[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-80243":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":23,"topics":25,"createdAt":10,"pushedAt":10,"updatedAt":46,"readmeContent":47,"aiSummary":48,"trendingCount":16,"starSnapshotCount":16,"syncStatus":49,"lastSyncTime":50,"discoverSource":51},80243,"vibecode-pro-max-kit","withkynam\u002Fvibecode-pro-max-kit","withkynam","Your AI forgets. This remembers. Spec-driven coding harness for vibecoders, product owners, CEOs and real builders — self-improving context memory, 12 agents, 32 skills. Kills context rot, ships features, not spaghetti. Claude Code & Codex. Any stack. 30 seconds","",null,"JavaScript",844,197,6,15,0,4,78,624,40,88.89,"MIT License",false,"main",[26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45],"agentic","ai-agents","ai-coding-assistant","ai-development","ai-workflow","anthropic","claude","claude-code","cli-tools","code-quality","codex","coding-agents","cursor","developer-tools","llm","openai","prompt-engineering","typescript","vibe-coding","vibecoding","2026-06-12 04:01:27","\u003Cp align=\"center\">\n  \u003Cstrong>English\u003C\u002Fstrong> |\n  \u003Ca href=\"docs\u002Fi18n\u002FREADME.zh-CN.md\">简体中文\u003C\u002Fa> |\n  \u003Ca href=\"docs\u002Fi18n\u002FREADME.ja-JP.md\">日本語\u003C\u002Fa> |\n  \u003Ca href=\"docs\u002Fi18n\u002FREADME.ko-KR.md\">한국어\u003C\u002Fa> |\n  \u003Ca href=\"docs\u002Fi18n\u002FREADME.vi-VN.md\">Tiếng Việt\u003C\u002Fa> |\n  \u003Ca href=\"docs\u002Fi18n\u002FREADME.pt-BR.md\">Portugues\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cdiv align=\"center\">\n\n\u003Ca href=\"https:\u002F\u002Fflowser.ai\">\n  \u003Cimg src=\"assets\u002Fflowser-logo.svg\" alt=\"Flowser\" width=\"120\">\n\u003C\u002Fa>\n\n*Built by world-class engineers, for vibecoders at*\u003Cbr>\n*[flowser.ai](https:\u002F\u002Fflowser.ai) — AI Agents with computers for GTM*\n\n\u003Cbr>\n\n# vibecode-pro-max-kit\n\n**Stop your AI from coding before it thinks — and forgetting every detailed prompt from you.\u003Cbr>This harness turns any AI coding agent into a spec-driven engineering team\u003Cbr>that researches, plans, ships production-grade code, and self-improves its memory to survive context-rotting even 6 months later.**\n\n\u003Cbr>\n\n\u003Cp align=\"center\">\n  \u003Cimg src=\"https:\u002F\u002Fmedia.tenor.com\u002Fq_5em_iLaxoAAAAC\u002Ftanjiro-i-water-style.gif\" alt=\"Flow like water\" width=\"480\">\n  \u003Cbr>\u003Cbr>\n  \u003Cem>\"Total Concentration — Spec Breathing, Tenth Form: Constant Flux.\u003Cbr>A continuous development cycle that grows stronger with each feature shipped.\u003Cbr>Context compounds. The flow never breaks.\"\u003C\u002Fem>\u003Cbr>\n  \u003Cstrong>— Tanjiro Kamado\u003C\u002Fstrong>\n\u003C\u002Fp>\n\n🔬 Spec-driven development for AI agents\u003Cbr>\n📋 Auto-generates PRDs, manages backlogs, routes context automatically\u003Cbr>\n🧠 Self-improving knowledge base that compounds as you ship\u003Cbr>\n⚡ Runs autonomously for hours on large tasks without losing state\u003Cbr>\n🤝 Plans and specs are shareable — devs, PMs, and stakeholders review the same artifacts\n\n\u003Cp>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fstargazers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fstars\u002Fwithkynam\u002Fvibecode-pro-max-kit\" alt=\"Stars\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fnetwork\u002Fmembers\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fforks\u002Fwithkynam\u002Fvibecode-pro-max-kit\" alt=\"Forks\">\u003C\u002Fa>\n  \u003Ca href=\"LICENSE\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flicense\u002Fwithkynam\u002Fvibecode-pro-max-kit\" alt=\"License\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fgraphs\u002Fcontributors\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Fcontributors\u002Fwithkynam\u002Fvibecode-pro-max-kit\" alt=\"Contributors\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Factions\u002Fworkflows\u002Fvalidate.yml\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Factions\u002Fworkflow\u002Fstatus\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fvalidate.yml\" alt=\"CI\">\u003C\u002Fa>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fcommits\u002Fmain\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fgithub\u002Flast-commit\u002Fwithkynam\u002Fvibecode-pro-max-kit\" alt=\"Last Commit\">\u003C\u002Fa>\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fagents-12-orange\" alt=\"Agents\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fskills-32-purple\" alt=\"Skills\">\n  \u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Ftools-7_%E2%9C%93-38BDF8\" alt=\"7 Tools\">\n\u003C\u002Fp>\n\n\u003Cp>\n  \u003Cstrong>The simplest, most flexible, team-friendly coding harness for\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fanthropics\u002Fclaude-code\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FClaude_Code-191919?style=for-the-badge&logo=anthropic&logoColor=white\" alt=\"Claude Code\">\u003C\u002Fa> \n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fopenai\u002Fcodex\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCodex_CLI-412991?style=for-the-badge&logo=openai&logoColor=white\" alt=\"Codex CLI\">\u003C\u002Fa> \n  \u003Ca href=\"https:\u002F\u002Fcursor.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCursor-000?style=for-the-badge&logo=cursor&logoColor=white\" alt=\"Cursor\">\u003C\u002Fa> \n  \u003Ca href=\"https:\u002F\u002Fwindsurf.com\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FWindsurf-09B6A2?style=for-the-badge&logo=codeium&logoColor=white\" alt=\"Windsurf\">\u003C\u002Fa>\u003Cbr>\n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fgoogle-gemini\u002Fgemini-cli\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FAntigravity-8E75B2?style=for-the-badge&logo=googlegemini&logoColor=white\" alt=\"Antigravity\">\u003C\u002Fa> \n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fopencode-ai\u002Fopencode\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FOpenCode-333?style=for-the-badge\" alt=\"OpenCode\">\u003C\u002Fa> \n  \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Ffeatures\u002Fcopilot\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FGitHub_Copilot-000?style=for-the-badge&logo=githubcopilot&logoColor=white\" alt=\"GitHub Copilot\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp>\n  \u003Cem>Works across any tech stack, any language, any project\u003C\u002Fem>\u003Cbr>\u003Cbr>\n  \u003Cpicture>\n    \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ts%2Cjs%2Creact%2Cnextjs%2Cvue%2Cnuxt%2Csvelte%2Cangular%2Cnodejs%2Cexpress%2Cbun%2Cpython%2Cdjango%2Cflask%2Cfastapi&theme=dark&perline=15\" \u002F>\n    \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ts%2Cjs%2Creact%2Cnextjs%2Cvue%2Cnuxt%2Csvelte%2Cangular%2Cnodejs%2Cexpress%2Cbun%2Cpython%2Cdjango%2Cflask%2Cfastapi&theme=light&perline=15\" \u002F>\n    \u003Cimg alt=\"Tech Stack Row 1\" src=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ts,js,react,nextjs,vue,nuxt,svelte,angular,nodejs,express,bun,python,django,flask,fastapi&perline=15\" \u002F>\n  \u003C\u002Fpicture>\n  \u003Cbr>\n  \u003Cpicture>\n    \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ruby%2Crails%2Cgo%2Crust%2Cjava%2Cspring%2Ckotlin%2Cswift%2Cphp%2Claravel%2Ccs%2Cdotnet%2Celixir%2Cgraphql%2Cprisma&theme=dark&perline=15\" \u002F>\n    \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ruby%2Crails%2Cgo%2Crust%2Cjava%2Cspring%2Ckotlin%2Cswift%2Cphp%2Claravel%2Ccs%2Cdotnet%2Celixir%2Cgraphql%2Cprisma&theme=light&perline=15\" \u002F>\n    \u003Cimg alt=\"Tech Stack Row 2\" src=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=ruby,rails,go,rust,java,spring,kotlin,swift,php,laravel,cs,dotnet,elixir,graphql,prisma&perline=15\" \u002F>\n  \u003C\u002Fpicture>\n  \u003Cbr>\n  \u003Cpicture>\n    \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=supabase%2Cfirebase%2Cpostgres%2Cmongodb%2Credis%2Cdocker%2Ckubernetes%2Caws%2Cgcp%2Cazure%2Cvercel%2Ccloudflare%2Ctailwind%2Celectron&theme=dark&perline=15\" \u002F>\n    \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=supabase%2Cfirebase%2Cpostgres%2Cmongodb%2Credis%2Cdocker%2Ckubernetes%2Caws%2Cgcp%2Cazure%2Cvercel%2Ccloudflare%2Ctailwind%2Celectron&theme=light&perline=15\" \u002F>\n    \u003Cimg alt=\"Tech Stack Row 3\" src=\"https:\u002F\u002Fskillicons.dev\u002Ficons?i=supabase,firebase,postgres,mongodb,redis,docker,kubernetes,aws,gcp,azure,vercel,cloudflare,tailwind,electron&perline=15\" \u002F>\n  \u003C\u002Fpicture>\n  \u003Cbr>\n  \u003Csub>React · Next.js · Vue · Nuxt · Svelte · Angular · React Native · Electron · Node.js · Express · Bun · Hono · Python · Django · FastAPI · Flask · Ruby · Rails · Go · Rust · Java · Spring Boot · Kotlin · Swift · PHP · Laravel · C# · .NET · Elixir · TypeScript · Prisma · Supabase · Firebase · PostgreSQL · MongoDB · Redis · GraphQL · Docker · Kubernetes · Terraform · AWS · GCP · Azure · Vercel · Cloudflare · Tailwind · shadcn\u002Fui · and any other stack your project uses\u003C\u002Fsub>\n\u003C\u002Fp>\n\n\u003C\u002Fdiv>\n\n---\n\n## 🚀 Install (30 seconds)\n\n> **Run this inside your project folder.