_test.py (minimal smoke test to run first in a new environment). Workflow, timeline extraction, and track creation methods in references/jianying-export.md; requires venv + pip install pyJianYingDraft.
  • gallery/ static gallery: Prefer directing the user to the online version https://vincentwei1021.github.io/video-shotcraft/library.html ; to run locally, first gallery/fetch-media.sh to pull sample clips (mp4s not in git), then cd gallery && python3 -m http.server 4178. 157 cards with 214 dynamic sample clips available for browsing/searching/multi-select copying of card names—ideal for letting the user choose shots while viewing samples.

相关技能