ClipWisp
Get the package
Inside the skills package

Find the skill
for your next project.

29 entries in library 0.2.1: 26 workflow and support entries, two selectors, and one YouTube compatibility alias.

Supporting entries help other workflows. The nine written formats belong to one repurposing skill.

This catalog describes library version 0.2.1, the current package release.

The landing page shows 18 real files these skills produced: launch videos, clips, animated carousels, YouTube thumbnails, and edits. See the examples.

Get the skills package →

Start here

Content workflow selector

creator-content-router

Choose the correct workflow from the combined creator and ClipWisp library.

Outputs and requirements

Outputs

  • The selected workflow and its requested deliverables.

Required setup

  • An agent with access to this skill library and the current project.

Scope

  • This selector does not render media itself.
  • Each selected workflow has separate runtime requirements.

ClipWisp workflow selector

clipwisp

Select transcription, clip creation, or written repurposing from one bundled local engine.

Outputs and requirements

Outputs

  • Only the selected ClipWisp workflow's deliverables.

Required setup

  • Python 3.10 or newer and the current agent's model connection.
  • The engine ships inside scripts/engine with its helpers, font, and license.

Scope

  • The local creator pack has three workflows, plus this selector.
  • No ClipWisp account or separate model API key is required.
  • The existing agent connection can still have usage costs and limits.

Creator workflows

Talking-head video editing

clean-my-talking-head-video

Remove mistakes and restarts, improve dialogue, and apply the accepted Silpho video presets.

Outputs and requirements

Outputs

  • An edit proposal, timed cut decisions, preview, and verified final video.
  • Approved graphics and a recoverable cleanup record when cleanup runs.

Required setup

  • FFmpeg, ffprobe, Python, and a transcription tool.
  • HyperFrames supports the saved animated graphics workflow.

Scope

  • Five presets cover clean footage, presenter, hook presenter, landscape overlays, and Screen Studio footage.
  • The personal default has no subtitles, music, or voice pitch changes.
  • Green belongs to video backgrounds. Yellow is the only accent inside video graphics.
  • Run cleanup only after verification. Keep the original recording outside the edit directory.

Conversational video editing

video-use

Transcribe footage, plan cuts in plain language, and render a clean edit with FFmpeg helpers.

Outputs and requirements

Outputs

  • A word-level transcript and packed takes file.
  • An edit decision list and a rendered video.

Required setup

  • Python 3.10 or newer with the packages in pyproject.toml.
  • FFmpeg and ffprobe.
  • An ElevenLabs API key for Scribe transcription.

Scope

  • Transcription sends audio to ElevenLabs.
  • Animated overlays need HyperFrames, Remotion, or Manim installed separately.

YouTube publication package

package-youtube-video

Build a supported title and thumbnail pair, then prepare the complete YouTube publication text.

Outputs and requirements

Outputs

  • A recommended title and thumbnail pair with four alternatives.
  • An approved thumbnail, description, chapters, and verification notes.

Required setup

  • Source access, frame inspection, and web tools for current facts.
  • An image tool with reference-image support for thumbnail generation.

Scope

  • The current version retains ClipWisp's workflow and adds pose and anatomy checks.
  • Three saved profiles cover personal light, personal dark, and the red-and-black two-host podcast style.
  • Other creators use a custom profile. There is no global personal default.
  • The two personal channel URLs are not configured yet. Use explicit preset IDs.
  • The skill does not supply an image model or upload the video.
  • Thumbnail performance remains unknown without audience testing.

Compatibility

YouTube package (alternate name)

creator-youtube-package

The same YouTube workflow as package-youtube-video, under a second name.

Outputs and requirements

Outputs

  • The same title, thumbnail, description, chapters, and verification notes.

Required setup

  • The same source and image tools as package-youtube-video.

Scope

  • This is a compatibility alias, not an additional content capability.
  • The current creator router selects this copy.

Marketing assets

App store screenshot editor

app-store-screenshots

Build a local screenshot editor with device frames, editable copy, and export layouts for mobile stores.

Outputs and requirements

Outputs

  • An editable Next.js screenshot project and exported marketing images.

Required setup

  • Node.js, a package manager, and a browser.
  • The reusable editor template ships with the skill.

Scope

  • This editor consumes app captures. It does not independently prove product features.
  • App-specific copy, images, and store sizes need review before submission.

Landing page design

landing-page-design

Turn an offer and supporting evidence into a clear landing page structure, copy, and design direction.

Outputs and requirements

Outputs

  • A page structure, copy, layout decisions, and implementation guidance or page changes within the requested project.

Required setup

  • The current agent and access to the website project when implementation is requested.

Scope

  • Conversion results require real audience data.
  • Use only supported testimonials, numbers, guarantees, and product claims.

Research and writing

Recent topic research

last30days

Research recent public discussions and source activity to support content planning.

