color plugins

by Hunter D

Framework-agnostic design principles for visual perception and hierarchy: how the eye groups elements, where it lands first, and what makes a composition read as ordered. Covers Hierarchy, Proximity, Signal-to-Noise Ratio, Color, Alignment, Legibility, Readability, Gestalt Similarity, Closure, and Figure-Ground Relationship.

by HKTITAN

Design engineering for AI agents — animation timing, transitions, typography, color, components, layout, accessibility, anti-patterns, sound, SVG, launch films. Plus a SOUL.md identity layer and ten workflow subagents. Distills Emil Kowalski, Benji Taylor, Jakub Antalik (transitions.dev), guidelines.sh, Vercel, Ben DC, DiceBear, lucide-animated, Google Labs design.md, Agentation.

by vikash ruhil

Loomwright is a plan-first, multi-agent development system for Claude Code. Launch Pad turns a goal into a reviewed brief; Supervisor executes it autonomously across parallel git worktrees with self-heal; dedicated agents cover code review, red-teaming, QA, and standalone PR review-and-heal. 14 agent roles, 24 slash commands, 42 skills, 43 quality gate hooks. Full history in CHANGELOG.md.

by akbun

Video editing skills — plan and run DaVinci Resolve vlog edits: speech-free travel vlogs (cut/stabilize, chronological timeline, color-grading workflow (Log LUT, exposure, white balance, contrast, saturation on labeled nodes), four separate creative looks on new LOOK nodes, and a city montage cut style, face mosaic, Gmarket Sans subtitles, audio mix, 4K delivery, vertical YouTube Shorts, and a YouTube hook finder that proposes cold-open/intro candidates and builds an approved hook timeline) and faceless talking developer vlogs (transcript-driven storytelling, beat-ordered assembly, sentence-level cuts, dialogue/keyword subtitles, HyperFrames graphic cards, Epidemic Sound SFX and BGM selection in essay/project/reflection styles, plus a pre-shoot guide, shot sketches and a storyboard with OBS screen-recording list)

by ogxo

Audit color usage and implement a color token system, with contrast and palette calculations from a bundled script.

by Colton Bearden

Document craft with the carrel CLI: /redact, /sign, /form, /proof, /color, a document-clerk agent (redact → verify → sign manifest) and a redaction-and-provenance skill.

by SSFSKIM

A working UI-design practice for Claude: read the project before touching it, derive one visual stance from the brief and commit to it in writing, build on semantic tokens, use real content, and check the result against an accessibility and taste floor before calling it done. Includes an aesthetic-ingredient sampler that pulls builds away from trained defaults, thirteen in-depth reference files, and an optional Essentialist designer persona distilled from the Bauhaus, Dieter Rams, and Jonathan Ive lineage. Ships a second, independent skill, the Materialist: the Liquid Glass aesthetic as a decision function and its laws, with the measured optics of the material and a cookbook for vitrea, its web implementation. Reconstructs the workflow method observed in Figma Make; not affiliated with Figma, Inc., and not a Figma integration.

by AkaraChen

Collaborative UI and visual design skills for audits, animation, and visual craft.

by Mike Sherrard

The AI Editing Studio: a full-time video editor inside the agent's Claude. 28 skills across five stages: DIRECT (creative direction, record coach, transcript read, section map), CUT (remove the fluff, hook finder, rescue cut, interview edit), POLISH (audio enhance, color and look, captions, music, B-roll, B-roll library, brand graphics, energy), REPURPOSE (short-form repurpose, weekly batch, podcast, trailer), SHIP (quality check, review and revise, publish), plus setup, a plain-English navigator that also resumes any stopped edit, and three full recipes (long-form, reel, listing). Claude edits directly in the agent's own Riverside account with precise, verifiable operations: no per-edit AI credits, every change proven by a revision diff, every visual checked on real frames, and publishing to YouTube, Shorts, TikTok, Instagram, Facebook, LinkedIn, and X only on an explicit yes. One owner per output: it edits video and never writes titles, descriptions, chapter names, post captions, or designs; those come from the YouTube System, the Short-Form System, and Claude Design. Requires Plugin 1 (the Brain) and a Riverside Grow plan or higher.

by groot99-droid

A local BM25 search catalog of UI styles, product palettes, font pairings, UX rules, icons, motion presets and stacks, with a contrast-checked design-system generator, a brief router, and multipart search and page-review agents.