{"id":36646690,"url":"https://github.com/popup-studio-ai/bkit-claude-code","last_synced_at":"2026-04-01T16:51:21.422Z","repository":{"id":332002248,"uuid":"1130876747","full_name":"popup-studio-ai/bkit-claude-code","owner":"popup-studio-ai","description":"bkit Vibecoding Kit - PDCA methodology + Claude Code mastery for AI-native development","archived":false,"fork":false,"pushed_at":"2026-03-18T09:50:30.000Z","size":10495,"stargazers_count":399,"open_issues_count":0,"forks_count":104,"subscribers_count":2,"default_branch":"main","last_synced_at":"2026-03-18T20:42:48.762Z","etag":null,"topics":["ai-agents","ai-driven-work","ai-native-development","bkit","claude-code","multi-language","pdca","pipeline","plugin-repo"],"latest_commit_sha":null,"homepage":"https://bkit.ai","language":"JavaScript","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/popup-studio-ai.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":"NOTICE","maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-01-09T06:25:08.000Z","updated_at":"2026-03-18T09:49:50.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/popup-studio-ai/bkit-claude-code","commit_stats":null,"previous_names":["popup-studio-ai/bkit-claude-code"],"tags_count":40,"template":false,"template_full_name":null,"purl":"pkg:github/popup-studio-ai/bkit-claude-code","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/popup-studio-ai%2Fbkit-claude-code","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/popup-studio-ai%2Fbkit-claude-code/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/popup-studio-ai%2Fbkit-claude-code/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/popup-studio-ai%2Fbkit-claude-code/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/popup-studio-ai","download_url":"https://codeload.github.com/popup-studio-ai/bkit-claude-code/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/popup-studio-ai%2Fbkit-claude-code/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31290538,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-01T13:12:26.723Z","status":"ssl_error","status_checked_at":"2026-04-01T13:12:25.102Z","response_time":53,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-agents","ai-driven-work","ai-native-development","bkit","claude-code","multi-language","pdca","pipeline","plugin-repo"],"created_at":"2026-01-12T10:01:04.163Z","updated_at":"2026-04-01T16:51:21.410Z","avatar_url":"https://github.com/popup-studio-ai.png","language":"JavaScript","funding_links":[],"categories":[],"sub_categories":[],"readme":"# bkit - Vibecoding Kit\n\n[![License](https://img.shields.io/badge/License-Apache%202.0-blue.svg)](https://opensource.org/licenses/Apache-2.0)\n[![Claude Code](https://img.shields.io/badge/Claude%20Code-v2.1.81+-purple.svg)](https://code.claude.com)\n[![Version](https://img.shields.io/badge/Version-2.0.6-green.svg)](CHANGELOG.md)\n[![Author](https://img.shields.io/badge/Author-POPUP%20STUDIO-orange.svg)](https://popupstudio.ai)\n\n\u003e **PDCA methodology + CTO-Led Agent Teams + AI coding assistant mastery for AI-native development**\n\nbkit is a Claude Code plugin that transforms how you build software with AI. It provides structured development workflows, automatic documentation, and intelligent code assistance through the PDCA (Plan-Do-Check-Act) methodology.\n\n![bkit Introduction](images/bkit-intro.png)\n\n---\n\n## What is Context Engineering?\n\n**Context Engineering** is the systematic curation of context tokens for optimal LLM inference—going beyond simple prompt crafting to build entire systems that consistently guide AI behavior.\n\n```\nTraditional Prompt Engineering:\n  \"The art of writing good prompts\"\n\nContext Engineering:\n  \"The art of designing systems that integrate prompts, tools, and state\n   to provide LLMs with optimal context for inference\"\n```\n\n**bkit is a practical implementation of Context Engineering**, providing a systematic context management system for Claude Code.