start-vibing 2.0.38 → 2.0.40

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "start-vibing",
-	"version": "2.0.38",
+	"version": "2.0.40",
 	"description": "Setup Claude Code agents, skills, and hooks in your project. Smart copy that preserves your custom domains and configurations.",
 	"type": "module",
 	"bin": {

package/template/.claude/hooks/user-prompt-submit.ts

@@ -1,738 +1,11 @@
 #!/usr/bin/env node
 /**
- * UserPromptSubmit Hook - TypeScript version (fallback when Python not available)
+ * UserPromptSubmit Hook
  *
- * This hook runs BEFORE Claude processes any user prompt and ENFORCES:
- * 1. MANDATORY detailed todo list creation from prompt
- * 2. MANDATORY research agent for new features
- * 3. STRICT workflow: audit -> branch -> implement -> document -> quality -> merge to main
- * 4. Separate UI for mobile/tablet/desktop (NOT just responsive)
- *
- * AGENT SYSTEM: 82 specialized agents in 14 categories
- * SKILL SYSTEM: 22 skills auto-loaded by agents
+ * Concise instructions injected before Claude processes each user prompt.
+ * Enforces: todo-list creation, sequential execution, commit rules, CLAUDE.md update.
  */
 
-import { existsSync, readFileSync } from 'fs';
-import { join } from 'path';
-
-const PROJECT_DIR = process.env['CLAUDE_PROJECT_DIR'] || process.cwd();
-const WORKFLOW_STATE_PATH = join(PROJECT_DIR, '.claude', 'workflow-state.json');
-
-const STRICT_WORKFLOW = `
-┌─────────────────────────────────────────────────────────────────┐
-│                    STRICT WORKFLOW (MANDATORY)                  │
-├─────────────────────────────────────────────────────────────────┤
-│ 0. INIT TASK    → Run commit-manager to REGISTER task start    │
-│ 1. TODO LIST    → Create DETAILED todo list from prompt        │
-│ 2. RESEARCH     → Run research agent for NEW features          │
-│ 3. AUDIT        → Check last audit docs OR run fresh audit     │
-│ 4. BRANCH       → Create feature/ | fix/ | refactor/ | test/   │
-│ 5. IMPLEMENT    → Follow rules + analyzer approval             │
-│ 6. DOCUMENT     → Document ALL modified files (MANDATORY)      │
-│ 7. UPDATE       → Update CLAUDE.md with session changes        │
-│ 8. QUALITY      → typecheck + lint + test (Husky enforced)     │
-│ 9. VALIDATE     → RUN: npx tsx .claude/hooks/stop-validator.ts │
-│ 10. FINISH      → Only after "ALL CHECKS PASSED"               │
-└─────────────────────────────────────────────────────────────────┘
-
-╔═════════════════════════════════════════════════════════════════╗
-║  ⚠️  MANDATORY: RUN STOP VALIDATOR BEFORE COMPLETING ANY TASK   ║
-╠═════════════════════════════════════════════════════════════════╣
-║                                                                 ║
-║  COMMAND: npx tsx .claude/hooks/stop-validator.ts               ║
-║                                                                 ║
-║  YOU MUST run this command manually BEFORE saying "task done".  ║
-║  Fix ALL issues one by one. Run validator again after EACH fix.║
-║  Only complete task when output shows "ALL CHECKS PASSED".      ║
-║                                                                 ║
-╚═════════════════════════════════════════════════════════════════╝
-
-⚠️  STOP HOOK BLOCKS task completion if:
-    - NOT on main branch (work must be merged to main first)
-    - Git tree NOT clean (all changes must be committed)
-    - CLAUDE.md NOT updated (must reflect session changes)
-    - CLAUDE.md missing required sections (Last Change, Stack, etc.)
-    - CLAUDE.md exceeds 40,000 characters (must compact)
-    - Source files NOT documented (must run documenter agent)
-`;
-
-// ============================================================================
-// AGENT SYSTEM - 82 Specialized Agents in 14 Categories
-// ============================================================================
-
-interface AgentInfo {
-	triggers: string[];
-	when: string;
-	skills: string[];
-	can_veto?: boolean;
-	mandatory_for?: string[];
-}
-
-interface CategoryInfo {
-	description: string;
-	agents: Record<string, AgentInfo>;
-}
-
-const AGENT_CATEGORIES: Record<string, CategoryInfo> = {
-	'01-orchestration': {
-		description: 'Workflow coordination and task management',
-		agents: {
-			orchestrator: {
-				triggers: ['implement feature', 'build', 'create', 'full workflow', 'multi-step'],
-				when: 'Task requires >2 agents or touches >3 files',
-				skills: ['codebase-knowledge'],
-			},
-			'task-decomposer': {
-				triggers: ['complex task', 'break down', 'multiple steps'],
-				when: 'Task has >3 steps or touches >3 files',
-				skills: ['codebase-knowledge'],
-			},
-			'workflow-router': {
-				triggers: ['route', 'which agent', 'unclear'],
-				when: 'At task start to route to correct agent',
-				skills: ['codebase-knowledge'],
-			},
-			'parallel-coordinator': {
-				triggers: ['parallel', 'concurrent', 'simultaneous'],
-				when: 'Multiple independent agents should run simultaneously',
-				skills: ['codebase-knowledge'],
-			},
-			'context-manager': {
-				triggers: ['context', 'compress', 'long conversation'],
-				when: 'Context grows large or between major phases',
-				skills: ['codebase-knowledge'],
-			},
-			'checkpoint-manager': {
-				triggers: ['checkpoint', 'save state', 'backup'],
-				when: 'BEFORE risky operations (git, deletions, refactors)',
