Screen scoping from the brief
The workflow starts by scoping screens from the product brief before any color or typography choices are made, so the layout system begins with structure instead of guesswork.
Tastemaker is a workflow skill for coding agents that turns product context and visual references into UI screens, assets, contrast-checked color systems, and local design decisions.
Tastemaker is a workflow skill for coding agents that helps them build interfaces with a more consistent visual system. It takes product context and reference material, then turns that into screens, assets, checks, and remembered design decisions that the agent can reuse in later work.
Rather than asking the agent to “have better taste,” the product focuses on concrete steps: scope the screens, ground the palette, verify contrast, assemble assets, build the interface, and store the decisions locally. The homepage positions it for agents such as Claude Code, Windsurf, and Gemini CLI, with outputs that live beside the project instead of behind a hosted service.
The workflow starts by scoping screens from the product brief before any color or typography choices are made, so the layout system begins with structure instead of guesswork.
It can read reference pixels from source imagery or generate a new palette for the project mood, then checks all legal color pairings with contrast math.
Tastemaker outputs a coherent asset set: icons, imagery when needed, a constructed mark, and favicons, instead of emoji placeholders or partial visual systems.
The same pass can produce UI mockups, flows, charts, before-and-after states, and GSAP motion, so the screen is built as evidence rather than as descriptive copy.
Resolved project preferences persist in local files, with developer decisions carried forward as a profile for later work.
The site shows a plain-text, file-based output model using Markdown, Python, SVG, CSS, and JavaScript, which makes the result inspectable in the repository.
Use it when you want a coding agent to produce a landing page with a coherent palette, proper type pairing, real icons, and motion instead of a generic first draft.
Use it for product or app shells that need empty, loading, and populated states built from a consistent visual system.
Use it when a reference image, mood board, or existing brand direction needs to become actual tokens and assets the agent can build from.
Use it for comparison work where you want to inspect how the same prompt behaves before and after the skill is installed.
Use it when a project needs local style memory so later screens keep the same visual decisions instead of drifting between sessions.
Tastemaker is installed into a coding agent workflow with a command like `npx skills add codeswithroh/tastemaker`. The homepage also says it can trigger when you ask your agent to build, style, or improve a UI, without a special phrase or hosted API.
It is designed for coding agents such as Claude Code, Windsurf, and Gemini CLI. The homepage also mentions a Claude Code plugin marketplace and manual install path.
The source shows a free core and sponsor-exclusive Aesthetic Modes available for $8/month. A separate pricing page is not available on the site.
The workflow produces concrete project files and assets such as `.tastemaker/style-lock.md`, `decisions.log`, `profile.md`, `design/assets`, and implementation files for the actual screens.
It is aimed at UI work such as landing pages, app shells, product screens, and comparisons where the agent needs reference-based visuals, checked color contrast, assets, and motion.
Paper is a design tool that connects canvas, code, and AI agents so teams can create, share, and ship work in one workflow. Includes desktop app and MCP access.
Orca is an Agent Development Environment for shipping with coding agents, running multiple CLI agents in parallel across isolated worktrees, with desktop and mobile workflows.
AI Magicx is a unified AI workspace for chat, image, video, voice, music, email and developer tasks, helping teams and creators manage multiple models in one place.
blop is a QA agent that writes browser tests as code in your repo, runs them in CI, clusters repeated failures, and can open PRs to fix broken tests.
RLAMA is a local AI platform for building RAG systems and intelligent agents on macOS, Linux, and Windows, with HTTP API support.
Kastra authorization infrastructure for AI systems checks prompts, tool calls, shell commands, API requests, and browser actions before execution.