\n\n### bkit's Context Engineering Architecture\n\n![bkit Context Engineering Architecture](images/bkit-architecture.png)\n\nbkit implements Context Engineering through three interconnected layers:\n\n| Layer | Components | Purpose |\n|-------|------------|---------|\n| **Domain Knowledge** | 37 Skills | Structured expert knowledge (phases, levels, specialized domains) |\n| **Behavioral Rules** | 32 Agents | Role-based constraints with model selection (opus/sonnet/haiku) |\n| **State Management** | ~620+ Functions | PDCA state machine, workflow engine, automation control, audit, quality gates, intent detection, team coordination |\n\n### 6-Layer Hook System\n\nContext injection occurs at six distinct layers:\n\n```\nLayer 1: hooks.json (Global)     → SessionStart, UserPromptSubmit, PreCompact, PostCompact, PreToolUse, PostToolUse, Stop, StopFailure + 10 more (18 events)\nLayer 2: Skill Frontmatter       → Domain-specific hooks (deprecated in v1.4.4, use hooks.json)\nLayer 3: Agent Frontmatter       → Task-specific hooks with constraints\nLayer 4: Description Triggers    → Semantic matching in 8 languages\nLayer 5: Scripts (21 modules)    → Actual Node.js execution logic with unified handlers\nLayer 6: Plugin Data Backup      → ${CLAUDE_PLUGIN_DATA} persistent state management\n```\n\n\u003e **Learn more**: See [Context Engineering Principles](bkit-system/philosophy/context-engineering.md) for detailed implementation.\n\n---\n\n## Features\n\n![bkit Features](images/bkit-features.png)\n\n- **CC v2.1.86 Compatibility + Skills Discoverability (v2.0.8)** - 34 skills description optimized for 250-char cap (/skills listing), Hook `if` conditional field documentation (CC v2.1.85+), Enterprise org policy documentation, CC v2.1.86+ recommended, 52 consecutive compatible releases (v2.1.34~v2.1.86)\n- **Living Context System + Self-Healing + PDCA Handoff Fix (v2.0.6)** - 4-Layer Living Context (`lib/context/` 7 modules), Self-Healing agent (opus) for automated error recovery, Deploy skill with 3-environment state machine, PDCA Handoff Loss Fix Phase 2+3 (PRD→Code context preservation 30-40% → 75-85%), 11 infrastructure templates (ArgoCD, Terraform, observability), 88 lib modules, 37 skills, 32 agents, 57 scripts\n- **PM Skills Integration + Interactive Checkpoints (v2.0.3)** - PM Agent Team expanded from 9→43 frameworks ([pm-skills](https://github.com/phuryn/pm-skills) MIT), PDCA Interactive Checkpoints 1~5 (AskUserQuestion-gated), code-analyzer Confidence-Based Filtering (≥80%), Design phase 3 Architecture Options (Minimal/Clean/Pragmatic), btw CTO Team Integration, pm-prd template v2.0 with Execution Deliverables (Pre-mortem, User/Job Stories, Test Scenarios, Stakeholder Map, Growth Loops)\n- **Cross-Project Isolation Fix (v2.0.1)** - PLUGIN_DATA backup/restore project identity guard via meta.json, globalCache project namespace, prevents cross-project PDCA state leakage ([#48](https://github.com/popup-studio-ai/bkit-claude-code/issues/48))\n- **AI Native Development OS (v2.0.0)** - Declarative PDCA state machine (20 transitions), YAML workflow DSL (3 presets), L0-L4 controllable AI automation, CLI dashboard (progress-bar, workflow-map, control-panel), audit logging + decision tracing, quality gates (7 stages), checkpoint/rollback, destructive operation detector (8 rules), MCP servers (bkit-pdca + bkit-analysis), 3,175+ TC, ~620+ exports, 88 lib modules, 37 skills, 32 agents, 18 hook events\n- **CC v2.1.78 Full Integration (v1.6.2)** - 14 ENH items (ENH-117~130), PostCompact/StopFailure hooks, ${CLAUDE_PLUGIN_DATA} persistent backup, agent effort/maxTurns, 1M context default, 128K output, 1186 TC (99.7%), 260+ exports, 54 scripts, CC v2.1.78 recommended\n- **CTO Orchestration Redesign + Quality Hardening (v1.6.1)** - Main Session as CTO pattern (CC v2.1.69+ compatibility), P0 bug fixes (4), Config-Code synchronization, 3-Tier Agent Security Model, 1073 TC comprehensive test (99.6% pass), CE Level 5 assessment (88/100), 72 files ~1,400 LOC\n- **Skills 2.0 Complete Integration (v1.6.0)** - 19 ENH items (ENH-85~103), Skill Evals framework with 28 eval definitions, Skill Classification (Workflow/Capability/Hybrid), A/B testing, template-validator, frontmatter hooks migration, context:fork deprecation, PM Agent Team integration\n- **Executive Summary \u0026 Preview UX (v1.5.9)** - Auto-generated 4-perspective summaries (Problem/Solution/Function \u0026 UX Effect/Core Value), AskUserQuestion with rich Markdown previews, ENH-74~81 CC v2.1.69 compatibility, 199 exports\n- **Studio Support \u0026 Path Registry (v1.5.8)** - Centralized state file path management (`lib/core/paths.js`), PDCA doc path registry, config cleanup (dead keys removed, missing keys added), state directory migration to `.bkit/{state,runtime,snapshots}/`, auto-migration with EXDEV fallback, 190 exports\n- **/simplify + /batch PDCA Integration (v1.5.7)** - CC built-in /simplify and /batch commands integrated into PDCA Check→Report flow, CC_COMMAND_PATTERNS 8-language awareness, HTTP Hooks documentation, English conversion for 3 stop scripts\n- **Auto-Memory Integration (v1.5.6)** - CC v2.1.59 auto-memory official support, /copy guidance, multi-agent memory optimization, ENH-48~51 enhancements, 186 exports\n- **Plan Plus Skill (v1.5.5)** - Brainstorming-enhanced PDCA planning with intent discovery, alternatives exploration, and YAGNI review\n- **bkend MCP Accuracy Fix (v1.5.4)** - MCP tool coverage 19→28+, accurate tool names, dynamic Base URL, search_docs workflow\n- **Team Visibility \u0026 State Writer (v1.5.3)** - Agent Teams state management with `.bkit/agent-state.json` for Studio IPC\n- **SubagentStart/SubagentStop Hooks (v1.5.3)** - 2 new hook events for agent lifecycle tracking (10 hook events total)\n- **Output Styles Auto-Discovery (v1.5.3)** - `outputStyles` in plugin.json + 4th style `bkit-pdca-enterprise`\n- **CTO-Led Agent Teams (v1.5.1)** - CTO agent orchestrates parallel PDCA execution with multi-agent teams (Dynamic: 3, Enterprise: 5 teammates)\n- **Output Styles (v1.5.1)** - Level-based response formatting (bkit-learning, bkit-pdca-guide, bkit-enterprise, bkit-pdca-enterprise)\n- **Agent Memory (v1.5.1)** - Cross-session context persistence for all 32 agents (auto-active)\n- **Natural Feature Discovery (v1.5.1)** - Auto-trigger integration aligned with \"Automation First\" philosophy\n- **Task Management + PDCA Integration (v1.4.7)** - Task Chain Auto-Creation, Task ID Persistence, Check↔Act Iteration\n- **Core Modularization (v1.4.7)** - lib/common.js split into 5 modules (lib/core/, lib/pdca/, lib/intent/, lib/task/, lib/team/)\n- **Context Engineering (v1.4.4)** - Systematic context curation with 7 library modules and unified hook system\n- **PDCA Methodology** - Structured development workflow with automatic documentation\n- **PDCA Skill Integration (v1.4.4)** - Unified `/pdca` skill with 8 actions (plan, design, do, analyze, iterate, report, status, next)\n- **Evaluator-Optimizer Pattern** - Automatic iteration cycles from Anthropic's agent architecture\n- **9-Stage Development Pipeline** - From schema design to deployment\n- **3 Project Levels** - Starter (static), Dynamic (fullstack), Enterprise (microservices)\n- **Multilingual Support** - 8 languages (EN, KO, JA, ZH, ES, FR, DE, IT)\n- **37 Skills** - Domain-specific knowledge (18 Workflow / 18 Capability / 1 Hybrid)\n- **32 Agents** - Specialized AI assistants (11 opus / 19 sonnet / 2 haiku) including CTO/PM Team + PDCA Eval agents\n- **21 Hook Scripts** - Hook execution with unified handlers across 18 event types\n- **~620+ Utility Functions** - 88 lib modules across 11 subdirectories (core, pdca, intent, task, team, ui, audit, control, quality, adapters, context)\n- **Check-Act Iteration Loop** - Automatic gap analysis and fix cycles with max 5 iterations (90% threshold)\n\n---\n\n## Skill Evals: Data-Driven Skill Quality Management\n\nClaude Code introduced **Skill Evals** in Skills 2.0—a framework for measuring skill quality through automated testing. bkit extends this concept into a **complete skill lifecycle management system** that answers a question no other plugin addresses: *\"Are my skills still worth keeping?\"*\n\n### What Are Skill Evals?\n\nSkill Evals run automated quality checks against skills by sending test prompts and comparing outputs against expected results. Think of them as **unit tests for AI skills**—they catch regressions when models update and measure whether a skill still adds value.