-				skills: ['codebase-knowledge'],
-			},
-			'error-recovery': {
-				triggers: ['failed', 'retry', 'timeout', 'unexpected error'],
-				when: 'An agent fails or returns unexpected results',
-				skills: ['codebase-knowledge', 'debugging-patterns'],
-			},
-			'agent-selector': {
-				triggers: ['select agent', 'best agent', 'which agent'],
-				when: 'Multiple agents could handle a task',
-				skills: ['codebase-knowledge'],
-			},
-		},
-	},
-	'02-typescript': {
-		description: 'TypeScript strict mode, types, and module resolution',
-		agents: {
-			'ts-strict-checker': {
-				triggers: ['strict mode', 'null check', 'process.env'],
-				when: 'AFTER editing any .ts file',
-				skills: ['typescript-strict'],
-			},
-			'ts-types-analyzer': {
-				triggers: ['type error', 'inference', 'generic problem'],
-				when: 'On type errors or typecheck fails',
-				skills: ['typescript-strict'],
-			},
-			'ts-generics-helper': {
-				triggers: ['generic', 'type parameter', 'complex type'],
-				when: 'Creating generic functions/types',
-				skills: ['typescript-strict'],
-			},
-			'ts-migration-helper': {
-				triggers: ['migrate', 'convert to typescript', '.js'],
-				when: 'Migrating JavaScript to TypeScript',
-				skills: ['typescript-strict'],
-			},
-			'type-definition-writer': {
-				triggers: ['new model', 'new entity', 'interface needed'],
-				when: 'BEFORE implementing new entities',
-				skills: ['typescript-strict'],
-			},
-			'import-alias-enforcer': {
-				triggers: ['import', 'alias', '@types'],
-				when: 'AFTER editing .ts files - enforces $types/*, @common, @db',
-				skills: ['typescript-strict'],
-			},
-			'esm-resolver': {
-				triggers: ['module error', 'import error', 'cannot find module'],
-				when: 'On module errors',
-				skills: ['typescript-strict', 'bun-runtime'],
-			},
-			'bun-runtime-expert': {
-				triggers: ['bun', 'runtime', 'package management'],
-				when: 'Using Bun runtime',
-				skills: ['bun-runtime'],
-			},
-			'zod-validator': {
-				triggers: ['validation', 'zod', 'schema'],
-				when: 'BEFORE commit when API routes exist',
-				skills: ['zod-validation'],
-			},
-			'zod-schema-designer': {
-				triggers: ['new endpoint', 'form input', 'user input'],
-				when: 'BEFORE implementing any API endpoint',
-				skills: ['zod-validation', 'typescript-strict'],
-			},
-		},
-	},
-	'03-testing': {
-		description: 'Unit tests (Vitest) and E2E tests (Playwright)',
-		agents: {
-			'vitest-config': {
-				triggers: ['test setup', 'vitest config', 'coverage config'],
-				when: 'Setting up tests or coverage issues arise',
-				skills: ['test-coverage'],
-			},
-			'tester-unit': {
-				triggers: ['unit test', 'function test', 'utility test'],
-				when: 'AFTER implementing any function or utility',
-				skills: ['test-coverage'],
-			},
-			'tester-integration': {
-				triggers: ['integration test', 'api test', 'service test'],
-				when: 'AFTER implementing API endpoints or services',
-				skills: ['test-coverage'],
-			},
-			'test-data-generator': {
-				triggers: ['mock data', 'test fixture', 'factory'],
-				when: 'BEFORE writing tests that need data',
-				skills: ['test-coverage'],
-			},
-			'test-cleanup-manager': {
-				triggers: ['flaky test', 'test isolation', 'state leakage'],
-				when: 'Tests are flaky or share state',
-				skills: ['test-coverage'],
-			},
-			'playwright-e2e': {
-				triggers: ['e2e test', 'user flow', 'end to end'],
-				when: 'AFTER implementing any user-facing feature',
-				skills: ['test-coverage', 'playwright-automation'],
-			},
-			'playwright-fixtures': {
-				triggers: ['fixture', 'shared setup', 'auth helper'],
-				when: 'Creating E2E tests that need shared setup',
-				skills: ['test-coverage', 'playwright-automation'],
-			},
-			'playwright-page-objects': {
-				triggers: ['page object', 'page model', 'page interaction'],
-				when: 'Creating E2E tests for new pages',
-				skills: ['test-coverage', 'playwright-automation'],
-			},
-			'playwright-multi-viewport': {
-				triggers: ['viewport', 'responsive test', 'mobile test'],
-				when: 'AFTER any UI implementation - tests desktop/tablet/mobile',
-				skills: ['test-coverage', 'playwright-automation'],
-			},
-			'playwright-assertions': {
-				triggers: ['assertion', 'expect', 'test validation'],
-				when: 'Creating comprehensive test assertions',
-				skills: ['test-coverage', 'playwright-automation'],
-			},
-		},
-	},
-	'04-docker': {
-		description: 'Docker containerization and deployment',
-		agents: {
-			'dockerfile-optimizer': {
-				triggers: ['dockerfile', 'docker build', 'image size'],
-				when: 'Creating or modifying Dockerfile',
-				skills: ['docker-patterns'],
-			},
-			'docker-multi-stage': {
-				triggers: ['multi-stage', 'build optimization', 'production docker'],
-				when: 'Dockerfile can benefit from multi-stage',
-				skills: ['docker-patterns'],
-			},
-			'docker-compose-designer': {
-				triggers: ['docker-compose', 'multi-service', 'local dev'],
-				when: 'Multi-service setup is needed',
-				skills: ['docker-patterns'],
-			},
-			'docker-env-manager': {
-				triggers: ['docker env', 'secrets', 'env vars'],
-				when: 'Docker uses environment variables',
-				skills: ['docker-patterns'],
-			},
-			'container-health': {
-				triggers: ['health check', 'container monitoring', 'service health'],