Outputs and requirements

Outputs

  • A sourced research report with platform coverage and stated gaps.

Required setup

  • The skill's Python runtime and available source tools.
  • Some providers need their own credentials or paid access.

Scope

  • Source availability changes with platform access and provider settings.
  • This bundle contains instructions and scripts, not provider accounts or private source data.

Hormozi transcript research

ask-hormozi

Answer a business question from an installed MoreMozi transcript index with video timestamp citations.

Outputs and requirements

Outputs

  • An attributed answer with source excerpts or paraphrases and clickable timestamps.

Required setup

  • The external ask-hormozi runtime or HQ integration.
  • The transcript corpus and search index are not included.

Scope

  • Copying this skill alone does not install its corpus or search command.
  • Do not invent quotes or treat search gaps as evidence of an opinion.

Writing cleanup

unslop

Remove repetitive, inflated, or generic wording while preserving the writer's meaning.

Outputs and requirements

Outputs

  • A revised draft with clearer wording and fewer repeated patterns.

Required setup

  • The current agent session. No external program is required.

Scope

  • Editing style does not verify the draft's facts.
  • The user's wording and accessibility requirements take priority over stylistic preferences.

Video toolkit

HyperFrames workflow selector

hyperframes

Select and coordinate the HyperFrames workflow for an editable HTML video composition.

Outputs and requirements

Outputs

  • A chosen production workflow, editable composition, preview, and requested render.

Required setup

  • Node.js 22 or newer, HyperFrames CLI, FFmpeg, and browser rendering support.

Scope

  • The toolkit refers to optional upstream workflows that are not all installed here.
  • Check availability before selecting an unbundled skill or provider.

Animation and scene transitions

hyperframes-animation

Choose and implement animation timing, scene transitions, and supported motion adapters.

Outputs and requirements

Outputs

  • Editable animation sequences and scene transitions.

Required setup

  • HyperFrames and the selected animation adapter.

Scope

  • Optional adapters can require extra packages or skills.
  • Motion must follow the current project's style and accessibility needs.

HyperFrames audio mixing

hyperframes-audio

Mix placed dialogue, music, and effects with fades, gain automation, ducking, and supported audio processing.

Outputs and requirements

Outputs

  • Editable audio levels, fades, effect chains, or grouped mix controls.

Required setup

  • HyperFrames and the source audio files.

Scope

  • This skill mixes existing audio. It does not find or generate the source assets.
  • Noise processing can damage speech when applied too strongly.

HyperFrames previews and renders

hyperframes-cli

Operate the CLI for project checks, snapshots, previews, local renders, and supported remote render routes.

Outputs and requirements

Outputs

  • Check reports, preview links, snapshots, or encoded video files.

Required setup

  • Node.js 22 or newer, HyperFrames, and FFmpeg.
  • Remote render routes need the relevant accounts and configuration.

Scope

  • A successful file check does not prove visual or editorial quality.
  • Hosted rendering and provider charges are separate from this library.

Composition structure and timing

hyperframes-core

Define the renderable HTML structure, tracks, clips, source times, and media placement.

Outputs and requirements

Outputs

  • Correct composition markup and timing relationships.

Required setup

  • HyperFrames and local source assets.

Scope

  • Use hyperframes-keyframes for visual camera motion and hyperframes-audio for mix changes.

Video design direction

hyperframes-creative

Plan palettes, typography, composition, narrative structure, and visual hierarchy for HyperFrames projects.

Outputs and requirements

Outputs

  • A scene plan and consistent design choices tied to the message.

Required setup

  • The current agent and the project assets.
  • Rendering uses the separate HyperFrames runtime.

Scope

  • Creative direction does not supply footage or prove a product claim.
  • Use the user's accepted preset before generic examples.

Zooms and camera motion

hyperframes-keyframes

Author predictable punch-ins, reframing, pans, and other visual property animation.

Outputs and requirements

Outputs

  • Editable keyframes and checked camera motion.

Required setup

  • HyperFrames and an existing composition.

Scope

  • Keyframes do not repair cut timing. Use hyperframes-core for trims or source-range changes.

Reusable motion components

hyperframes-registry

Find and install existing HyperFrames components for a named effect or visual treatment.

Outputs and requirements

Outputs

  • A selected component with the required assets and project wiring.

Required setup

  • HyperFrames CLI and access to its component catalog.

Scope

  • A catalog result still needs adaptation and visual review.
  • Some components require optional dependencies.

Custom HyperFrames videos

general-video

Build longer or custom HTML video compositions when a specialized workflow does not fit.

Outputs and requirements

Outputs

  • An editable multi-scene composition and requested preview or video.

Required setup

  • HyperFrames and its core, creative, animation, and media support skills.