\n\n### How bkit Enhances Skill Evals\n\nbkit builds three layers on top of Claude Code's native Evals:\n\n| Layer | Claude Code Native | bkit Enhancement |\n|-------|-------------------|------------------|\n| **Eval Execution** | Basic eval runner | `evals/runner.js` with benchmark mode, 29 pre-built eval definitions |\n| **A/B Testing** | Not available | `evals/ab-tester.js` compares skill performance across models (e.g., Sonnet 4.6 vs Opus 4.6) |\n| **Skill Classification** | Not available | All 37 skills classified as Workflow (18) / Capability (18) / Hybrid (1) with deprecation-risk scoring |\n\n```\nevals/\n├── config.json              # Global settings (thresholds, classifications)\n├── runner.js                # Eval execution engine (CLI + module)\n├── reporter.js              # Markdown/JSON result reporting\n├── ab-tester.js             # Model comparison + parity testing\n├── workflow/{10 skills}/     # Eval definitions for permanent skills\n├── capability/{18 skills}/  # Eval definitions for model-dependent skills\n└── hybrid/{1 skill}/        # Eval definitions for single dual-purpose skill\n```\n\n### Skill Classification \u0026 Lifecycle\n\nNot all skills age the same way. bkit classifies each skill to manage its lifecycle:\n\n| Classification | Count | Purpose | What Evals Measure |\n|---------------|:-----:|---------|-------------------|\n| **Workflow** | 17 | Process automation (PDCA, pipelines) | Quality regression only—these skills are permanent |\n| **Capability** | 18 | Model ability augmentation (mockups, APIs) | **Parity testing**—can the model match this skill's output without it? |\n| **Hybrid** | 1 | Both process + capability | Both regression and parity |\n\nWhen a model upgrade makes a Capability skill redundant, the **Model Parity Test** detects it:\n\n```bash\n# Does the model now produce equivalent results without this skill?\nnode evals/ab-tester.js --parity phase-3-mockup --model claude-opus-4-6\n\n# Compare skill performance between two models\nnode evals/ab-tester.js --skill pdca --modelA claude-sonnet-4-6 --modelB claude-opus-4-6\n\n# Run all 29 skill evaluations\nnode evals/runner.js --benchmark\n```\n\n### What Changes for Users\n\n| Before (v1.5.9) | After (v1.6.1 with Evals) |\n|-----------------|--------------------------|\n| 28 skills, no quality measurement | 37 skills, 29 with automated eval definitions |\n| No way to know if a skill degraded after model update | Benchmark detects regression across all skills |\n| Manual judgment on skill usefulness | Data-driven deprecation recommendations |\n| Skills accumulate indefinitely | Skill lifecycle: create → eval → deprecate → remove |\n| \"Does this skill help?\" is a guess | Parity test gives a quantified answer |\n\n### Integration with PDCA\n\nSkill Evals connect directly to bkit's PDCA workflow:\n\n- **Skill Creator** (`skill-creator/`) generates new skills with eval templates pre-included\n- **Template Validator** ensures PDCA documents maintain required sections\n- **Classification metadata** in every SKILL.md frontmatter enables automated lifecycle decisions\n- **Quarterly benchmarks** track skill quality trends over time\n\n\u003e **Philosophy**: bkit's third principle is *\"No Guessing.\"* Skill Evals replace intuition with measurement—you never have to guess whether a skill is still earning its place in your workflow.\n\n---\n\n### New to Claude Code?\n\n\u003e **First time using Claude Code?**\n\u003e\n\u003e Start with [bkit-starter](https://github.com/popup-studio-ai/bkit-starter)!\n\u003e\n\u003e - Beginner-friendly guide\n\u003e - No programming experience required\n\u003e - Build your first project hands-on\n\u003e\n\u003e ```bash\n\u003e /plugin enable bkit-starter\n\u003e ```\n\u003e\n\u003e bkit is the advanced extension designed for users who have mastered bkit-starter.