-				when: 'Creating Docker containers',
-				skills: ['docker-patterns'],
-			},
-			'deployment-validator': {
-				triggers: ['deploy', 'pre-deploy', 'docker validation'],
-				when: 'BEFORE deploying',
-				skills: ['docker-patterns'],
-			},
-		},
-	},
-	'05-database': {
-		description: 'MongoDB/Mongoose database operations',
-		agents: {
-			'mongoose-schema-designer': {
-				triggers: ['schema', 'model', 'collection'],
-				when: 'BEFORE creating any database model',
-				skills: ['mongoose-patterns'],
-			},
-			'mongoose-index-optimizer': {
-				triggers: ['index', 'slow query', 'query performance'],
-				when: 'Database queries are slow',
-				skills: ['mongoose-patterns'],
-			},
-			'mongoose-aggregation': {
-				triggers: ['aggregation', 'pipeline', 'reporting'],
-				when: 'Complex data queries needed',
-				skills: ['mongoose-patterns'],
-			},
-			'mongodb-query-optimizer': {
-				triggers: ['n+1', 'query optimizer', 'database performance'],
-				when: 'Queries are slow',
-				skills: ['mongoose-patterns'],
-			},
-			'database-seeder': {
-				triggers: ['seed', 'sample data', 'dev data'],
-				when: 'Setting up development environment',
-				skills: ['mongoose-patterns'],
-			},
-			'data-migration': {
-				triggers: ['migration', 'schema change', 'data transform'],
-				when: 'Schema changes are needed',
-				skills: ['mongoose-patterns'],
-			},
-		},
-	},
-	'06-security': {
-		description: 'Security auditing and OWASP compliance (CAN VETO)',
-		agents: {
-			'security-auditor': {
-				triggers: ['security', 'audit', 'vulnerability'],
-				when: 'BEFORE committing auth/user/API code',
-				can_veto: true,
-				skills: ['security-scan'],
-			},
-			'owasp-checker': {
-				triggers: ['owasp', 'top 10', 'security review'],
-				when: 'BEFORE committing any API or security code',
-				can_veto: true,
-				skills: ['security-scan'],
-			},
-			'input-sanitizer': {
-				triggers: ['sanitize', 'user input', 'xss'],
-				when: 'Handling user input',
-				skills: ['security-scan', 'zod-validation'],
-			},
-			'auth-session-validator': {
-				triggers: ['auth', 'session', 'login', 'jwt'],
-				when: 'Implementing auth or session code',
-				skills: ['security-scan'],
-			},
-			'permission-auditor': {
-				triggers: ['permission', 'authorization', 'access control'],
-				when: 'Implementing protected routes',
-				skills: ['security-scan'],
-			},
-			'sensitive-data-scanner': {
-				triggers: ['sensitive', 'pii', 'data leak'],
-				when: 'Implementing API responses or logging',
-				skills: ['security-scan'],
-			},
-		},
-	},
-	'07-documentation': {
-		description: 'Documentation and domain knowledge updates',
-		agents: {
-			documenter: {
-				triggers: ['document', 'docs', 'explain code'],
-				when: 'AFTER any code implementation completes',
-				skills: ['docs-tracker', 'codebase-knowledge'],
-			},
-			'domain-updater': {
-				triggers: ['domain', 'learnings', 'session end'],
-				when: 'BEFORE commit-manager at session end',
-				skills: ['codebase-knowledge', 'docs-tracker'],
-			},
-			'readme-generator': {
-				triggers: ['readme', 'project docs', 'setup guide'],
-				when: 'Project structure changes significantly',
-				skills: ['docs-tracker'],
-			},
-			'jsdoc-generator': {
-				triggers: ['jsdoc', 'function docs', 'api docs'],
-				when: 'Complex functions lack documentation',
-				skills: ['docs-tracker', 'typescript-strict'],
-			},
-			'changelog-manager': {
-				triggers: ['changelog', 'release notes', 'version'],
-				when: 'BEFORE committing any feature or fix',
-				skills: ['docs-tracker', 'git-workflow'],
-			},
-			'api-documenter': {
-				triggers: ['api docs', 'swagger', 'openapi'],
-				when: 'AFTER creating or modifying API endpoints',
-				skills: ['docs-tracker'],
-			},
-		},
-	},
-	'08-git': {
-		description: 'Git workflow and version control',
-		agents: {
-			'branch-manager': {
-				triggers: ['branch', 'feature branch', 'fix branch'],
-				when: 'BEFORE making source changes on main branch',
-				skills: ['git-workflow'],
-			},
-			'commit-manager': {
-				triggers: ['commit', 'push', 'finalize', 'merge'],
-				when: 'FINAL AGENT when implementation is complete - commits and merges to main',
-				skills: ['git-workflow', 'docs-tracker', 'codebase-knowledge'],
-			},
-		},
-	},
-	'09-quality': {
-		description: 'Code quality and review',
-		agents: {
-			'quality-checker': {
-				triggers: ['quality', 'typecheck', 'lint', 'build'],
-				when: 'BEFORE any commit',
-				skills: ['quality-gate', 'codebase-knowledge'],
-			},
-			'code-reviewer': {
-				triggers: ['review', 'code review', 'code quality'],
-				when: 'AFTER significant code is written',
-				skills: ['quality-gate', 'codebase-knowledge'],
-			},
-		},
-	},
-	'10-research': {
-		description: 'Web research and best practices (MANDATORY for new features)',
-		agents: {
-			'research-web': {
-				triggers: ['search', 'find info', 'look up'],
-				when: 'BEFORE implementing any new feature or technology',
-				mandatory_for: ['feature'],
-				skills: ['research-cache'],
-			},
-			'research-cache-manager': {
-				triggers: ['cache', 'cached research', 'previous research'],
-				when: 'BEFORE any web research',
-				skills: ['research-cache'],
-			},
-			'best-practices-finder': {
-				triggers: ['best practice', 'recommended', 'how should'],
