Skip to content

digitalsamba/claude-code-video-toolkit / .claude/skills/video-design

video-design

Verified

Visual design for videos - colour by act, type scale, chart and layout rules for 16:9 and 9:16 scenes, a design.md handoff that any design skill can fill, and a contact-sheet critique of the render. Use when designing or refining how a video looks (title cards, data scenes, slides, a whole video's palette), when a scene looks generic or busy, when choosing colours for a video, or when /design or /scene-review asks for visual work. Works on its own with a brand, or takes direction from a design source such as taste-skill, Anthropic's frontend-design, or the user's own design skill.

FileKindSize
SKILL.mdskill md9.4 KB
critique.mdmarkdown4.4 KB
handoff.mdmarkdown3.8 KB
idents.mdmarkdown5.4 KB
palettes.mdmarkdown3.4 KB
sources.mdmarkdown6.7 KB
# Marks and idents

A mark is the logo's shape. An ident is the mark moving, with a sound, at the start and the end of a video. This
file is the bar for both and the method that produced the toolkit's own. Making a 3D ident needs Blender, which
is optional; a mark and its sheet need nothing but the toolkit.

## Marks

**The bar.** A mark has to work where it will be seen, and most of those places are small:

- at 16 px (a favicon, a browser tab) and 32 px, where a thin gap closes and a small part disappears;
- on the brand's light ground, its dark ground and its own colour, which usually takes two or three versions
  (as drawn, on dark, one colour);
- inside a circle, because avatars are round and the corners of the square are cut off;
- in one colour at 70% in the corner of a video frame, as a watermark;
- as the last frame of an ident, so it has to be something that can stand still.

**The method.**

1. **Start from the subject, in words.** Write what the thing is in one line ("a terminal that makes video") and
   list what a viewer already associates with each half. Ideas come from joining the two lists, not from
   decorating a letter.
2. **Draw five directions, not one.** Flat SVG, by hand, one colour plus the brand colour at most. Give every
   shape an `id`: a 3D ident moves the parts by name. One of the five should be the obvious idea, so the others
   are measured against it.
3. **Run the sheet on each.** `uv run tools/mark_sheet.py --svg mark.svg --brand <name>` draws the sizes, the
   pixel zooms, the grounds, the circle and the watermark on one PNG and warns about contrast under 3:1, a part
   the circle cuts, and a part lost at 16 px. Add `--on-dark` and `--mono` once those versions exist.
4. **Say how each would move, in one sentence.** A mark with no move in it makes a weak ident. The toolkit's
   mark is a slash and a play triangle; its move is a clap, so the sound is the move.
5. **Critique each against the bar, rank them, recommend one.** The still at 16 px decides more often than the
   drawing at 512 px. The toolkit's first draft (a prompt chevron and a caret) ranked third because at rest it is
   the `>_` every terminal app has.
6. **Search for lookalikes before it is public.** An image search of the mark on white. Note the nearest shapes
   and anything in the same field with the same colours. That is a first pass, not a trademark search.

Keep the mark in `brands/<name>/assets/` as `mark.svg`, `mark-on-dark.svg` and `mark-white.svg`, with the same
shapes and ids in each.

## Idents

**The bar.**

- **One mark, one move, one sound.** The move and the sound do the work; the mark stays simple. If the move needs
  three sentences it is three moves.
- **Short.** An open of 1.5 to 2.5 s and a close of 3 to 5 s. The open must not delay the hook, so a Short gets
  the close only, or the short open.
- **Siblings.** The close is the open played back or completed, so the pair reads as one thing.
- **It works with the sound off.** The move alone has to say what happened.
- **It ends as the mark.** The last frame of the open is the avatar. Anything else on screen then is a mistake.
- **The brand's colours survive it.** A flash or a glow that turns the brand colour into another colour is not
  the brand, even for three frames at rest. At the hit, a brighter version of the same hue is the limit.
- **It is yours.** No other brand's mark, and no copy of a famous ident's move or sound.