\n\n---\n\n## Requirements\n\n| Requirement | Minimum Version | Notes |\n|-------------|:---------------:|-------|\n| **Claude Code** | **v2.1.78+** | Required. bkit v2.0.4 uses agent frontmatter (effort/maxTurns/disallowedTools), 18 hook events, MCP servers, and ${CLAUDE_PLUGIN_DATA}. Recommended: v2.1.81+. |\n| Node.js | v18+ | For hook script execution |\n| Agent Teams (optional) | Set `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1` | Required only for CTO-Led Agent Teams feature |\n\n\u003e **Troubleshooting**: If you see `\"Failed to load hooks\"` error after installation, update Claude Code to the latest version:\n\u003e ```bash\n\u003e claude update\n\u003e ```\n\n---\n\n## Quick Start\n\n\u003e **Note**: bkit is designed for **Claude Code**. For Gemini CLI, see [bkit-gemini](https://github.com/popup-studio-ai/bkit-gemini).\n\n### Option 1: Marketplace Installation (Recommended)\n\nThe easiest way to install bkit is through the Claude Code marketplace.\n\n```bash\n# Step 1: Add bkit marketplace\n/plugin marketplace add popup-studio-ai/bkit-claude-code\n\n# Step 2: Install bkit plugin\n/plugin install bkit\n```\n\n#### Managing Marketplaces\n\nUse `/plugin` command and navigate to **Marketplaces** tab to manage your plugin sources:\n\n![bkit Marketplace](images/bkit-marketplace.png)\n\n- **bkit-marketplace**: Contains bkit and bkit-starter plugins\n- **claude-plugins-official**: Official Anthropic plugins\n\n#### Discovering Plugins\n\nNavigate to **Discover** tab to browse and install available plugins:\n\n![bkit Marketplace Plugins](images/bkit-marketplace-plugins.png)\n\n| Plugin | Description | Best For |\n|--------|-------------|----------|\n| **bkit** | Full PDCA methodology + Claude Code mastery | Experienced developers |\n| **bkit-starter** | Korean learning guide for beginners | First-time Claude Code users |\n\n#### Auto-Update Configuration\n\nKeep your plugins up-to-date automatically by configuring auto-update in your settings:\n\n```json\n// ~/.claude/settings.json\n{\n  \"plugins\": {\n    \"autoUpdate\": true\n  }\n}\n```\n\n**Update Commands:**\n- Press `u` in the Marketplaces view to update all plugins\n- Press `r` to remove a marketplace\n- Use `Space` to toggle plugin selection in Discover view\n\n### Plugin Structure\n\n```\nbkit-claude-code/\n├── .claude-plugin/\n│   ├── plugin.json          # Claude Code plugin manifest\n│   └── marketplace.json     # Marketplace registry\n├── agents/                  # Specialized AI agents\n├── commands/                # CLI command definitions\n├── skills/                  # Domain knowledge\n├── hooks/                   # Event hooks (hooks.json)\n├── evals/                   # Skill eval definitions \u0026 runner\n├── scripts/                 # Hook execution scripts\n├── servers/                 # MCP servers (bkit-pdca, bkit-analysis)\n├── lib/                     # Shared utilities (88 modules across 11 subdirs)\n├── output-styles/           # Level-based response formatting\n├── templates/               # Document templates\n└── bkit.config.json         # Centralized configuration\n```\n\n---\n\n## Customization\n\nAfter installing bkit via the marketplace, you can customize any component by copying it to your project's `.claude/` folder.\n\n\u003e **Comprehensive Guide**: See **[CUSTOMIZATION-GUIDE.md](CUSTOMIZATION-GUIDE.md)** for detailed instructions on customizing bkit for your organization, including platform-specific paths, component examples, and license attribution requirements.\n\n### How It Works\n\nClaude Code searches for configuration files in this priority order:\n1. **Project `.claude/`** (highest priority - your customizations)\n2. **User `~/.claude/`**\n3. **Plugin installation** (default)\n\n### Customization Steps\n\n```bash\n# Step 1: Find the plugin installation location\nls ~/.claude/plugins/bkit/\n\n# Step 2: Copy only the files you want to customize\nmkdir -p .claude/skills/starter\ncp ~/.claude/plugins/bkit/skills/starter/SKILL.md .claude/skills/starter/\n\n# Step 3: Edit the copied file in your project\n# Your project's .claude/skills/starter/SKILL.md will override the plugin's version\n\n# Step 4: Commit to version control (optional)\ngit add .claude/\ngit commit -m \"feat: customize bkit starter skill\"\n```\n\n### Available Components for Customization\n\n| Component | Location | Description |\n|-----------|----------|-------------|\n| **Skills** | `~/.claude/plugins/bkit/skills/` | Domain knowledge, context and slash commands (e.g., `/pdca plan`) |\n| **Agents** | `~/.claude/plugins/bkit/agents/` | Specialized AI assistants |\n| **Templates** | `~/.claude/plugins/bkit/templates/` | Document templates |\n| **Scripts** | `~/.claude/plugins/bkit/scripts/` | Hook scripts |\n| **Config** | `~/.claude/plugins/bkit/bkit.config.json` | Central configuration |\n\n### Important Notes\n\n- **Customized files don't receive plugin updates.