-				when: 'BEFORE implementing any new pattern',
-				skills: ['research-cache'],
-			},
-			'pattern-researcher': {
-				triggers: ['pattern', 'architecture', 'design decision'],
-				when: 'Facing architectural decisions',
-				skills: ['research-cache'],
-			},
-			'competitor-analyzer': {
-				triggers: ['competitor', 'similar app', 'how does'],
-				when: 'Designing UI or features with existing market solutions',
-				skills: ['research-cache'],
-			},
-			'tech-evaluator': {
-				triggers: ['compare', 'evaluate', 'which library'],
-				when: 'Choosing between technologies',
-				skills: ['research-cache'],
-			},
-		},
-	},
-	'11-ui-ux': {
-		description: 'UI/UX - SEPARATE UIs for mobile/tablet/desktop (NOT just responsive)',
-		agents: {
-			'ui-mobile': {
-				triggers: ['mobile', 'touch', '375px'],
-				when: 'Implementing any UI feature',
-				skills: ['ui-ux-audit', 'react-patterns', 'tailwind-patterns', 'shadcn-ui'],
-			},
-			'ui-tablet': {
-				triggers: ['tablet', 'ipad', '768px'],
-				when: 'Implementing any UI feature',
-				skills: ['ui-ux-audit', 'react-patterns', 'tailwind-patterns', 'shadcn-ui'],
-			},
-			'ui-desktop': {
-				triggers: ['desktop', 'sidebar', '1280px'],
-				when: 'Implementing any UI feature',
-				skills: ['ui-ux-audit', 'react-patterns', 'tailwind-patterns', 'shadcn-ui'],
-			},
-			'skeleton-generator': {
-				triggers: ['skeleton', 'loading', 'placeholder'],
-				when: 'AFTER creating any component that loads data',
-				skills: ['react-patterns', 'tailwind-patterns'],
-			},
-			'design-system-enforcer': {
-				triggers: ['design system', 'consistency', 'component style'],
-				when: 'AFTER creating UI components',
-				skills: ['ui-ux-audit', 'shadcn-ui', 'tailwind-patterns'],
-			},
-			'accessibility-auditor': {
-				triggers: ['a11y', 'accessibility', 'wcag', 'screen reader'],
-				when: 'AFTER any UI implementation',
-				skills: ['ui-ux-audit'],
-			},
-		},
-	},
-	'12-performance': {
-		description: 'Performance profiling and optimization',
-		agents: {
-			'performance-profiler': {
-				triggers: ['slow', 'performance', 'profile', 'bottleneck'],
-				when: 'Application is slow',
-				skills: ['performance-patterns'],
-			},
-			'memory-leak-detector': {
-				triggers: ['memory', 'leak', 'heap'],
-				when: 'Memory issues are suspected',
-				skills: ['performance-patterns'],
-			},
-			'bundle-analyzer': {
-				triggers: ['bundle', 'build size', 'lighthouse'],
-				when: 'Build is large or slow',
-				skills: ['performance-patterns'],
-			},
-			'api-latency-analyzer': {
-				triggers: ['api slow', 'response time', 'latency'],
-				when: 'API endpoints are slow',
-				skills: ['performance-patterns'],
-			},
-			'query-optimizer': {
-				triggers: ['slow query', 'n+1', 'database slow'],
-				when: 'Database queries are slow',
-				skills: ['performance-patterns', 'mongoose-patterns'],
-			},
-			'render-optimizer': {
-				triggers: ['re-render', 'react slow', 'component slow'],
-				when: 'React components re-render excessively',
-				skills: ['performance-patterns', 'react-patterns'],
-			},
-		},
-	},
-	'13-debugging': {
-		description: 'Error analysis and debugging',
-		agents: {
-			debugger: {
-				triggers: ['bug', 'error', 'not working', 'broken', 'fails'],
-				when: 'Any bug or error occurs',
-				skills: ['debugging-patterns'],
-			},
-			'type-error-resolver': {
-				triggers: ['ts error', 'type error', 'typecheck fails'],
-				when: 'On TypeScript type errors',
-				skills: ['debugging-patterns', 'typescript-strict'],
-			},
-			'runtime-error-fixer': {
-				triggers: ['crash', 'exception', 'runtime error'],
-				when: 'On runtime crashes or exceptions',
-				skills: ['debugging-patterns'],
-			},
-			'network-debugger': {
-				triggers: ['fetch error', 'api error', 'cors', 'network'],
-				when: 'On network or API errors',
-				skills: ['debugging-patterns'],
-			},
-			'error-stack-analyzer': {
-				triggers: ['stack trace', 'trace', 'call stack'],
-				when: 'Error includes stack trace',
-				skills: ['debugging-patterns'],
-			},
-			'build-error-fixer': {
-				triggers: ['build failed', 'compile error', 'bundler error'],
-				when: 'Build fails',
-				skills: ['debugging-patterns'],
-			},
-		},
-	},
-	'14-validation': {
-		description: 'Final validation before commit (CAN VETO)',
-		agents: {
-			'final-validator': {
-				triggers: ['final check', 'validate', 'ready to commit'],
-				when: 'BEFORE commit-manager - last check',
-				can_veto: true,
-				skills: ['final-check'],
-			},
-		},
-	},
-};
-
-// ============================================================================
-// SKILL SYSTEM - 22 Skills Auto-loaded by Agents
-// ============================================================================
-
-const SKILLS: Record<string, string> = {
-	'bun-runtime': 'Bun runtime patterns, package management, scripts',
-	'codebase-knowledge': 'Project domain knowledge, file mapping, recent commits',
-	'debugging-patterns': 'Stack traces, runtime errors, build errors, network issues',
-	'docker-patterns': 'Containerization, multi-stage builds, Docker Compose, security',
-	'docs-tracker': 'Documentation maintenance, git diff detection, changelog',
-	'final-check': 'Final validation, tests pass, docs updated, security audited',
