Token导航 LogoToken导航

Agent Skill · 研究检索

react-flow-architect(React flow 架构师)

用于辅助前端页面、组件、样式和交互逻辑的开发与维护。它适合让 Agent 生成或审查 React、Next.js、Vue、Tailwind、CSS 等相关代码,整理组件结构,或定位布局和性能问题。使用时需要结合项目现有设计系统、路由和构建方式,避免只生成孤立片段;涉及页面改动时,应配合本地预览和构建检查确认视觉效果。

react-flow-architect 是研究检索类 Agent Skill,可用宿主为 Codex、Claude、Cursor、Gemini CLI,通过GitHub 仓库获取,权限为权限未标注;许可证信息还不明确,第三方扫描已通过,按安装量在研究检索类的 73,675 个 Skill 里排第 28,973。

总安装 3,128周安装 133GitHub Stars 35,726下载 1,096核验 2026-05-01

安装说明

本站只整理中文说明,不托管安装包,也不代用户安装。
GitHub
许可证
unknown

安装方式

通过对话安装

复制提示词发给支持本地命令或 Skills 的 AI 助手,先确认命令和权限,再让它执行。

请帮我安装这个 Agent Skill:react-flow-architect(React flow 架构师)
来源仓库:https://github.com/sickn33/antigravity-awesome-skills
仓库路径:skills/react-flow-architect
安装命令:
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill react-flow-architect
安装前请先检查当前环境是否支持对应 CLI,并向我确认将要执行的命令、安装目录、联网范围和文件读写权限;确认后再执行。

命令行安装

复制命令到本机终端执行。该命令会通过 npx skills 从第三方来源获取 Skill;本站只展示命令,不托管安装包,也不自动执行。

skills.shnpx skills
npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill react-flow-architect

这个 Skill 是做什么的

研究检索类 Skill 主要解决「找资料、核对事实、把来源整理清楚」的问题,通常会让助手先检索再总结。

专注 React Flow 架构设计与优化建议。适用于大型流程图系统的模块拆分与数据流管理。提供节点通信、状态共享和渲染性能的架构指导。通过 GitHub 安装,需根据项目规模调整实现策略。推荐在本地验证架构变更对构建和运行的影响。

react-flow-architect 属于研究检索类 Skill,可作为该场景下的辅助能力补充。

来源仓库内路径:skills/react-flow-architect

上手步骤

4 步
  1. 1

    确认自己的宿主工具

    公开资料标注支持 Codex、Claude、Cursor、Gemini CLI;先确认自己用的是其中之一,再按对应工具的 Skills 目录放置。

  2. 2

    拿到安装命令

    可以直接复制的命令是 npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill rea…;命令里的占位符换成自己的路径或密钥再执行。

  3. 3

    安装后重启宿主

    命令执行完要完全退出宿主工具再打开,让新装的 Skill 被重新扫描到,否则列表里可能看不到。

  4. 4

    按最小权限试用

    权限没有明确标注,第一次使用时只给一个无关紧要的任务,观察它到底读了哪些文件和网络。

常见问题

6 条

react-flow-architect 怎么安装?

在终端执行 npx skills add https://github.com/sickn33/antigravity-awesome-skills --skill react-flow-ar…。装完完全退出宿主工具再打开,让它重新扫描本地 Skill。

react-flow-architect 支持哪些 AI 工具?

公开资料标注支持 Codex、Claude、Cursor、Gemini CLI。其他工具只要能读取本地 Skill 目录,一般也能手动放进去,但要自己确认触发方式是否一致。

react-flow-architect 需要什么权限?

权限级别没有明确标注。第一次使用时建议只给无关紧要的任务,观察它实际读写了哪些内容。

react-flow-architect 是免费的吗?

公开资料没有标注明确许可证,收费和商用限制都无法确认。正式使用前请到来源仓库核对授权说明。

react-flow-architect 用的人多吗?

公开数据是累计安装 3,128 次,近一周安装 133 次,下载 1,096 次,GitHub 35,726 星。这些数字来自第三方目录统计,只能当作热度参考,不代表质量或安全性。

react-flow-architect 安全吗?