** When bkit is updated, your customized files remain unchanged while non-customized files are updated automatically.\n- **Check the [CHANGELOG](CHANGELOG.md)** periodically for updates that might affect your customizations.\n- **Delete a customized file** to revert to the plugin's default version.\n- **Attribution required**: When creating derivative plugins, follow the [License \u0026 Attribution](CUSTOMIZATION-GUIDE.md#license--attribution) guidelines.\n\n---\n\n## Usage\n\n### Start Learning\n```bash\n/claude-code-learning\n```\n\n### Initialize a Project\n```bash\n/starter      # Static website (Starter level)\n/dynamic      # Fullstack with BaaS (Dynamic level)\n/enterprise   # Microservices with K8s (Enterprise level)\n```\n\n### PDCA Workflow (v2.0.3)\n\n\u003e **v2.0.3**: Each phase now includes Interactive Checkpoints that pause for user confirmation before proceeding. Plan confirms requirements, Design presents 3 architecture options, Do confirms implementation scope, Check offers fix strategy choices.\n\n```bash\n/pdca pm {feature}       # PM analysis \u0026 PRD generation (43 frameworks)\n/pdca plan {feature}     # Create plan document (Checkpoint 1-2)\n/pdca design {feature}   # Create design document (Checkpoint 3: 3 architecture options)\n/pdca do {feature}       # Implementation guide\n/pdca analyze {feature}  # Run gap analysis\n/pdca iterate {feature}  # Auto-fix with Evaluator-Optimizer pattern\n/pdca report {feature}   # Generate completion report\n/pdca status             # Check current PDCA status\n/pdca next               # Guide to next PDCA step\n```\n\n### CTO-Led Agent Teams (v1.5.1)\n\nCTO-Led Agent Teams enable parallel PDCA execution with multiple AI agents orchestrated by a CTO lead agent.\n\n```bash\n# Start CTO Team for a feature\n/pdca team {feature}\n\n# Monitor team progress\n/pdca team status\n\n# Cleanup team resources\n/pdca team cleanup\n```\n\n**How it works:**\n1. CTO lead agent (opus) analyzes the feature and selects the optimal team composition\n2. Teammates are spawned in parallel (Dynamic: 3, Enterprise: 5 agents)\n3. Each teammate handles a specific area (QA, frontend, backend, security, etc.)\n4. CTO orchestrates task assignment, progress monitoring, and result aggregation\n5. Team is cleaned up after work is complete\n\n**Requirements:**\n- Set environment variable: `CLAUDE_CODE_EXPERIMENTAL_AGENT_TEAMS=1`\n- Claude Code v2.1.32+\n\n**Available Team Agents:**\n\n| Team | Agent | Model | Role |\n|------|-------|-------|------|\n| CTO | cto-lead | opus | Team orchestration, PDCA workflow management |\n| CTO | frontend-architect | sonnet | UI/UX design, component architecture |\n| CTO | product-manager | sonnet | Requirements analysis, feature prioritization |\n| CTO | qa-strategist | sonnet | Test strategy, quality metrics coordination |\n| CTO | security-architect | opus | Vulnerability analysis, auth design review |\n| PM | pm-lead | opus | PM Team orchestration, PRD synthesis |\n| PM | pm-discovery | sonnet | Opportunity Solution Tree analysis |\n| PM | pm-strategy | sonnet | Value Proposition, Lean Canvas |\n| PM | pm-research | sonnet | Personas, competitors, market sizing |\n| PM | pm-prd | sonnet | PRD document generation |\n\n### PM Agent Team (v1.6.0)\n\nPM Agent Team runs **before** the Plan phase to produce a comprehensive PRD (Product Requirements Document) through automated product discovery.