-	'git-workflow': 'Branch management, conventional commits, merge to main, hooks',
-	'mongoose-patterns': 'MongoDB schema design, queries, indexes, aggregations',
-	'nextjs-app-router': 'Next.js 15 App Router, Server/Client components, data fetching',
-	'performance-patterns': 'React optimization, bundle analysis, memory leaks, API latency',
-	'playwright-automation': 'E2E tests, browser automation, visual testing, API testing',
-	'quality-gate': 'Quality checks (typecheck, lint, test, build)',
-	'react-patterns': 'React 19 patterns, hooks, state management, performance',
-	'research-cache': 'Cached research findings, best practices by topic',
-	'security-scan': 'OWASP Top 10, user ID validation, sensitive data detection',
-	'shadcn-ui': 'shadcn/ui components, customization, theming, accessibility',
-	'tailwind-patterns': 'Tailwind CSS, responsive design, dark mode, animations',
-	'test-coverage': 'Playwright E2E, Vitest unit tests, coverage tracking',
-	'trpc-api': 'tRPC type-safe APIs, routers, procedures, middleware',
-	'typescript-strict': 'TypeScript strict mode, index access, null checks, generics',
-	'ui-ux-audit': 'UI/UX audits, competitor research, WCAG 2.1, responsiveness',
-	'zod-validation': 'Zod schemas, input validation, type inference, error handling',
-};
-
-const WORKFLOWS: Record<string, string[]> = {
-	feature: [
-		'analyzer',
-		'research-web',
-		'ui-ux-reviewer',
-		'documenter',
-		'tester',
-		'security-auditor',
-		'quality-checker',
-		'final-validator',
-		'domain-updater',
-		'commit-manager',
-	],
-	fix: [
-		'debugger',
-		'analyzer',
-		'tester',
-		'security-auditor',
-		'quality-checker',
-		'final-validator',
-		'domain-updater',
-		'commit-manager',
-	],
-	refactor: [
-		'analyzer',
-		'code-reviewer',
-		'tester',
-		'quality-checker',
-		'final-validator',
-		'domain-updater',
-		'commit-manager',
-	],
-	config: ['quality-checker', 'domain-updater', 'commit-manager'],
-};
-
-interface WorkflowState {
-	currentTask?: {
-		id?: string;
-		type?: string;
-		description?: string;
-		agents?: Record<string, { executed?: boolean }>;
-		approvedFiles?: string[];
-	};
-}
-
-function loadWorkflowState(): WorkflowState | null {
-	if (!existsSync(WORKFLOW_STATE_PATH)) return null;
-	try {
-		return JSON.parse(readFileSync(WORKFLOW_STATE_PATH, 'utf8'));
-	} catch {
-		return null;
-	}
-}
-
-function detectTaskType(prompt: string): string {
-	const promptLower = prompt.toLowerCase();
-
-	if (
-		['bug', 'fix', 'error', 'broken', 'not working', 'issue', 'debug'].some((w) =>
-			promptLower.includes(w)
-		)
-	) {
-		return 'fix';
-	} else if (
-		['refactor', 'restructure', 'reorganize', 'clean up', 'improve'].some((w) =>
-			promptLower.includes(w)
-		)
-	) {
-		return 'refactor';
-	} else if (
-		['config', 'setting', 'env', 'package.json', 'tsconfig'].some((w) =>
-			promptLower.includes(w)
-		)
-	) {
-		return 'config';
-	}
-	return 'feature';
-}
-
-function detectBestAgents(prompt: string): Array<[string, string, string]> {
-	const promptLower = prompt.toLowerCase();
-	const matches: Array<[string, string, string]> = [];
-
-	for (const [category, info] of Object.entries(AGENT_CATEGORIES)) {
-		for (const [agentName, agentInfo] of Object.entries(info.agents)) {
-			for (const trigger of agentInfo.triggers) {
-				if (promptLower.includes(trigger)) {
-					matches.push([category, agentName, `Matched '${trigger}'`]);
-					break;
-				}
-			}
-		}
-	}
-
-	matches.sort((a, b) => a[0].localeCompare(b[0]));
-
-	if (matches.length === 0) {
-		return [
-			['01-orchestration', 'orchestrator', 'No specific trigger - orchestrator will analyze'],
-		];
-	}
-
-	return matches.slice(0, 3);
-}
-
-function formatAgentCategories(): string {
-	const lines: string[] = [];
-	for (const [category, info] of Object.entries(AGENT_CATEGORIES).sort()) {
-		const agentCount = Object.keys(info.agents).length;
-		const vetoAgents = Object.entries(info.agents)
-			.filter(([, a]) => a.can_veto)
-			.map(([n]) => n);
-		const vetoStr = vetoAgents.length > 0 ? ` [VETO: ${vetoAgents.join(', ')}]` : '';
-		lines.push(`  ${category}: ${info.description} (${agentCount} agents)${vetoStr}`);
-	}
-	return lines.join('\n');
-}
-
-function formatSkills(): string {
-	return Object.entries(SKILLS)
-		.sort()
-		.map(([name, desc]) => `  ${name}: ${desc}`)
-		.join('\n');
-}
-
-function formatWorkflowStatus(state: WorkflowState | null): string {
-	if (!state || !state.currentTask) {
-		return 'STATUS: No active task. Start with orchestrator agent';
-	}
-
-	const task = state.currentTask;
-	const agents = task.agents || {};
-
-	const executed = Object.entries(agents)
-		.filter(([, s]) => s.executed)
-		.map(([n]) => n);
-	const pending = Object.entries(agents)
-		.filter(([, s]) => !s.executed)
-		.map(([n]) => n);
-
-	return `STATUS: Task active
-  ID: ${task.id || 'unknown'}
-  Type: ${task.type || 'unknown'}
-  Description: ${task.description || 'unknown'}
-  Executed: ${executed.length > 0 ? executed.join(', ') : 'none'}
-  Pending: ${pending.length > 0 ? pending.join(', ') : 'none'}
-  Approved files: ${(task.approvedFiles || []).length}`;
-}
-
 async function readStdinWithTimeout(timeoutMs: number): Promise<string> {
 	return new Promise((resolve) => {
 		const timeout = setTimeout(() => {
@@ -773,122 +46,32 @@ async function main(): Promise<void> {
 	}
 
