Video skills
videoclaw ships a curated video skill library — a layered set of agent skills that turn a plain-English request ("make me a presenter video", "clone this ad", "storyboard this brief") into the exact vclaw video ... commands and artifacts that produce a finished, identity-locked video. The library is not a flat bag of equally-preferred entry points: it is organized into two canonical front doors, a ring of specialists that sit underneath them, and a handful of compatibility aliases and reference surfaces.
Start here, every time
There are exactly two canonical entry points. Begin at one of them and let it route you — specialize only when the mode is already obvious.
video-framework— any generic video request (copy, create, narrate, film, UGC).brand-presenter— any narrated, host-led / slide-deck presenter video.
How the layers fit together
Diagram source (live Mermaid)
The flow is always the same: canonical front door → specialist (only when the mode is known) → concrete vclaw video commands and machine-readable artifacts. Aliases and the deprecated reference surface stay discoverable but are never first-choice.
42
video skills
category video in catalog.json
2
canonical entry points
video-framework, brand-presenter
35
specialists
one production lane each
4
compatibility aliases
over brand-presenter
1
deprecated reference
video-replicator
Canonical entry points
These are the two surfaces an agent or human should reach for first. They route broad intent into the right specialist so you never have to memorize the catalog.
| Skill | Role | What it does |
|---|---|---|
video-framework | canonical-entry | Unified front door for creating videos — routes across copy, create, narrated, presentation, long-form, film, and UGC workflows, reusing proven engines behind adapters. Start here for any generic video request. |
brand-presenter | canonical-entry | Generic narrated presenter-video workflow: turn a slide deck or structured topic into an intro/slides/outro presentation using a branded host profile. Start here for any host-led presenter video. |
Specialists
Reach for a specialist only once the mode is clearly known. Each one specializes video-framework (or, for the presenter variants, brand-presenter) and owns a single production lane.
| Skill | Role | Specializes | What it does |
|---|---|---|---|
ai-director | specialist | video-framework | Act as an AI Animation Director — run the staged director system to lock a project-level visual blueprint (visual identity, master color system, lighting grammar, per-character blueprint, environment blueprint, camera bible, performance rules) and persist it as project-blueprint.json so every filmmaking-prompts scene packet inherits it. |
3d-animation-short | specialist | video-framework | A one-line story idea → a finished stylized 3D animated short: story spine, labelled cast, no-people plates, a per-second shot table gated by six automated checks before anything renders. |
ai-film-director | specialist | video-framework | Platform-aware scene director: a minimal idea (plus character / first-frame / last-frame references) → a complete production prompt for Seedance or Google Veo, with continuous multi-scene handoff. |
brand-explainer | specialist | mograph | Narrated 60–70 s brand explainer films in the house motion-graphics style: one shared engine + a .brand data pack per film. |
garden-days | specialist | video-framework | Baby Ram & Sita "Garden Days" three-lane series factory: status, new episodes, driver ops, troubleshooting, franchise cloning. |
mograph | specialist | video-framework | Style-locked motion graphics: one motion-sheet style authority + action-only motion-pack blocks, rendered through the batch queue. |
rap-avatar-mv | specialist | rhyme-factory | A portrait + a supplied or generated track → rap MV variants: measured vocal windows, image and voice references, clip reuse, timing checks and verified delivery. |
rhyme-factory | specialist | garden-days | Character pack + nursery-rhyme idea → a YouTube-ready SUNG music video (song, sing-along captions, polish, thumbnail, upload pack) on top of garden-days. |
runway | specialist | video-framework | Draft renders on Runway through its free queue; paid credits stay opt-in. |
story-bible-builder | specialist | ai-director | Interview-driven canon builder: a story world → one dense story bible that ai-director and the storyboard consume. |
dance-breakdown | specialist | video-framework | A reference dance video → a measured per-shot choreography breakdown (scene detection + each shot read by a multimodal model). |
match-highlights-local | specialist | video-framework | Cricket highlights read straight off the broadcast scoreboard overlay — no model, no network, no spend; the zero-cost sibling of vclaw video match-highlights. |
vox-explainer | specialist | video-framework | Vox-inspired collage explainers from an approved narration and real archival photos: public domain / CC0 / CC BY / CC BY-SA only from Wikimedia Commons and the Library of Congress, object-capable cutouts, a contact sheet you approve before anything renders. |
ad-intel | specialist | video-framework | Competitor-ad teardown with our own tools: the public Meta Ad Library via the Higgsfield browser session → whisper transcript → Gemini reading → one artifact per ad and a brand teardown. |
ai-filmmaking | specialist | video-framework | Generate production-grade prompts the videoclaw way: an 8-shot Character Sheet to lock identity, a 9-panel Cinematic Storyboard Grid with CAM/MOVE/MOOD annotation strips to lock the scene, and Seedance 2.0 video prompts in three reference variants to render each shot — the chained character-sheet → storyboard-grid → seedance-packets workflow. |
brand-agency | specialist | video-framework | Full creative-agency brand launch inside videoclaw: from a product photo or website, lock a brand system (palette/voice/typography) and produce the 10-image + 5-video launch package — orchestrating go-bananas, gen-image --backend flow, provider routes, and the preview portal, persisting the system via vclaw video brand-definition. |