\n\n```bash\n# Run PM analysis before planning\n/pdca pm user-authentication\n\n# Then proceed with PDCA planning (PRD auto-referenced)\n/pdca plan user-authentication\n```\n\n**How it works:**\n1. pm-lead (opus) collects project context and git history\n2. Three agents run in parallel with 43 frameworks:\n   - discovery: Opportunity Solution Tree + Brainstorm + Assumption Risk Assessment\n   - strategy: JTBD + Lean Canvas + SWOT + PESTLE + Porter's Five Forces + Growth Loops\n   - research: Personas + Competitors + TAM/SAM/SOM + Customer Journey Map + ICP\n3. pm-prd synthesizes all findings into a PRD with 6 sections + Execution Deliverables (Pre-mortem, User/Job Stories, Test Scenarios, Stakeholder Map, Battlecards) at `docs/00-pm/{feature}.prd.md`\n4. Plan phase automatically references the PRD for higher-quality planning\n\n**Frameworks**: Based on [pm-skills](https://github.com/phuryn/pm-skills) by Pawel Huryn (MIT License)\n\n---\n\n## Project Levels\n\n| Level | Description | Stack |\n|-------|-------------|-------|\n| **Starter** | Static websites, portfolios | HTML, CSS, JS |\n| **Dynamic** | Fullstack applications | Next.js, BaaS |\n| **Enterprise** | Microservices architecture | K8s, Terraform, MSA |\n\n---\n\n## Is bkit Only for Development?\n\n![bkit for Non-Development](images/to-use-non-development.png)\n\nbkit is **primarily designed for software development**. However, some components can inspire structured workflows beyond coding:\n\n| Component | Beyond Development Uses |\n|-----------|------------------------|\n| **PDCA Methodology** | Project management, process improvement |\n| **Document Templates** | Planning any structured project |\n| **Gap Analysis** | Comparing any plan vs. actual outcome |\n\n\u003e **Note**: For general writing, research, or non-technical tasks, plain Claude Code (without bkit) is better suited.\n\n---\n\n## Documentation\n\n### Customization Guide\n\n- **[CUSTOMIZATION-GUIDE.md](CUSTOMIZATION-GUIDE.md)** - Complete guide to customizing bkit for your organization\n  - Platform-specific configuration paths (macOS, Linux, Windows, WSL)\n  - Component customization (agents, skills, commands, hooks, templates)\n  - License attribution requirements for derivative works\n  - bkit design philosophy and architecture decisions\n\n### Current Reference (bkit-system/)\n\n- **[System Architecture](bkit-system/README.md)** - Plugin structure and trigger system overview\n- **[Context Engineering](bkit-system/philosophy/context-engineering.md)** - LLM context curation principles (v1.4.2)\n- **[Core Mission \u0026 Philosophy](bkit-system/philosophy/core-mission.md)** - 3 core philosophies (Automation First, No Guessing, Docs=Code)\n- **[AI-Native Principles](bkit-system/philosophy/ai-native-principles.md)** - AI-Native development and 3 core competencies\n- **[PDCA Methodology](bkit-system/philosophy/pdca-methodology.md)** - PDCA cycle and 9-stage pipeline relationship\n- **[Graph Index](bkit-system/_GRAPH-INDEX.md)** - Obsidian-optimized component graph\n\n### Component Reference\n\n- [Development Pipeline](skills/development-pipeline/SKILL.md) - 9-stage pipeline skill\n- [Skills Reference](skills/) - 36 domain skills (17 Workflow / 18 Capability / 1 Hybrid)\n- [Agents Reference](agents/) - 31 specialized agents (10 opus / 19 sonnet / 2 haiku)\n\n### PDCA Documents\n\n- [Active PDCA](docs/pdca/) - Current plan/design/analysis documents\n- [Archive](docs/archive/) - Completed PDCA + legacy documents\n\n### Other\n\n- **[Changelog](CHANGELOG.md)** - Version history and release notes\n\n### Visualize with Obsidian\n\nThe `bkit-system/` documentation is optimized for [Obsidian](https://obsidian.md/)'s Graph View:\n\n1. Open `bkit-system/` as an Obsidian vault\n2. Press `Ctrl/Cmd + G` to open Graph View\n3. Explore component relationships visually\n\nSee **[bkit-system/README.md](bkit-system/README.md#viewing-with-obsidian)** for detailed instructions.\n\n---\n\n## Language Support\n\nbkit automatically detects your language from trigger keywords:\n\n| Language | Trigger Keywords |\n|----------|-----------------|\n| English | static website, beginner, API design |\n| Korean | 정적 웹, 초보자, API 설계 |\n| Japanese | 静的サイト, 初心者, API設計 |\n| Chinese | 静态网站, 初学者, API设计 |\n| Spanish | sitio web estático, principiante |\n| French | site web statique, débutant |\n| German | statische Webseite, Anfänger |\n| Italian | sito web statico, principiante |\n\n### Setting Response Language\n\nClaude Code supports configuring your preferred response language through the `language` setting in your settings file.