** Open a terminal and `cd` into the project you want the harness installed into *before* running the command — it installs into the current directory.\n>\n> Prefer to drive it from your agent? Open Claude Code or Codex **with that project folder as the working directory**, then paste the [full setup prompt](#-full-agent-setup-prompt) below.\n\n```bash\ncurl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fmain\u002Finstall.sh | bash\n```\n\nThen open Claude Code and say:\n\n```\nRun vc-setup\n```\n\nThat's it. The setup skill detects your stack, asks about your project (a real conversation, not a checklist), scaffolds the process directory, deep-scans your codebase, and populates context files with actual content — not placeholders.\n\n\u003Cbr>\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>📦 What gets installed\u003C\u002Fstrong>\u003C\u002Fsummary>\n\n\u003Cbr>\n\n```\nyour-project\u002F\n├── .claude\u002F\n│   ├── agents\u002F              # 🤖 12 specialized agent definitions\n│   │   ├── vc-research-agent.md\n│   │   ├── vc-execute-agent.md\n│   │   └── ...\n│   ├── skills\u002F              # ⚡ 31 auto-discovered skills\n│   │   ├── vc-generate-plan\u002F\n│   │   ├── vc-security\u002F\n│   │   ├── vc-scout\u002F\n│   │   └── ...\n│   └── hooks\u002F               # 🪝 7 lifecycle hooks\n│       ├── privacy-block.cjs\n│       ├── scout-block.cjs\n│       └── ...\n├── .codex\u002F\n│   └── agents\u002F              # 🔄 Mirrored agents for Codex\n├── CLAUDE.md                # 📋 Orchestrator + routing rules\n├── AGENTS.md                # 📖 Agent registry\n└── process\u002F                 # 🧠 Created by vc-setup (not install)\n    └── ...\n```\n\n- **Fresh project?** Installs the full harness, then `vc-setup` studies your codebase\n- **Existing `.claude\u002F` config?** Backs up to `.vibecode-backup\u002F`, installs fresh, restores your `settings.json`\n- **Existing `process\u002F` directory?** Never touched by install — `vc-setup` handles migration intelligently\n- **Existing `CLAUDE.md`?** Backed up as `CLAUDE.md.pre-vibecode`, harness version installed\n\n\u003C\u002Fdetails>\n\n\u003Ca id=\"-full-agent-setup-prompt\">\u003C\u002Fa>\n\n\u003Cdetails>\n\u003Csummary>\u003Cstrong>🤖 Full agent setup prompt\u003C\u002Fstrong> (copy-paste this into Claude Code for maximum control)\u003C\u002Fsummary>\n\n> **First, open Claude Code or Codex with your project folder as the working directory** (launch it from inside the project, or `cd` there first). The harness installs into the current directory, so this must be your project — then paste the prompt below.\n\n```\nFirst, install the vibecode-pro-max-kit agent harness by running this command:\n\ncurl -fsSL https:\u002F\u002Fraw.githubusercontent.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fmain\u002Finstall.sh | bash\n\nAfter the install completes, run vc-setup to configure everything for this project.\n\nFollow the full interactive flow:\n\n1. DETECT — Read package.json, detect my stack (framework, package manager, monorepo\n   structure, test framework, database, auth). Also check if I have any existing .claude\u002F,\n   process\u002F, or context files from a previous setup.\n\n2. SHOW ME WHAT YOU FOUND — Present a summary of the detection results and wait for me\n   to confirm before continuing. If this is an existing project with process\u002F folders or\n   context files, tell me what you found and what looks good vs what could be improved.\n\n3. ASK ME ABOUT THE PROJECT — Before scaffolding or scanning, have a real conversation\n   with me about this project. Don't just ask a fixed list of questions and move on — ask\n   follow-ups based on my answers, probe deeper on anything vague, and keep going until\n   you genuinely understand the project. Start with the basics (what is this? who uses it?),\n   then dig into architecture, features, conventions, pain points, and anything else that\n   matters. Summarize your understanding back to me and confirm it's correct before moving on.\n\n4. SCAFFOLD — Create the process\u002F directory structure. If I already have process\u002F folders,\n   show me what you plan to change and wait for my approval before reorganizing anything.\n   Never silently move or delete my existing files.\n\n5. STUDY — Deep-scan the codebase and populate process\u002Fcontext\u002Fall-context.md with REAL\n   content based on what you find AND what I told you. Include: repo structure, tech stack\n   with versions, key patterns and conventions, import aliases, env vars, API routes,\n   database schema, test setup. Do not leave placeholder text.