**The method.**

1. **Animatics first.** Three ideas as rough drafts with a sound sketch each, shown side by side. Choose with
   the sound on. A choice made from stills picks the best still, not the best move.
2. **One family, not one ident.** Keep the move and the sound; vary the staging. The toolkit has the mark alone
   on a dark stage (everywhere, and the only one for Shorts) and the same clap on a keycap (full episodes).
3. **Build from a spec.** `tools/blender_render.py` with a `logo` prop and an `ident` shot (`docs/blender.md`,
   "A logo and an ident"; the `blender` skill has the verbs). The timing is the list of beats, so a change is a
   number, and the vertical version is the same spec at `--res 1080x1920`.
4. **Put the sound on the picture's biggest change.** That frame is what the eye takes as the hit. `render`
   checks it for every `hit`. With `tools/sync_check.py`, measure a hit with `--kind peak`: as an `impact` it
   reads late, because a hit sets off something bigger that stops later.
5. **Check the last frame as a still.** `check` compares it with the flat SVG: parts in place, extras gone, the
   lens square on, each colour within 10 Delta E.
6. **Leave room.** If a wordmark or a title may sit with the mark, move the mark in the frame with the shot's
   `offset` now, so adding the words later does not change the ident.
7. **Say what nobody has heard.** Levels and sync can be measured. Whether the sound is right cannot: a person
   listens before it is used.

8. **Put it in the brand's slot.** `brands/<name>/ident/` with an `ident.json` (lengths, the hit's frame, the
   frame a title may come in) and the flat mark in `brands/<name>/assets/mark.svg`; the format is in
   `docs/creating-brands.md`. A template that reads the slot plays the ident, and keeps its own 2D opening when
   the slot is empty, so a video never depends on the ident being there.

Without Blender, the same bar holds for a 2D ident in Remotion: the mark's shapes as SVG, one move on a `cue()`
frame, one sound on that frame.
canvas-designanthropics/skillsCreate beautiful visual art in .png and .pdf documents using design philosophy. You should use this skill when the user asks to create a poster, piece of art, design, or other static piece. Create original visual designs, never copying existing artists' work to avoid copyright violations.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
docxanthropics/skillsUse this skill whenever the user wants to create, read, edit, or manipulate Word documents (.docx files) or Word templates (.dotx files). Triggers include: any mention of 'Word doc', 'word document', '.docx', '.dotx', or requests to produce professional documents with formatting like tables of contents, headings, page numbers, or letterheads. Also use when extracting or reorganizing content from .docx or .dotx files, inserting or replacing images in documents, performing find-and-replace in Word files, working with tracked changes or comments, or converting content into a polished Word document. If the user asks for a 'report', 'memo', 'letter', 'template', or similar deliverable as a Word or .docx file, use this skill. Do NOT use for PDFs, spreadsheets, Google Docs, or general coding tasks unrelated to document generation.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
pdfanthropics/skillsUse this skill whenever the user wants to do anything with PDF files. This includes reading or extracting text/tables from PDFs, combining or merging multiple PDFs into one, splitting PDFs apart, rotating pages, adding watermarks, creating new PDFs, filling PDF forms, encrypting/decrypting PDFs, extracting images, and OCR on scanned PDFs to make them searchable. If the user mentions a .pdf file or asks to produce one, use this skill.Stars: 198kUpdated Claude CodeCodexVerifiedInstall
pptxanthropics/skillsUse this skill any time a .pptx or .potx file is involved in any way — as input, output, or both. This includes: creating slide decks, pitch decks, or presentations; reading, parsing, or extracting text from any .pptx or .potx file (even if the extracted content will be used elsewhere, like in an email or summary); editing, modifying, or updating existing presentations; combining or splitting slide files; working with templates (.potx), layouts, speaker notes, or comments. Trigger whenever the user mentions "deck," "slides," "presentation," or references a .pptx or .potx filename, regardless of what they plan to do with the content afterward. If a .pptx or .potx file needs to be opened, created, or touched, use this skill.Stars: 198kUpdated Claude CodeCodexVerifiedInstall