\n\n#### Configuration Files (Priority Order)\n\n| File | Scope | Git Tracked |\n|------|-------|-------------|\n| `.claude/settings.local.json` | Project (personal) | No (gitignored) |\n| `.claude/settings.json` | Project (shared) | Yes |\n| `~/.claude/settings.json` | User (global) | N/A |\n\n#### How to Configure\n\nAdd the `language` key to any settings file:\n\n```json\n{\n  \"language\": \"korean\"\n}\n```\n\n#### Supported Languages\n\n| Language | Setting Value |\n|----------|---------------|\n| English | `\"english\"` (default) |\n| Korean | `\"korean\"` |\n| Japanese | `\"japanese\"` |\n| Chinese | `\"chinese\"` |\n| Spanish | `\"spanish\"` |\n| French | `\"french\"` |\n| German | `\"german\"` |\n| Italian | `\"italian\"` |\n\n\u003e **Note**: Trigger keywords work in any language. The `language` setting only affects Claude's response language.\n\n---\n\n## Contributing\n\nWe welcome contributions! Please see [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n### Branch Protection\n\n- Only `admin` team members can merge to `main`\n- All changes require pull request review\n- Version releases are managed through Git tags\n\n---\n\n## Design Philosophy\n\nbkit is not a productivity hack. It is an attempt to bring **engineering discipline** to AI-native development.\n\nThe software industry has spent decades refining how humans write code—version control, code review, CI/CD, testing pyramids. But when AI enters the development loop, most of that discipline evaporates. Developers prompt, accept, and ship. The feedback loop disappears. Documentation becomes an afterthought. Quality becomes a matter of luck.\n\n**bkit exists because we believe AI-assisted development deserves the same rigor as traditional engineering.**\n\n### What we optimize for\n\n- **Process over output.** A single feature built through proper planning, design, implementation, and verification is worth more than ten features hacked together. The PDCA cycle is not overhead—it is the product.\n\n- **Verification over trust.** AI generates plausible code. Plausible is not correct. Every implementation goes through gap analysis against its design document. If the match rate falls below 90%, the system iterates automatically. We do not ship hope.\n\n- **Context over prompts.** A well-structured prompt helps once. A well-structured context system helps every time. bkit's ~620+ functions across 88 modules, 37 skills, and 32 agents exist to ensure the AI receives the right context at the right moment—not through clever prompting, but through systematic engineering.\n\n- **Constraints over features.** We intentionally limit what bkit does. Three project levels, not infinite configuration. A fixed 9-stage pipeline, not a customizable workflow builder. Opinionated defaults, not a framework for frameworks. Constraints eliminate decision fatigue and make the system learnable.\n\n### What this means in practice\n\nWhen you use bkit, you will write a plan document before writing code. You will generate a design specification before implementation. You will run gap analysis after every feature. You will produce a completion report that captures what was built, what was verified, and what was learned.\n\nThis is slower than prompting and shipping. It is also how software that lasts gets built.\n\n\u003e *\"We do not offer a hundred features. We engineer each one through proper design and verification. That is the difference between a tool and a discipline.\"*\n\n---\n\n## License\n\nCopyright 2024-2026 POPUP STUDIO PTE. LTD.\n\nLicensed under the Apache License, Version 2.0. See [LICENSE](LICENSE) for details.\n\nYou must include the [NOTICE](NOTICE) file in any redistribution.\n\n---\n\n## Support\n\n- **Issues**: [GitHub Issues](https://github.com/popup-studio-ai/bkit-claude-code/issues)\n- **Email**: contact@popupstudio.ai\n\n---\n\nMade with AI by [POPUP STUDIO](https://popupstudio.ai)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpopup-studio-ai%2Fbkit-claude-code","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpopup-studio-ai%2Fbkit-claude-code","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpopup-studio-ai%2Fbkit-claude-code/lists"}