\n\n6. VALIDATE — Run all the validation checks to make sure everything is wired correctly.\n\nImportant rules:\n- If I have existing context files or a well-written CLAUDE.md, read them first and\n  preserve what is good. Merge intelligently — do not replace good content with generic scans.\n- Show me a summary of what you plan to create or change at each major step and wait\n  for my OK before proceeding.\n- Do not create empty placeholder files. Only create files that have real content.\n- Ask before reorganizing. If my existing setup works, tell me what you would improve\n  and let me decide.\n```\n\n\u003C\u002Fdetails>\n\n\u003Cbr>\n\n\u003Cdetails>\n\u003Csummary>Table of Contents\u003C\u002Fsummary>\n\n- [The Problem](#-the-problem)\n- [The Fix](#-the-fix)\n- [The Vibe Coding Revolution](#the-vibe-coding-revolution)\n- [Who Is This For?](#who-is-this-for)\n- [At a Glance](#at-a-glance)\n- [Why Teams Use This](#-why-teams-use-this)\n- [How This Compares](#how-this-compares)\n- [What Makes This Different](#-what-makes-this-different)\n- [What's Inside](#-whats-inside)\n- [How It Works](#-how-it-works)\n- [Built-in Safety Systems](#-built-in-safety-systems)\n- [Contributing](#contributing)\n- [Star History](#-star-history)\n\n\u003C\u002Fdetails>\n\n---\n\n## 🔥 The Problem\n\nYou ask Claude to \"add webhook support.\" It immediately starts writing code. No questions about your architecture. No check on existing patterns. No plan. You get 400 lines that don't fit your codebase, and you spend an hour fixing it.\n\n**But that's just the surface.** The deeper problems hit harder:\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🧠\u003C\u002Fh1>\n\u003Cstrong>Context dies every session\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nYour agent forgets everything it learned. Same mistakes, same questions, every time. No memory, no compounding knowledge.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>📄\u003C\u002Fh1>\n\u003Cstrong>Docs go stale instantly\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nYou wrote great context docs last week. They're already outdated. Nothing auto-updates them as the codebase evolves.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>💥\u003C\u002Fh1>\n\u003Cstrong>Big tasks collapse mid-way\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nContext window fills up, state is lost, the agent starts hallucinating. You restart from scratch on hour 3.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🤝\u003C\u002Fh1>\n\u003Cstrong>No specs, no review, no collaboration\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nYour PM can't review what the agent is about to build. There's no artifact to share, discuss, or approve before code is written.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🎭\u003C\u002Fh1>\n\u003Cstrong>Architecture decisions are hallucinated\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nThe agent invents patterns instead of researching how other codebases solved the same problem.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n**Your agent has intelligence but no process, no memory, and no way to collaborate with your team.**\n\nWhether you're a developer, a PM, or a CEO who just started vibe coding — this problem hits everyone the same way. The fix is the same too: **give your agent a real development process.**\n\n---\n\n## 🛠️ The Fix\n\nThis harness installs a complete development system into your project — not just a CLAUDE.md file, but **12 specialized agents, 31 skills**, and a phase-locked workflow that forces your agent to **understand before it builds**.\n\n\u003Cbr>\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>📋\u003C\u002Fh1>\n\u003Cstrong>Spec-driven plans\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Csub>PMs and devs review the same plan artifact before any code is written\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🔄\u003C\u002Fh1>\n\u003Cstrong>Self-improving context\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Csub>Auto-updates every time a feature ships — docs never go stale\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>⚡\u003C\u002Fh1>\n\u003Cstrong>Autonomous execution\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Csub>Survives context compaction — runs for hours, not minutes\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🧬\u003C\u002Fh1>\n\u003Cstrong>Architecture research\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Csub>Studies real codebases before making design decisions\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🧭\u003C\u002Fh1>\n\u003Cstrong>Smart