 	const prompt = hookInput.user_prompt || hookInput.prompt || '';
-	const state = loadWorkflowState();
-	const taskType = prompt ? detectTaskType(prompt) : 'unknown';
-	const bestAgents = prompt
-		? detectBestAgents(prompt)
-		: [['01-orchestration', 'orchestrator', 'Default'] as [string, string, string]];
-
-	const isNewFeature =
-		taskType === 'feature' ||
-		['new', 'implement', 'create', 'add'].some((w) => prompt.toLowerCase().includes(w));
-	const isUiTask = [
-		'ui',
-		'component',
-		'page',
-		'design',
-		'layout',
-		'mobile',
-		'desktop',
-		'tablet',
-	].some((w) => prompt.toLowerCase().includes(w));
-	const isDebugTask = ['bug', 'error', 'broken', 'not working', 'debug', 'fix'].some((w) =>
-		prompt.toLowerCase().includes(w)
-	);
-
-	const recommendedStr = bestAgents
-		.map(([cat, agent, reason], i) => `    ${i + 1}. ${cat}/${agent} - ${reason}`)
-		.join('\n');
-
-	const claudeMdReminder = `
-================================================================================
-CLAUDE.MD UPDATE REQUIREMENT (MANDATORY)
-================================================================================
-
-Before completing ANY task, you MUST update CLAUDE.md at project root with:
-
-1. "## Last Change" section (ONLY keep the latest, do NOT stack):
-   **Branch:** your-branch-name
-   **Date:** ${new Date().toISOString().split('T')[0]}
-   **Summary:** What you implemented/fixed
-
-2. Update "## Architecture" if structure changed
-
-3. Add new rules/patterns learned to appropriate sections
-
-4. SYNTHESIZE: If user mentioned preferences or corrections, add as rules
-
-CHARACTER LIMIT: 40,000 max. If exceeded, COMPACT the file (keep critical sections).
-
-STOP HOOK WILL BLOCK if CLAUDE.md is not properly updated!
-================================================================================
-`;
-
-	const output = `
-================================================================================
-STRICT WORKFLOW ENFORCEMENT - UserPromptSubmit Hook
-================================================================================
-${STRICT_WORKFLOW}
-${claudeMdReminder}
-
-${isNewFeature ? '⚠️  NEW FEATURE DETECTED - RESEARCH AGENT IS MANDATORY!' : ''}
-${isUiTask ? '⚠️  UI TASK DETECTED - SEPARATE UIs FOR MOBILE/TABLET/DESKTOP REQUIRED!' : ''}
-${isDebugTask ? '🐛 DEBUG TASK DETECTED - Use debugger or type-error-resolver agents' : ''}
-
-================================================================================
-STEP 1: CREATE DETAILED TODO LIST (MANDATORY)
-================================================================================
-
-BEFORE doing anything, you MUST use TodoWrite to create a detailed todo list.
-Break down the user's prompt into specific, actionable items.
-
-================================================================================
-AGENT SYSTEM: 82 Specialized Agents in 14 Categories
-================================================================================
-
-CATEGORIES:
-${formatAgentCategories()}
-
-TASK ANALYSIS:
-  Detected type: ${taskType}
-  Recommended agents:
-${recommendedStr}
-  Research required: ${isNewFeature ? 'YES (MANDATORY)' : 'Optional'}
-  Separate UIs required: ${isUiTask ? 'YES (MANDATORY)' : 'N/A'}
-  Workflow sequence: ${(WORKFLOWS[taskType] || WORKFLOWS['feature']).join(' -> ')}
-
-${formatWorkflowStatus(state)}
-
-================================================================================
-SKILL SYSTEM: 22 Skills Auto-loaded by Agents
-================================================================================
-
-${formatSkills()}
-
-Skills are auto-loaded when an agent starts. Agents don't inherit parent skills.
-
-================================================================================
-AGENT INVOCATION (VIA TASK TOOL ONLY)
-================================================================================
+	if (!prompt.trim()) {
+		console.log(JSON.stringify({ continue: true }));
+		process.exit(0);
+	}
 