第三方扫描结果为通过,但这只覆盖已知的静态检查项,不能替代自己读一遍源码和安装命令。

值得选的地方和要先想清楚的地方

值得选的地方

  • 同时适配 4 种工具,换客户端不用重新找替代品。
  • 第三方扫描已通过,常见的高风险行为没有命中。
  • 来源仓库有 35,726 星,维护活跃度相对可查。

先想清楚的地方

  • 许可证没有明确标注,能不能商用、能不能再分发都不确定。
  • 第三方目录整理,版本更新和行为细节以来源仓库为准。

适不适合你

更适合

  • 已经在用 Codex,想直接扩展它能力的人。
  • 需要处理研究检索类任务、又不想每次都手写提示词的人。

不太适合

  • 需要明确授权才能用到公司项目里的场景。
  • 需要厂商级技术支持和 SLA 的生产系统。

研究检索类 Skill 对比

6 个同类
Skill总安装GitHub 星权限许可证扫描
find-skills

寻找技能 · Codex、Claude

31,199,81316,486操作浏览器MIT有提醒
self-improving-agent

自我改进 Agent · OpenClaw

16,464,0023,396接触敏感数据MIT-0通过
soultrace

灵魂追踪 · Codex、Claude

7,588,94412需要联网未明确有提醒
self-improving

自我完善 · OpenClaw

6,829,2841,046需要联网MIT-0通过
microsoft-foundry

微软代工厂 · Codex、Claude

6,614,144764调用外部服务未明确有提醒
react-flow-architect

React flow 架构师 · 本页

3,12835,726权限未标注未明确通过

按安装量取同分类前几名;数字来自第三方目录统计,只作热度参考,不代表质量结论。

SKILL.md 原文

ReactFlow Architect

Build production-ready ReactFlow applications with hierarchical navigation, performance optimization, and advanced state management.

Quick Start

Create basic interactive graph:

import ReactFlow, { Node, Edge } from "reactflow";

const nodes: Node[] = [
  { id: "1", position: { x: 0, y: 0 }, data: { label: "Node 1" } },
  { id: "2", position: { x: 100, y: 100 }, data: { label: "Node 2" } },
];

const edges: Edge[] = [{ id: "e1-2", source: "1", target: "2" }];

export default function Graph() {
  return <ReactFlow nodes={nodes} edges={edges} />;
}

Core Patterns

Hierarchical Tree Navigation

Build expandable/collapsible tree structures with parent-child relationships.

Node Schema
interface TreeNode extends Node {
  data: {
    label: string;
    level: number;
    hasChildren: boolean;
    isExpanded: boolean;
    childCount: number;
    category: "root" | "category" | "process" | "detail";
  };
}
Incremental Node Building
const buildVisibleNodes = useCallback(
  (allNodes: TreeNode[], expandedIds: Set<string>, otherDeps: any[]) => {
    const visibleNodes = new Map<string, TreeNode>();
    const visibleEdges = new Map<string, TreeEdge>();

    // Start with root nodes
    const rootNodes = allNodes.filter((n) => n.data.level === 0);

    // Recursively add visible nodes
    const addVisibleChildren = (node: TreeNode) => {
      visibleNodes.set(node.id, node);

      if (expandedIds.has(node.id)) {
        const children = allNodes.filter((n) => n.parentNode === node.id);
        children.forEach((child) => addVisibleChildren(child));
      }
    };

    rootNodes.forEach((root) => addVisibleChildren(root));

    return {
      nodes: Array.from(visibleNodes.values()),
      edges: Array.from(visibleEdges.values()),
    };
  },
  [],
);

Performance Optimization

Handle large datasets with incremental rendering and memoization.

Incremental Rendering
const useIncrementalGraph = (
  allNodes: Node[],
  allEdges: Edge[],
  expandedList: string[],
) => {
  const prevExpandedListRef = useRef<Set<string>>(new Set());
  const prevOtherDepsRef = useRef<any[]>([]);

  const { visibleNodes, visibleEdges } = useMemo(() => {
    const currentExpandedSet = new Set(expandedList);
    const prevExpandedSet = prevExpandedListRef.current;

    // Check if expanded list changed
    const expandedChanged = !areSetsEqual(currentExpandedSet, prevExpandedSet);

    // Check if other dependencies changed
    const otherDepsChanged = !arraysEqual(otherDeps, prevOtherDepsRef.current);

    if (expandedChanged && !otherDepsChanged) {
      // Only expanded list changed - incremental update
      return buildIncrementalUpdate(
        cachedVisibleNodesRef.current,
        cachedVisibleEdgesRef.current,
        allNodes,
        allEdges,
        currentExpandedSet,
        prevExpandedSet,
      );
    } else {
      // Full rebuild needed
      return buildFullGraph(allNodes, allEdges, currentExpandedSet);
    }
  }, [allNodes, allEdges, expandedList, ...otherDeps]);

  return { visibleNodes, visibleEdges };
};
Memoization Patterns
// Memoize node components to prevent unnecessary re-renders
const ProcessNode = memo(({ data, selected }: NodeProps) => {
  return (
    <div className={`process-node ${selected ? 'selected' : ''}`}>
      {data.label}
    </div>
  );
}, (prevProps, nextProps) => {
  // Custom comparison function
  return (
    prevProps.data.label === nextProps.data.label &&
    prevProps.selected === nextProps.selected &&
    prevProps.data.isExpanded === nextProps.data.isExpanded
  );
});