Scope

  • This is a composition workflow, not an automatic substitute for footage editing judgment.
  • Optional Figma or media providers require their own available tools.

Product launch videos

product-launch-video

Turn a product URL, script, or brief into a launch or promo video with a storyboard, voice, music, and designed frames.

Outputs and requirements

Outputs

  • An editable HyperFrames project with a storyboard, a script, and one composition per frame.
  • A rendered 16:9 MP4 by default.
  • A 9:16 reel or 1:1 version when the storyboard format is set to portrait or square. Each format is a separate build.

Required setup

  • Node.js 22 or newer, HyperFrames CLI, FFmpeg, and the bundled HyperFrames and media-use skills.
  • Web access to capture a URL. A HeyGen account or the offline Kokoro voice for narration.

Scope

  • Music lookup needs a HeyGen account. Without one, the video has no background music.
  • Figma sources need a separate Figma skill that this library does not include.
  • Its cross-workflow tests expect upstream HyperFrames workflows that are not bundled here.

Designed video overlays

talking-head-recut

Add timed information graphics, titles, side panels, and picture-in-picture elements to existing talking footage.

Outputs and requirements

Outputs

  • An editable HyperFrames composition with synchronized graphic overlays.

Required setup

  • HyperFrames, FFmpeg, source access, and the bundled design assets.

Scope

  • Graphic overlays differ from word-by-word subtitles.
  • Paweł's current preset rules override the toolkit's generic design examples.

Media services

Media sourcing and preparation

media-use

Find, generate, or prepare images, footage, audio, logos, fonts, and treatments for a HyperFrames project.

Outputs and requirements

Outputs

  • Local adopted assets with provenance and instructions for project use.

Required setup

  • Node.js and the relevant local or hosted provider tools.
  • Authentication and usage charges depend on the selected provider.

Scope

  • This library does not include provider accounts or a stock-media subscription.
  • Optional provider access can fail even when the skill files are installed.

Generated music

music

Create music with ElevenLabs from a prompt or a structured composition plan.

Outputs and requirements

Outputs

  • Generated music audio and available generation metadata.

Required setup

  • Internet access, the ElevenLabs SDK or supported API client, and an ElevenLabs API key.

Scope

  • Provider usage and terms are separate from this bundle.
  • Paweł's video cleanup preset does not add music by default.

Generated sound effects

sound-effects

Create a non-speech sound effect or ambience with ElevenLabs.

Outputs and requirements

Outputs

  • Generated audio ready for review and placement.

Required setup

  • Internet access, the ElevenLabs SDK or supported API client, and an ElevenLabs API key.

Scope

  • This is an optional paid provider route.
  • The personal Silpho preset prefers local or procedural transition sounds by default.

ClipWisp workflows

ClipWisp content repurposing

clipwisp-repurpose

Analyze a source once and create selected written formats with linked source evidence.

Outputs and requirements

Outputs

  • A reusable source analysis, structured content package, and exported Markdown drafts.
  • LinkedIn posts, X posts, X threads, articles, newsletters, show notes, scripts, carousel copy, or infographic briefs.

Required setup

  • Python 3.10 or newer and the bundled ClipWisp engine.
  • The agent supplies writing through its existing connection. Existing transcripts need no media tools.

Scope

  • Nine formats are options inside one skill.
  • Carousel and infographic outputs are text briefs, not rendered visuals.
  • Untimed text cannot establish chapter timestamps.
  • The helper checks structure and evidence references. The agent still checks meaning.

ClipWisp clip creation

clipwisp-clips

Find complete moments in a recording and render clips with selected framing and optional captions.

Outputs and requirements

Outputs

  • Reviewed clip candidates with source times and rendered H.264/AAC MP4 files.

Required setup

  • Python 3.10 or newer, ffprobe, and FFmpeg with libx264 and the subtitles filter.
  • The bundled DejaVu caption font. Optional yt-dlp supports online sources.

Scope

  • Portrait output is 1080 × 1920. Landscape output is 1920 × 1080. Both use 30 fps.
  • Framing supports fit_blur, crop_fill, and horizontal. It does not track the speaker.
  • Omitting captions still requires FFmpeg subtitle support in the current renderer.
  • Clip scores express editorial judgment, not predicted views.

ClipWisp local transcription

clipwisp-transcribe

Create a timed transcript from local speech with segment and word timestamps.

Outputs and requirements

Outputs

  • Transcript JSON with text, detected language, segments, and word timestamps.

Required setup

  • Python 3.10 or newer, ffprobe, and faster-whisper.
  • A downloaded speech model or an existing local model directory.

Scope

  • The default model is base and runs on the CPU.
  • First use can download model files.
  • Transcription can mishear names and does not replace a manual accuracy check.

Start with the tools you have.

Writing from an existing transcript needs fewer programs than video rendering. Check the requirements for your selected workflow.

Read the setup guide →