-You MUST use the Task tool with subagent_type to invoke agents.
-DO NOT execute agent logic manually - INVOKE the agent properly.
+	const today = new Date().toISOString().split('T')[0];
 
-Example:
-  Task(subagent_type="debugger", prompt="Fix the TypeError in user.ts")
-  Task(subagent_type="playwright-e2e", prompt="Create E2E tests for login")
+	const systemMessage = `TASK WORKFLOW (English only):
 
-================================================================================
-`;
+1. CREATE a detailed todo-list (TaskCreate) breaking down the user's request into subtopics/steps.
+2. WORK through each item sequentially — mark in_progress when starting, completed when done.
+3. COMMIT using conventional commits (feat/fix/refactor/docs/chore). Use the commit-manager agent for proper git workflow: branch → implement → commit → merge to main.
+4. UPDATE CLAUDE.md before finishing — this is the project's fast-recall memory across sessions:
+   a. "## Last Change" section (date: ${today}, branch, summary). Keep only the latest, never stack.
+   b. If you changed how the app works (new flows, changed workflows, new patterns): update the relevant CLAUDE.md sections (Architecture, Workflow, Rules, UI patterns, etc.) so the next session knows.
+   c. If you added new rules, gotchas, or forbidden patterns: add them to the appropriate section.
+   d. If you changed config, stack, or tooling: update the Stack/Configuration sections.
+   CLAUDE.md is the single source of truth for quick context. Anything not recorded here will be forgotten next session.
+5. RUN stop-validator before finishing: npx tsx .claude/hooks/stop-validator.ts`;
 