context routing\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Csub>Loads only what's relevant — not your entire knowledge base every time\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n\u003Cbr>\n\n```mermaid\n%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '16px', 'lineColor': '#8888AA'}} }%%\nflowchart TD\n    R[\"🔍 RESEARCH\\nRead codebase, gather facts\"]\n    I[\"💡 INNOVATE\\nExplore 2-3 approaches\"]\n    P[\"📋 PLAN\\nWrite detailed spec\"]\n    E[\"⚡ EXECUTE\\nImplement the plan\"]\n    T[\"✅ tester → reviewer → git-manager\"]\n    U[\"🧠 UPDATE PROCESS\\nCapture learnings\"]\n\n    R -->|\"you say 'go'\"| I\n    I -->|\"you say 'go'\"| P\n    P -->|\"ENTER EXECUTE MODE\"| E\n    E --> T\n    E -->|\"recommended\"| U\n\n    style R fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n    style I fill:#E65100,stroke:#BF360C,color:#FFFFFF\n    style P fill:#2E7D32,stroke:#1B5E20,color:#FFFFFF\n    style E fill:#C62828,stroke:#B71C1C,color:#FFFFFF\n    style T fill:#6A1B9A,stroke:#4A148C,color:#FFFFFF\n    style U fill:#00695C,stroke:#004D40,color:#FFFFFF\n```\n\nEvery transition requires your **explicit approval**. Nothing auto-advances. You stay in control.\n\n---\n\n## The Vibe Coding Revolution\n\n\u003Cdiv align=\"center\">\n\u003Ch3>\u003Cem>\"The hottest new programming language is English.\"\u003C\u002Fem>\u003C\u002Fh3>\n\u003Cstrong>— Andrej Karpathy\u003C\u002Fstrong>\n\u003C\u002Fdiv>\n\n\u003Cbr>\n\n**Vibe coding changed who can build software. Spec-driven development changes what they can ship.**\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\">\n\u003Ch3>63%\u003C\u002Fh3>\n\u003Csub>of vibe coding users are \u003Cstrong>NOT\u003C\u002Fstrong> developers\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\">\n\u003Ch3>16.2M\u003C\u002Fh3>\n\u003Csub>citizen developers worldwide\u003Cbr>(38% YoY growth)\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\">\n\u003Ch3>$4.7B\u003C\u002Fh3>\n\u003Csub>vibe coding market\u003Cbr>growing 38% annually\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\">\n\u003Ch3>25%\u003C\u002Fh3>\n\u003Csub>of YC W25 startups had 95%+ AI-generated codebases\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\nMost tools help you start a project. This harness helps you **finish it** — with plans your team can review, context that never goes stale, and safety systems that catch mistakes before they ship.\n\n---\n\n## Who Is This For?\n\n\u003Cdiv align=\"center\">\n\u003Ch3>\u003Cem>\"The point isn't who typed it. It's what shipped.\"\u003C\u002Fem>\u003C\u002Fh3>\n\u003Cstrong>— Garry Tan, YC\u003C\u002Fstrong>\n\u003C\u002Fdiv>\n\n\u003Cbr>\n\nWhether you just discovered vibe coding or you're a staff engineer shipping production systems — this harness adapts to your workflow.\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🧑‍💼\u003C\u002Fh1>\n\u003Cstrong>CEO \u002F Founder\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Cem>\"Build me a SaaS with auth, billing, and a landing page\"\u003C\u002Fem>\u003Cbr>\u003Cbr>\nThe agent researches your stack, writes an architecture plan you can review, implements with tests, and captures every decision for your technical co-founder to audit later.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>📊\u003C\u002Fh1>\n\u003Cstrong>Product Manager\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Cem>\"Create a dashboard showing MRR, churn, and growth metrics\"\u003C\u002Fem>\u003Cbr>\u003Cbr>\nIt generates a PRD-style spec, gets your approval before writing code, implements to spec, and archives the plan as searchable project history.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🎨\u003C\u002Fh1>\n\u003Cstrong>Designer\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Cem>\"Match this Figma screenshot pixel-perfect\"\u003C\u002Fem>\u003Cbr>\u003Cbr>\nThe design-aware agent analyzes your mockup, implements component-by-component with your design tokens, and spawns visual comparison checks.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>⚙️\u003C\u002Fh1>\n\u003Cstrong>Engineer\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Cem>\"Refactor the auth module to support RBAC with zero downtime\"\u003C\u002Fem>\u003Cbr>\u003Cbr>\nIt researches your current auth code and how other codebases solved RBAC, writes a migration plan with blast radius analysis, implements safely with rollback notes.