Repository navigation
Clip art
The clip art library keeps pictures that are free to use: old engravings and prints, drawings, emoji and icons. They're for covers, the weekly issue, story cuts and explainers. It is the still-picture partner of the b-roll library. Each picture keeps where it came from, its credit line and its licence.
-
commons: Wikimedia Commons. Drawings (SVG files: clip art, icons, diagrams) come first. Only when a find still wants more does it search pictures (photos and scans), which are mostly photos and often need a credit;[clipart] commons_pictures = falsekeeps it to drawings. Add "engraving" to a search for old prints. An SVG, or a picture wider than[clipart] max_px, comes as Commons' own render at about that width (Commons rounds it to one of its standard sizes; SVGs come as PNG, with their transparent background). No key. -
met: The Metropolitan Museum of Art's open access: prints, trade cards, ads and drawings the Met marks public domain, which it releases as CC0. No key. -
cleveland: the Cleveland Museum of Art's open access, only works it marks CC0, at 3400 pixels. No key. -
iconify: emoji and icons through Iconify (api.iconify.design). By default the sets are[clipart] iconify_sets: Fluent Emoji (Microsoft's 3D-look emoji, and their flat version), Noto Emoji (Google), Tabler and Phosphor (arrows, shapes, symbols). Files are SVG. Each set's licence comes from its own record. A logo or brand set is never used, whatever its licence says, because logos are trademarks. Withiconify_sets = []every set whose licence is in[clipart] licencesis searched. Each SVG is sized[clipart] max_pxtall, so a renderer doesn't draw it 16 pixels high. No key. -
pixabay: Pixabay's vectors ([clipart] pixabay_type, orillustrationorall) under the Pixabay Content License. It needs a free key: sign up at pixabay.com, copy your key from pixabay.com/api/docs intosecrets/pixabay_secret.txt([paths] pixabay_key), or setPIXABAY_API_KEY(b-roll uses the same key for Pixabay's video). The key is never written toclipart.jsonor printed. Without a key, Pixabay is skipped. Pictures come at most 1280 pixels wide, unless Pixabay grants your key full access.pixabay_transparent = truekeeps only pictures with a transparent background.
The Art Institute of Chicago's image server refuses scripts, so it isn't a source. Openclipart needs an account's key for its API and may come later. Until then, Openclipart's art is on Pixabay under the user OpenClipart-Vectors.
-
clippa clipart find "cow engraving": searches every source in[clipart] sourcesand adds up to[clipart] per_sourcenew pictures from each, tagged with the search.--source metsearches one source;-n 10takes more. -
clippa clipart list [--tag cow] [--source iconify] [--status new|kept] [--json]: the library, newest first, with each picture's size, credit and licence.--jsonadds each file's path. -
clippa clipart add <address|set:icon|file|folder> [--tag X]: one picture. It can be:- a Commons
File:page, a metmuseum.org/art/collection/search, clevelandart.org/art or pixabay.com page; - an Iconify name such as
fluent-emoji:cow-face(from icon-sets.iconify.design, whose addresses work too); - your own picture file, or a folder of them with all its subfolders. That needs
--creditand usually--licence.
A downloaded pack goes in whole, for example the CC0 packs from kenney.nl (speech bubbles, emotes):
clippa clipart add Downloads/kenney_emotes-pack --credit "Kenney (kenney.nl)" --licence CC0 --tag emote. - a Commons
-
clippa clipart keep|drop <id> ...: a dropped picture's file is deleted, and the next find doesn't bring its source back. Adding it by hand brings it back.
The files and clipart.json are in [paths] clipart. A file is <id>.<ext>: .png, .jpg,
.gif, .webp or .svg. A museum original over [clipart] max_mb falls back to the
museum's web-size copy.
A picture is kept only if all three of these hold, and otherwise it's skipped with the reason printed:
- Its licence is in
[clipart] licences. By default that ispd(public domain),cc0,permissive(MIT, Apache, ISC, BSD and OFL) andpixabay. - Nothing in its title, description or credit names an outside owner (©, "copyright", "all rights reserved", Getty ...).
- Commons doesn't mark it trademarked or as a person's likeness.
Noncommercial, no-derivatives and GPL licences are never kept.
-
Credits. None of the default licences needs a credit when you use a picture. The credit
line is kept anyway; it's good manners in a description. Adding
cc-bytolicenceskeeps attribution-licensed art too, such as Twemoji, game-icons.net and many Commons files. Then every video or post that uses one must credit its author.cc-by-sa(OpenMoji, a lot of Commons) also makes what you build from it share-alike, a cover included. Leave it out. - People and brands. Pixabay's licence bans using a picture that shows a brand or a logo commercially, and showing a recognisable person in a bad light. Use Pixabay for objects and drawings, never as the face of a cover.
- Old characters. A public-domain comic or ad can show a character someone still owns as a trademark. Don't put a recognisable mascot on a cover.
-
Permissive licences (Fluent Emoji MIT, Noto Apache): the licence notice belongs beside
copies you hand on. The library keeps the set's name and licence in
clipart.json, and a video needs no credit.
Some sources are good but have no API, or don't want scripts; for these, download what you
pick and clippa clipart add the file:
- Old Book Illustrations, engravings on white, ideal for cut-outs.
- The Public Domain Review's image archive.
- The Library of Congress's "Free to use and reuse" sets (old ads, WPA posters). Its API allows only 20 requests a minute.
- LottieFiles, for free animations.
- Kenney's CC0 packs.
Avoid "free PNG" sites (PNGWing, StickPNG, CleanPNG) and GIF sites (Giphy): they hold no rights to what they serve.
[clipart] in clippa/defaults.toml: sources, per_source, licences, commons_pictures, max_px,
max_mb, iconify_sets, pixabay_type, pixabay_transparent, timeout_s. clippa doctor
checks the source and licence names, warns when Pixabay has no key, and counts the library.
Home · Getting started · clippa doctor · These pages are built from docs/wiki/ in the repo, so edits made here are overwritten: how to edit them
Run it
- Overview
- Getting started
- Roles and services
- Configuration
- Pipeline
- Review
- Reviewing
- War room
- Clippa messenger
- Scheduling and uploads
- Costs
- Checkup
- Audit
- Backups
Install
- Install
- Install checklist
- Kitsu
- Discord app
- YouTube sign-in
- Windows
- macOS
- Transcribe on Modal
- NAS recorder
- NAS hub
- Claude on the NAS
- Phone access
- Model providers
- Adding a show
- YouTube channel
Features
- Pre-shows
- Fetching VODs
- Speakers
- Faces
- Covers
- Cast library
- B-roll
- Clip art
- Weekly issue
- X posts
- TikTok
- News watcher
- Character pieces
- Documentaries
Research
Working on Clippa