-	const result = { continue: true, systemMessage: output };
-	console.log(JSON.stringify(result));
+	console.log(JSON.stringify({ continue: true, systemMessage }));
 	process.exit(0);
 }
 
 main().catch((err) => {
 	console.error('Hook error:', err);
-	const result = { continue: true, systemMessage: 'Hook error occurred, continuing...' };
-	console.log(JSON.stringify(result));
+	console.log(JSON.stringify({ continue: true, systemMessage: 'Hook error, continuing...' }));
 	process.exit(0);
 });

package/template/.claude/skills/codebase-knowledge/domains/claude-system.md

@@ -2,9 +2,9 @@
 
 ## Last Update
 
-- **Date:** 2026-01-06
-- **Commit:** 77e2263
-- **Session:** Fixed stop hook exit code from 2 to 0 for proper JSON processing
+- **Date:** 2026-02-04
+- **Commit:** pending
+- **Session:** Simplified user-prompt-submit hook (895→73 lines) and enhanced CLAUDE.md update instructions to cover flows, rules, architecture — not just Last Change
 
 ## Files
 
@@ -36,8 +36,7 @@
 - `.claude/hooks/run-hook.cmd` - **NEW** Batch wrapper for Windows
 - `.claude/hooks/stop-validator.py` - Stop hook enforcement: branch protection + documentation check
 - `.claude/hooks/stop-validator.ts` - **NEW** TypeScript fallback version of stop-validator
-- `.claude/hooks/user-prompt-submit.py` - Agent selection guidance before prompt processing
-- `.claude/hooks/user-prompt-submit.ts` - **NEW** TypeScript fallback version of user-prompt-submit
+- `.claude/hooks/user-prompt-submit.ts` - Concise 5-step workflow instructions (simplified from 895 to 73 lines)
 - `.claude/hooks/check-documentation.py` - Documentation verification hook
 - `.claude/hooks/SETUP.md` - Hook system documentation