// Memoize edge calculations
const styledEdges = useMemo(() => {
  return edges.map(edge => ({
    ...edge,
    style: {
      ...edge.style,
      strokeWidth: selectedEdgeId === edge.id ? 3 : 2,
      stroke: selectedEdgeId === edge.id ? '#3b82f6' : '#94a3b8',
    },
    animated: selectedEdgeId === edge.id,
  }));
}, [edges, selectedEdgeId]);

State Management

Complex node/edge state patterns with undo/redo and persistence.

Reducer Pattern
type GraphAction =
  | { type: "SELECT_NODE"; payload: string }
  | { type: "SELECT_EDGE"; payload: string }
  | { type: "TOGGLE_EXPAND"; payload: string }
  | { type: "UPDATE_NODES"; payload: Node[] }
  | { type: "UPDATE_EDGES"; payload: Edge[] }
  | { type: "UNDO" }
  | { type: "REDO" };

const graphReducer = (state: GraphState, action: GraphAction): GraphState => {
  switch (action.type) {
    case "SELECT_NODE":
      return {
        ...state,
        selectedNodeId: action.payload,
        selectedEdgeId: null,
      };

    case "TOGGLE_EXPAND":
      const newExpanded = new Set(state.expandedNodeIds);
      if (newExpanded.has(action.payload)) {
        newExpanded.delete(action.payload);
      } else {
        newExpanded.add(action.payload);
      }
      return {
        ...state,
        expandedNodeIds: newExpanded,
        isDirty: true,
      };

    default:
      return state;
  }
};
History Management
const useHistoryManager = (
  state: GraphState,
  dispatch: Dispatch<GraphAction>,
) => {
  const canUndo = state.historyIndex > 0;
  const canRedo = state.historyIndex < state.history.length - 1;

  const undo = useCallback(() => {
    if (canUndo) {
      const newIndex = state.historyIndex - 1;
      const historyEntry = state.history[newIndex];

      dispatch({
        type: "RESTORE_FROM_HISTORY",
        payload: {
          ...historyEntry,
          historyIndex: newIndex,
        },
      });
    }
  }, [canUndo, state.historyIndex, state.history]);

  const saveToHistory = useCallback(() => {
    dispatch({ type: "SAVE_TO_HISTORY" });
  }, [dispatch]);

  return { canUndo, canRedo, undo, redo, saveToHistory };
};

Advanced Features

Auto-Layout Integration

Integrate Dagre for automatic graph layout:

import dagre from "dagre";

const layoutOptions = {
  rankdir: "TB", // Top to Bottom
  nodesep: 100, // Node separation
  ranksep: 150, // Rank separation
  marginx: 50,
  marginy: 50,
  edgesep: 10,
};

const applyLayout = (nodes: Node[], edges: Edge[]) => {
  const g = new dagre.graphlib.Graph();
  g.setGraph(layoutOptions);
  g.setDefaultEdgeLabel(() => ({}));

  // Add nodes to graph
  nodes.forEach((node) => {
    g.setNode(node.id, { width: 200, height: 100 });
  });

  // Add edges to graph
  edges.forEach((edge) => {
    g.setEdge(edge.source, edge.target);
  });

  // Calculate layout
  dagre.layout(g);

  // Apply positions
  return nodes.map((node) => ({
    ...node,
    position: {
      x: g.node(node.id).x - 100,
      y: g.node(node.id).y - 50,
    },
  }));
};

// Debounce layout calculations
const debouncedLayout = useMemo(() => debounce(applyLayout, 150), []);

Focus Mode

Isolate selected nodes and their direct connections:

const useFocusMode = (
  selectedNodeId: string,
  allNodes: Node[],
  allEdges: Edge[],
) => {
  return useMemo(() => {
    if (!selectedNodeId) return { nodes: allNodes, edges: allEdges };

    // Get direct connections
    const connectedNodeIds = new Set([selectedNodeId]);
    const focusedEdges: Edge[] = [];

    allEdges.forEach((edge) => {
      if (edge.source === selectedNodeId || edge.target === selectedNodeId) {
        focusedEdges.push(edge);
        connectedNodeIds.add(edge.source);
        connectedNodeIds.add(edge.target);
      }
    });