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## At a Glance\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🤖\u003C\u002Fh1>\n\u003Ch3>12\u003C\u002Fh3>\n\u003Cstrong>Specialized Agents\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Domain experts that own each development phase\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>⚡\u003C\u002Fh1>\n\u003Ch3>32\u003C\u002Fh3>\n\u003Cstrong>Auto-Discovered Skills\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Reusable capabilities surfaced by keyword matching\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🪝\u003C\u002Fh1>\n\u003Ch3>7\u003C\u002Fh3>\n\u003Cstrong>Lifecycle Hooks\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Pre\u002Fpost execution guardrails and context injection\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>📜\u003C\u002Fh1>\n\u003Ch3>6\u003C\u002Fh3>\n\u003Cstrong>Development Protocols\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Shared workflow rules across all tools\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🛡️\u003C\u002Fh1>\n\u003Ch3>5\u003C\u002Fh3>\n\u003Cstrong>Safety Systems\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Phase-locking, blast radius, privacy, leak detection\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🔧\u003C\u002Fh1>\n\u003Ch3>7\u003C\u002Fh3>\n\u003Cstrong>Supported Tools\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>Claude Code, Codex, Cursor, Windsurf, Antigravity, OpenCode, Copilot\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>🌍\u003C\u002Fh1>\n\u003Ch3>6\u003C\u002Fh3>\n\u003Cstrong>Languages\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>EN · 中文 · 日本語 · 한국어 · Tiếng Việt · Português\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003Ctd align=\"center\" width=\"50%\" valign=\"top\">\n\u003Ch1>⚡\u003C\u002Fh1>\n\u003Ch3>30s\u003C\u002Fh3>\n\u003Cstrong>Install Time\u003C\u002Fstrong>\u003Cbr>\n\u003Csub>One curl command + auto-setup does the rest\u003C\u002Fsub>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 💎 Why Teams Use This\n\n> Most harnesses give you a CLAUDE.md and instructions. This gives you an **autonomous development system** that compounds intelligence over time.\n\n\u003Cbr>\n\n### 📋 Spec-Driven Development — Not Vibes-Driven\n\nEvery feature gets a **written plan with blast radius analysis** before a single line of code is written.\n\n> 💡 Auto-generates PRDs, manages backlogs, organizes feature groups. Works for both developers and product managers — your agent plans like a senior engineer, not an intern.\n\n**What's in every plan:**\n\n| Section | Purpose |\n|---|---|\n| 📍 **Touchpoints** | Every file that will be created or modified, listed upfront |\n| 📜 **Public contracts** | What API surfaces or interfaces change |\n| 💥 **Blast radius** | What could break, what tests to run, what to watch |\n| ✅ **Verification evidence** | How to prove the implementation is correct |\n| 🔄 **Resume handoff** | Enough context for any agent to pick up mid-plan |\n\n\u003Cbr>\n\n### 🔄 Autonomous Multi-Phase Execution — Hours of Hands-Free Work\n\nFor large tasks, the agent runs an **iterative phased loop**:\n\n```\n🔍 research → ⚡ execute → ✅ validate → 📄 report → 🔄 repeat\n```\n\n> 💡 It self-heals when stuck, self-reflects to improve approach, and writes durable progress reports to disk. **Context compaction can't kill it** — all state lives in files, not memory.\n\nWalk away and come back to completed work.\n\n\u003Cbr>\n\n### 🧬 Auto-Architecture Research — Learn From Any Codebase\n\nThe agent doesn't just read your code — it **studies other repositories** to learn how they solved similar problems (`vc-xia`).\n\n> 💡 It researches, compares approaches, and adapts the best patterns into your codebase. Architecture decisions are informed by real-world implementations, not hallucinated best practices.\n\n\u003Cbr>\n\n### 🧭 Persistent Smart Context Routing — Always the Right Context\n\nContext isn't one giant file. It's organized into **auto-routed knowledge domains**:\n\n```\nprocess\u002Fcontext\u002F\n├── all-context.md              # 🧭 Root router — reads your task, loads what's relevant\n├── tests\u002F\n│   └── all-tests.md            # 🧪 Test runners, commands, debugging\n├── container\u002F\n│   └── all-container.md        # 🐳 Docker, deployment, infra\n├── uxui\u002F\n│   └── all-uxui.md             # 🎨 Components, design tokens, patterns\n└── {your-domain}\u002F\n    └── all-{domain}.md         # 📚 Any domain with 3+ durable docs\n```\n\n> 💡 When the agent works on billing, it loads billing context — not your entire codebase docs. Context **auto-updates every time you complete a feature**, so it never goes stale.