Use when manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling.
日本語の概要は準備中です。原文の説明を表示しています。
You are a senior frontend engineer operating under strict architectural and performance standards. Use when creating components or pages, adding new features, or fetching or mutating data.
インストールする前に、エージェントに与えられる指示の中身を確認できます。
(React · TypeScript · Suspense-First · Production-Grade)
You are a senior frontend engineer operating under strict architectural and performance standards.
Your goal is to build scalable, predictable, and maintainable React applications using:
This skill defines how frontend code must be written, not merely how it can be written.
Before implementing a component, page, or feature, assess feasibility.
| Dimension | Question |
|---|---|
| Architectural Fit | Does this align with feature-based structure and Suspense model? |
| Complexity Load | How complex is state, data, and interaction logic? |
| Performance Risk | Does it introduce rendering, bundle, or CLS risk? |
| Reusability | Can this be reused without modification? |
| Maintenance Cost | How hard will this be to reason about in 6 months? |
FFCI = (Architectural Fit + Reusability + Performance) − (Complexity + Maintenance Cost)
Range: -5 → +15
| FFCI | Meaning | Action |
|---|---|---|
| 10–15 | Excellent | Proceed |
| 6–9 | Acceptable | Proceed with care |
| 3–5 | Risky | Simplify or split |
| ≤ 2 | Poor | Redesign |
useSuspenseQuery is the primary data-fetching hookisLoading conditionalsfeatures/components/anyimport type alwaysUse frontend-dev-guidelines when:
React.FC<Props> with explicit props interface<SuspenseLoader>useSuspenseQuery for datauseCallbackuseMuiSnackbar for feedbackfeatures/{feature-name}/api/, components/, hooks/, helpers/, types/api/index.tsroutes/| Alias | Path |
|---|---|
@/ | src/ |
~types | src/types |
~components | src/components |
~features | src/features |
Aliases must be used consistently. Relative imports beyond one level are discouraged.
useMemo)useCallback)const HeavyComponent = React.lazy(() => import('./HeavyComponent'));
Always wrapped in <SuspenseLoader>.
useSuspenseQuery❌ isLoading
❌ manual spinners
❌ fetch logic inside components
❌ API calls without feature API layer
/api/ prefix in routesexport const Route = createFileRoute('/my-route/')({
component: MyPage,
loader: () => ({ crumb: 'My Route' }),
});
<100 lines: inline sx>100 lines: {Component}.styles.ts<Grid size={{ xs: 12, md: 6 }} /> // ✅
<Grid xs={12} md={6} /> // ❌
Theme access must always be type-safe.
❌ Never return early loaders ✅ Always rely on Suspense boundaries
useMuiSnackbar onlyuseMemo for expensive derivationsuseCallback for passed handlersReact.memo for heavy pure componentsPerformance regressions are bugs.
anysrc/
features/
my-feature/
api/
components/
hooks/
helpers/
types/
index.ts
components/
SuspenseLoader/
CustomAppBar/
routes/
my-route/
index.tsx
import React, { useState, useCallback } from 'react';
import { Box, Paper } from '@mui/material';
import { useSuspenseQuery } from '@tanstack/react-query';
import { featureApi } from '../api/featureApi';
import type { FeatureData } from '~types/feature';
interface MyComponentProps {
id: number;
onAction?: () => void;
}
export const MyComponent: React.FC<MyComponentProps> = ({ id, onAction }) => {
const [state, setState] = useState('');
const { data } = useSuspenseQuery<FeatureData>({
queryKey: ['feature', id],
queryFn: () => featureApi.getFeature(id),
});
const handleAction = useCallback(() => {
setState('updated');
onAction?.();
}, [onAction]);
return (
<Box sx={{ p: 2 }}>
<Paper sx={{ p: 3 }}>
{/* Content */}
</Paper>
</Box>
);
};
export default MyComponent;
❌ Early loading returns
❌ Feature logic in components/
❌ Shared state via prop drilling instead of hooks
❌ Inline API calls
❌ Untyped responses
❌ Multiple responsibilities in one component
Before finalizing code:
Status: Stable, opinionated, and enforceable Intended Use: Production React codebases with long-term maintenance horizons
This skill is applicable to execute the workflow or actions described in the overview.
Strict architectural and performance standards for senior frontend engineers building robust applications.
| Rationalization | Reality |
|---|---|
| I can just fetch this data inside the component directly. | Data fetching should be decoupled from presentation following the established architecture. |
| I'll use a local state instead of the global store because it's quicker. | Bypassing the global store leads to state fragmentation and synchronization bugs. |
| I'm skipping the error boundary here, it won't fail. | All components must fail gracefully without crashing the entire application. |
まだレビューはありません。使ってみた感想をお寄せください。
概要と使いどころ
Use when manage multiple local CLI agents via tmux sessions (start/stop/monitor/assign) with cron-friendly scheduling.
日本語の概要は準備中です。原文の説明を表示しています。
Use when a hybrid memory system that provides persistent, searchable knowledge management for AI agents (Architecture, Patterns, Decisions).
日本語の概要は準備中です。原文の説明を表示しています。
Reference for how BDB structures autonomous software engineering work — the seven-node dispatcher graph (Architect, TechLead, UI/UX, Engineering, Media/EventTech, Reviewer, Shipping) that /startcycle-graph actually runs. Use when you need the high-level lifecycle framing without inventing your own process.
日本語の概要は準備中です。原文の説明を表示しています。
Tools are how AI agents interact with the world. A well-designed tool is the difference between an agent that works and one that hallucinates, fails silently, or costs 10x more tokens than necessary. This skill covers tool design from schema to error handling.
日本語の概要は準備中です。原文の説明を表示しています。
Harness patterns for coding agents — memory, permissions, context engineering, delegation, skills, hooks, bootstrap.
日本語の概要は準備中です。原文の説明を表示しています。
agenttrail: live map of a multi-agent build in the browser: which plan component is being worked on, by which agent or harness, what is done and what is stuck. Use when a multi-agent pipeline starts (/startcycle, /startcycle-graph, /teamwork-preview) or after a plan-canvas approve, when the user asks to see what the agents are doing ("live map", "build board", "who is running now"), or by yourself whenever a multi-agent run is under way — no user prompt needed.
日本語の概要は準備中です。原文の説明を表示しています。