character-ad | specialist | video-framework | Build a short character-led ad the native Google Flow way: register the cast (and product) as Flow Characters once, generate each scene straight from those references with native voice (R2V), then stitch with cross-dissolves and a music bed under the voice — tighter identity than keyframe→i2v, one render per scene. |
motion-reel | specialist | video-framework | Turn a talking-head video into a polished motion-graphics reel — synced lower-thirds, anchor-word headlines, and concept icons/infographics driven by what is being said. Renders a finished reel locally and emits a Flow hand-off pack for the premium Omni-animated version. |
video-storyboard | specialist | video-framework | Turn a brief or clone plan into an explicit scene-by-scene storyboard artifact with optional character-to-scene bindings. |
video-analyze-template | specialist | video-framework | Turn a reference video into a reusable template packet before clone or storyboard work begins (Gemini auto-mode). |
video-clone-ad | specialist | video-framework | Adapt a saved template or known style to a new product, offer, or brand while preserving the execution structure (clone-execute). |
movie-director | specialist | video-framework | Create any short film via Director mode — interview-driven, auto, or CLI-hybrid; 12 genres, cast building, Seedance-safe prompts with content-filter auto-fix, and the storyboard-review gate that never burns credits until approved. |
multi-shot-prompt | specialist | video-framework | Generate timed multi-shot cinematic prompts from a reference image, validated against a videoclaw preset (cinematic-15s, seedance-10s, veo-8s, runway-10s) for Seedance / Veo / Runway / Kling / Sora. |
ugc | specialist | video-framework | UGC Campaign Generator — belief-driven (E5 method) pipeline that turns a product URL into N targeted 30–60s ad videos with subtitles plus a campaign report. |
higgsfield-generate | specialist (external bridge) | video-framework | Optional bridge over Higgsfield's MIT-licensed public skills and CLI: Marketing Studio ad videos, product photoshoots, Soul identity training, image-to-video, and virality scoring. |
seedance-prompts | specialist | — | Browse the clean-room Seedance prompt reference library and apply current provider guidance (formulas, examples, structure) to Seedance-targeted scene writing. |
character-creator | specialist | — | Create Go Bananas characters with profile images and multi-view reference sheets for consistent characters across scenes. |
character-library | specialist | character-creator | Hygiene/repair companion to character-creator: audit polluted entries, list the library, patch base prompts in place, and delete bad anchors. |
video-production-handoff | specialist | video-framework | Handoff: move a project from create/storyboard/assets into Review UI, review-autopilot, status, doctor-project, and publish readiness with canonical review-report.json truth. |
video-review-ui-qa | specialist | video-framework | Review UI QA lane: validate the local Review UI, storyboard handoff, review-autopilot, browser/mobile/desktop behavior, and CLI alignment for publish readiness. |
video-portfolio-ops | specialist | video-framework | Portfolio operations: manage multiple projects, report blocked work, export CSV/Obsidian dashboards, assign metadata, and check portfolio-level review/publish readiness. |
video-release-readiness | specialist | video-framework | Release-readiness lane: prepare a source checkout, package dry-run, npm release handoff, and verification report for the CLI, Review UI, skills, docs, and artifact boundaries. |
video-post | specialist | video-framework | Post-production: verify final renders, create social/platform variants, extract thumbnails, build loop-friendly cuts, and archive finished projects. |
video-thumbnail-lab | specialist | video-framework | Turn a final render into a click-driving thumbnail plus square/vertical platform variants and packaging for YouTube, Shorts, and social. |
youtube-audio | specialist | — | Download audio (MP3) or video (MP4) from YouTube videos or playlists — trimming, resolution, batch URLs (requires yt-dlp + FFmpeg). |
Identity-locked characters
For locked characters, the proven path is character sheets → keyframe → image-to-video via character-creator / character-library feeding the Seedance routes — never raw text-to-video. See the project docs on the Asset Library for the 1080p identity-locked recipe.
Compatibility aliases
Personal and brand-specific presenter skills are kept as thin aliases over the shared brand-presenter surface so the repo stays generic while preserving discoverability. They all resolve to the same canonical presenter engine.
| Skill | Role | Alias of | What it does |
|---|---|---|---|
davendra-presenter | compatibility-alias | brand-presenter | Narrated video from a PDF slide deck featuring Davendra, a 3D Pixar-style executive presenter. |
nex-presenter | compatibility-alias | brand-presenter | Narrated video from a PDF slide deck featuring Nex, a 3D Pixar-style tech commentator. |
bunty | compatibility-alias | brand-presenter | "Match Day Analysis with Bunty" — a cricket scorecard or play-cricket URL becomes a narrated video with Bunty, the cartoon Indian commentator in an orange blazer. |
david-sales-presenter | compatibility-alias | brand-presenter | Narrated video from a PDF slide deck featuring David Sales; delegates into brand-presenter. |
Reference (do not trigger)
| Skill | Role | Superseded by | What it does |
|---|---|---|---|
video-replicator | reference (deprecated) | video-framework | Reference-only documentation of the legacy 7-mode replicator surface. Do not use to trigger work — it is preserved as historical context; use video-framework instead. |
Use the canonical front door
video-replicator is kept purely as historical documentation. New work for the modes it described now flows through video-framework.
Complete count
This page lists all 42 video skills whose category is video in skills/catalog.json: 2 canonical entry points (video-framework, brand-presenter), 35 specialists, 4 compatibility aliases, and 1 deprecated reference surface.