    // Get connected nodes
    const focusedNodes = allNodes.filter((n) => connectedNodeIds.has(n.id));

    return { nodes: focusedNodes, edges: focusedEdges };
  }, [selectedNodeId, allNodes, allEdges]);
};

// Smooth transitions for focus mode
const focusModeStyles = {
  transition: "all 0.3s ease-in-out",
  opacity: isInFocus ? 1 : 0.3,
  filter: isInFocus ? "none" : "blur(2px)",
};

Search Integration

Search and navigate to specific nodes:

const searchNodes = useCallback((nodes: Node[], query: string) => {
  if (!query.trim()) return [];

  const lowerQuery = query.toLowerCase();
  return nodes.filter(
    (node) =>
      node.data.label.toLowerCase().includes(lowerQuery) ||
      node.data.description?.toLowerCase().includes(lowerQuery),
  );
}, []);

const navigateToSearchResult = (nodeId: string) => {
  // Expand parent nodes
  const nodePath = calculateBreadcrumbPath(nodeId, allNodes);
  const parentIds = nodePath.slice(0, -1).map((n) => n.id);

  setExpandedIds((prev) => new Set([...prev, ...parentIds]));
  setSelectedNodeId(nodeId);

  // Fit view to node
  fitView({ nodes: [{ id: nodeId }], duration: 800 });
};

Performance Tools

Graph Performance Analyzer

Create a performance analysis script:

// scripts/graph-analyzer.js
class GraphAnalyzer {
  analyzeCode(content, filePath) {
    const analysis = {
      metrics: {
        nodeCount: this.countNodes(content),
        edgeCount: this.countEdges(content),
        renderTime: this.estimateRenderTime(content),
        memoryUsage: this.estimateMemoryUsage(content),
        complexity: this.calculateComplexity(content),
      },
      issues: [],
      optimizations: [],
      patterns: this.detectPatterns(content),
    };

    // Detect performance issues
    this.detectPerformanceIssues(analysis);

    // Suggest optimizations
    this.suggestOptimizations(analysis);

    return analysis;
  }

  countNodes(content) {
    const nodePatterns = [
      /nodes:\s*\[.*?\]/gs,
      /const\s+\w+\s*=\s*\[.*?id:.*?position:/gs,
    ];

    let totalCount = 0;
    nodePatterns.forEach((pattern) => {
      const matches = content.match(pattern);
      if (matches) {
        matches.forEach((match) => {
          const nodeMatches = match.match(/id:\s*['"`][^'"`]+['"`]/g);
          if (nodeMatches) {
            totalCount += nodeMatches.length;
          }
        });
      }
    });

    return totalCount;
  }

  estimateRenderTime(content) {
    const nodeCount = this.countNodes(content);
    const edgeCount = this.countEdges(content);

    // Base render time estimation (ms)
    const baseTime = 5;
    const nodeTime = nodeCount * 0.1;
    const edgeTime = edgeCount * 0.05;

    return baseTime + nodeTime + edgeTime;
  }

  detectPerformanceIssues(analysis) {
    const { metrics } = analysis;

    if (metrics.nodeCount > 500) {
      analysis.issues.push({
        type: "HIGH_NODE_COUNT",
        severity: "high",
        message: `Too many nodes (${metrics.nodeCount}). Consider virtualization.`,
        suggestion: "Implement virtualization or reduce visible nodes",
      });
    }

    if (metrics.renderTime > 16) {
      analysis.issues.push({
        type: "SLOW_RENDER",
        severity: "high",
        message: `Render time (${metrics.renderTime.toFixed(2)}ms) exceeds 60fps.`,
        suggestion: "Optimize with memoization and incremental rendering",
      });
    }
  }
}

Best Practices

Performance Guidelines

  1. Use React.memo for node components to prevent unnecessary re-renders
  2. Implement virtualization for graphs with 1000+ nodes
  3. Debounce layout calculations during rapid interactions
  4. Use useCallback for edge creation and manipulation functions
  5. Implement proper TypeScript types for nodes and edges

Memory Management

// Use Map for O(1) lookups instead of array.find
const nodesById = useMemo(
  () => new Map(allNodes.map((n) => [n.id, n])),
  [allNodes],
);

// Cache layout results
const layoutCacheRef = useRef<Map<string, Node[]>>(new Map());

// Proper cleanup in useEffect
useEffect(() => {
  return () => {
    // Clean up any lingering references
    nodesMapRef.current.clear();
    edgesMapRef.current.clear();
  };
}, []);

State Optimization

// Use useRef for objects that shouldn't trigger re-renders
const autoSaveDataRef = useRef({
  nodes: [],
  edges: [],
  lastSaved: Date.now(),
});