\n\n\u003Cbr>\n\n### 🧠 Self-Improving Knowledge Base — Gets Smarter as You Ship\n\nEvery completed feature feeds learnings back into the context system.\n\n> 💡 Research findings, architectural decisions, debugging insights, and coding patterns are **captured and indexed automatically**. Your 100th feature benefits from everything learned in the first 99. The knowledge compounds — it doesn't reset.\n\n---\n\n## How This Compares\n\n| Feature | vibecode-pro-max-kit | Superpowers | GSD | gstack |\n|---------|---------------------|-------------|-----|--------|\n| Spec-driven lifecycle | Full RIPER-5 (research → plan → execute → verify) | Mandatory workflows | Context-rot fix | Partial |\n| Phase-locked safety | Tool restrictions per mode (read-only research, no-write innovate) | Skill-based constraints | Phase separation | None |\n| Multi-tool support | 7 tools via AGENTS.md + native | Claude Code plugin | 14 runtimes | 1 tool |\n| Auto-improving context | Domain-routed context groups, updates after every feature | Plugin memory | Disk-persisted state | Manual |\n| Team collaboration | Shared specs, plans, and review artifacts | Solo | Solo | Solo |\n| Skills system | 32 auto-discovered, keyword-matched at every prompt | 86 composable skills | Meta-prompting | 23 role tools |\n| Multi-phase programs | Umbrella plans + phase-by-phase execution loop with regression checks | Single task | Single task | Single task |\n| Quality pipeline | 6-step chain (code-review → test → simplify → security → audit → commit) | Per-skill quality | No auto-chain | No auto-chain |\n| Installation | 30-second `curl` install + auto-setup | Plugin marketplace | npx one-liner | git clone |\n| Context routing | Domain-based routing table with grouped context packs | Flat skill context | Flat context | Single file |\n\n> **On runtime breadth:** GSD supports 14 runtimes. We support 7 deeply — with full agent harnesses, skill discovery, and lifecycle hooks on every platform. Breadth vs. depth: your choice.\n\n---\n\n## ⚡ What Makes This Different\n\nMost agent harnesses give you a big CLAUDE.md and some instructions. Here's what this one actually does:\n\n\u003Cbr>\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔒\u003C\u002Fh1>\n\u003Cstrong>Phase-Locked Tool Restrictions\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nYour agent literally \u003Cstrong>cannot\u003C\u002Fstrong> write code during research. RESEARCH is read-only, INNOVATE has no Bash, PLAN can only write to \u003Ccode>process\u002F\u003C\u002Fcode>. \u003Cstrong>Actual capability removal\u003C\u002Fstrong>, not suggestions.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🎯\u003C\u002Fh1>\n\u003Cstrong>Smart Auto-Routing\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nDetects your intent from natural language. \"build webhook support\" → full pipeline. \"login is broken\" → debugger. 6-level precedence, one clarifying question max.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔍\u003C\u002Fh1>\n\u003Cstrong>Automatic Skill Discovery\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nBefore routing any request, scans \u003Cstrong>32 skills\u003C\u002Fstrong> and matches keywords. Say \"add webhook support\" and \u003Ccode>vc-security\u003C\u002Fcode> + \u003Ccode>vc-scenario\u003C\u002Fcode> surface automatically.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>💾\u003C\u002Fh1>\n\u003Cstrong>Survives Context Compaction\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nPlans, reports, context docs, and learnings all live on disk. The session-init hook re-injects approval gates after compaction. \u003Cstrong>Nothing is lost.\u003C\u002Fstrong>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🛡️\u003C\u002Fh1>\n\u003Cstrong>Self-Policing Violation Detection\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nWhen the agent is about to cross a phase boundary, it stops itself: \u003Cem>\"PHASE JUMPING PREVENTED\"\u003C\u002Fem>. A \u003Cstrong>structural hallucination guard\u003C\u002Fstrong>.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔄\u003C\u002Fh1>\n\u003Cstrong>Works Across 7 AI Coding Tools\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nTwo open standards — \u003Ccode>AGENTS.md\u003C\u002Fcode> and \u003Ccode>SKILL.md\u003C\u002Fcode> — mean \u003Cstrong>zero adapters, zero plugins, zero configuration.\u003C\u002Fstrong> Start in Claude Code, switch to Cursor, continue in Codex.