Skip to main content

ai-chat-interface-pro

Use when designing AI chat, copilots, agent workspaces, prompt consoles, and assistant surfaces. Pro version focused on conversation flow, tool state, citations, safety, and trust; emphasizes deep information architecture, strong hierarchy, explicit states, responsive behavior, accessibility, and validation checks.

インストールへ移動

ソース情報

リポジトリ
Mr-Q526/TeamCC-Platform
ソースの最終更新活動
2026年4月15日 03:16
検出された SKILL.md の言語
英語
スター
8
フォーク
1

インストール方法

デフォルトでは、最初にソースを確認する Prompt が選択されています。直接コマンドに切り替えるか、ローカルコピーをダウンロードすることもできます。

ソースファイルを確認

インストールを決める前に、SKILL.md と SkillsMP に表示されている付属ファイルをお読みください。

ファイルエクスプローラー
4 ファイル

SKILL.md を表示中

SKILL.md
ソースの指示 · 読み取り専用プレビュー
schemaVersion
2026-04-11T00:00:00.000Z
skillId
frontend/ai-chat-interface-pro
name
ai-chat-interface-pro
displayName
AI Chat Interface PRO
description
Use when designing AI chat, copilots, agent workspaces, prompt consoles, and assistant surfaces. Pro version focused on conversation flow, tool state, citations, safety, and trust; emphasizes deep information architecture, strong hierarchy, explicit states, responsive behavior, accessibility, and validation checks.
aliases
["ai-chat-interface-pro","AI Chat Interface PRO","aichatinterfacepro","智能助手","chat","interface"]
version
0.1.0
sourceHash
sha256:bd263863b08c5b1bbc8aeb60d92c8ed6338f9b7c8e82ff3dc06356296d6cb76f
domain
frontend
departmentTags
["frontend-platform"]
sceneTags
["design"]
# AI Chat Interface PRO Use this skill when the task is to design AI chat, copilots, agent workspaces, prompt consoles, and assistant surfaces. Goal: produce a high-quality, production-ready page or interface that supports conversation flow, tool state, citations, safety, and trust. ## Variant intent - Use the rigorous version when quality, accessibility, conversion, and maintainability matter. - Optimize for deep information architecture, strong hierarchy, explicit states, responsive behavior, accessibility, and validation checks. - Prefer concrete UI decisions over generic advice. ## Default workflow 1. Define the primary user, task, and success action. 2. Map the minimum page structure needed for the scenario. 3. Establish hierarchy: what must be understood first, second, and third. 4. Design responsive behavior for mobile and desktop. 5. Add states: loading, empty, error, success, disabled, and permission where relevant. 6. Run the final checks before delivery. ## Design rules - Use clear section names and user-facing copy. - Make the primary action visually dominant and repeat it only when it helps. - Keep navigation and secondary actions subordinate to the main task. - Use spacing, typography, and alignment before adding decorative containers. - Do not hide critical information behind hover-only interactions. - Preserve keyboard access, readable contrast, and touch targets. ## Pro guidance - Build a strong visual and interaction thesis before writing components. - Include proof, context, and state transitions when they affect trust. - Use component variants intentionally: default, hover, active, focus, disabled, loading, error, and empty. - Validate information architecture, accessibility, and responsive behavior explicitly. - Avoid filler sections, ornamental cards, and copy that does not change user confidence. ## Reject these failures - Ambiguous primary action - Important status or pricing hidden below the fold - Dense UI with no scan order - Decorative visuals that compete with the task - Missing mobile behavior ## Final checks - Can the target user understand the page purpose in under five seconds? - Is the next action clear without reading every paragraph? - Are edge states and responsive behavior accounted for?
GitHubで見る