// Update properties without breaking reference
const updateAutoSaveData = (newNodes: Node[], newEdges: Edge[]) => {
  autoSaveDataRef.current.nodes = newNodes;
  autoSaveDataRef.current.edges = newEdges;
  autoSaveDataRef.current.lastSaved = Date.now();
};

Common Problems & Solutions

Performance Issues

  • Problem: Lag during node expansion
  • Solution: Implement incremental rendering with change detection
  • Problem: Memory usage increases over time
  • Solution: Proper cleanup in useEffect hooks and use WeakMap for temporary data

Layout Conflicts

  • Problem: Manual positioning conflicts with auto-layout
  • Solution: Use controlled positioning state and separate layout modes

Rendering Issues

  • Problem: Excessive re-renders
  • Solution: Use memo, useMemo, and useCallback with stable dependencies
  • Problem: Slow layout calculations
  • Solution: Debounce layout calculations and cache results

Complete Example

import React, { useState, useCallback, useMemo, useRef } from 'react';
import ReactFlow, { Node, Edge, useReactFlow } from 'reactflow';
import dagre from 'dagre';
import { debounce } from 'lodash';

interface GraphState {
  nodes: Node[];
  edges: Edge[];
  selectedNodeId: string | null;
  expandedNodeIds: Set<string>;
  history: GraphState[];
  historyIndex: number;
}

export default function InteractiveGraph() {
  const [state, setState] = useState<GraphState>({
    nodes: [],
    edges: [],
    selectedNodeId: null,
    expandedNodeIds: new Set(),
    history: [],
    historyIndex: 0,
  });

  const { fitView } = useReactFlow();
  const layoutCacheRef = useRef<Map<string, Node[]>>(new Map());

  // Memoized styled edges
  const styledEdges = useMemo(() => {
    return state.edges.map(edge => ({
      ...edge,
      style: {
        ...edge.style,
        strokeWidth: state.selectedNodeId === edge.source || state.selectedNodeId === edge.target ? 3 : 2,
        stroke: state.selectedNodeId === edge.source || state.selectedNodeId === edge.target ? '#3b82f6' : '#94a3b8',
      },
      animated: state.selectedNodeId === edge.source || state.selectedNodeId === edge.target,
    }));
  }, [state.edges, state.selectedNodeId]);

  // Debounced layout calculation
  const debouncedLayout = useMemo(
    () => debounce((nodes: Node[], edges: Edge[]) => {
      const cacheKey = generateLayoutCacheKey(nodes, edges);

      if (layoutCacheRef.current.has(cacheKey)) {
        return layoutCacheRef.current.get(cacheKey)!;
      }

      const layouted = applyDagreLayout(nodes, edges);
      layoutCacheRef.current.set(cacheKey, layouted);

      return layouted;
    }, 150),
    []
  );

  const handleNodeClick = useCallback((event: React.MouseEvent, node: Node) => {
    setState(prev => ({
      ...prev,
      selectedNodeId: node.id,
    }));
  }, []);

  const handleToggleExpand = useCallback((nodeId: string) => {
    setState(prev => {
      const newExpanded = new Set(prev.expandedNodeIds);
      if (newExpanded.has(nodeId)) {
        newExpanded.delete(nodeId);
      } else {
        newExpanded.add(nodeId);
      }

      return {
        ...prev,
        expandedNodeIds: newExpanded,
      };
    });
  }, []);

  return (
    <ReactFlow
      nodes={state.nodes}
      edges={styledEdges}
      onNodeClick={handleNodeClick}
      fitView
    />
  );
}

This comprehensive skill provides everything needed to build production-ready ReactFlow applications with hierarchical navigation, performance optimization, and advanced state management patterns.

When to Use

This skill is applicable to execute the workflow or actions described in the overview.

Limitations

  • Use this skill only when the task clearly matches the scope described above.
  • Do not treat the output as a substitute for environment-specific validation, testing, or expert review.
  • Stop and ask for clarification if required inputs, permissions, safety boundaries, or success criteria are missing.

平台分布

Codex

34.15%
按下载量换算374

Claude

30.52%
按下载量换算334

Cursor

17.77%
按下载量换算195

Gemini CLI

8.68%
按下载量换算95

安全审计

Gen Agent Trust Hub

通过

Socket

通过

Snyk

通过

权限和风险

权限未标注

当前来源未能明确判断权限范围,默认进入异常复核队列。

安装前确认

本站仅展示第三方公开信息,不托管安装包,不提供自动安装或运行环境。安装前应自行审查源码、依赖和命令行为。目前只有一个公开来源,建议同时核对源仓库。

继续浏览同类 Skills