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 🧭 How It Works\n\n```\nYour request\n  → Step 0: Skill Discovery (match keywords → surface relevant skills)\n  → Intent Detection (feature \u002F bug \u002F question \u002F refactor \u002F UI)\n  → Route to the right agent\n  → Phase-locked execution with explicit transitions\n```\n\nThe orchestrator **never does the work itself** — it routes, monitors, and manages transitions.\n\n\u003Cbr>\n\n### 📊 The Workflow\n\n| Phase | What happens | You say |\n|-------|-------------|---------|\n| 🔍 **RESEARCH** | Read-only fact gathering — codebase + web | *(automatic on feature requests)* |\n| 💡 **INNOVATE** | Explore 2-3 approaches with trade-offs | `go` |\n| 📋 **PLAN** | Write a detailed spec you can review | `go` |\n| ⚡ **EXECUTE** | Implement exactly what was planned | `ENTER EXECUTE MODE` |\n| 🧠 **UPDATE PROCESS** | Capture learnings, update context, archive plan | *(recommended after non-trivial work)* |\n\n> 💡 **Shortcuts:** `ENTER FAST MODE - [task]` compresses RESEARCH+INNOVATE+PLAN into one pass — still pauses before EXECUTE. Trivial fixes (single file, \u003C15 lines, no schema\u002Fauth changes) skip straight to execute.\n\n\u003Cbr>\n\n### 💻 Typical Session\n\n```\n# 🆕 Feature request\nYou: \"add webhook support to the API\"\n→ Skill discovery surfaces: vc-scenario, vc-security\n→ research-agent gathers context (read-only, can't touch code)\n→ You say \"go\" → innovate-agent explores approaches\n→ You say \"go\" → plan-agent writes spec with blast radius\n→ You review the plan, say \"ENTER EXECUTE MODE\"\n→ execute-agent implements → self-review → tester → code-reviewer → git-manager\n→ Closeout packet: what changed, what's verified, recommended next step\n```\n\n```\n# 🐛 Bug fix\nYou: \"login redirect is broken\"\n→ Routes to vc-debugger → evidence gathering → competing hypotheses\n→ Root cause identified with proof chain\n→ execute-agent implements the fix → quality pipeline\n```\n\n```\n# ⏩ Fast mode\nYou: \"ENTER FAST MODE - add rate limiting middleware\"\n→ Compressed research+innovate+plan in one pass\n→ Mandatory safety pause → you review → \"ENTER EXECUTE MODE\"\n```\n\n```\n# 🏗️ Large program\nYou: \"build a full testing platform\"\n→ Creates umbrella plan + phase plans in a feature folder\n→ Each phase: re-research → approve → execute → validate → durable report\n→ Progress survives context compaction — durable reports on disk\n```\n\n```\n# 🔄 Autonomous optimization\nYou: \"improve test coverage to 80% using vc-autoresearch\"\n→ Agent iterates: make change → commit → measure → keep\u002Frevert\n→ Stuck detection after 5 consecutive discards → strategy shift\n→ Full audit trail in TSV\n```\n\n---\n\n## 🛡️ Built-in Safety Systems\n\nThese aren't just guidelines — they're **structural enforcement** built into every agent.\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>⏸️\u003C\u002Fh1>\n\u003Cstrong>50% Mid-Implementation Check-In\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nAt approximately halfway through execution, the agent \u003Cstrong>pauses\u003C\u002Fstrong> to report progress, list completed and remaining items, and asks: \u003Cem>\"Continue with current approach or pause and return to PLAN?\"\u003C\u002Fem>\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🚫\u003C\u002Fh1>\n\u003Cstrong>Never Silently Deviate\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nIf the execute-agent hits a problem requiring deviation from the plan, it \u003Cstrong>immediately stops\u003C\u002Fstrong>, explains the issue, and returns to PLAN mode. No quiet improvising.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔙\u003C\u002Fh1>\n\u003Cstrong>Approach Abandonment Protocol\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nWhen an approach fails, the agent evaluates reusable components, documents lessons before deletion, creates an abandonment summary, and returns to PLAN.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔐\u003C\u002Fh1>\n\u003Cstrong>Privacy Guardrails Hook\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nThe agent is \u003Cstrong>blocked from reading\u003C\u002Fstrong> \u003Ccode>.env\u003C\u002Fcode>, credentials, SSH keys, and \u003Ccode>.pem\u003C\u002Fcode> files. Must ask for explicit approval.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>⚠️\u003C\u002Fh1>\n\u003Cstrong>High-Risk Evidence Packs\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nFor changes touching auth, billing, schema migrations, or public APIs — the system requires a formal evidence pack before calling work \"done.\"\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>📊\u003C\u002Fh1>\n\u003Cstrong>Drift Signal Scoring\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\nAfter execution, the system scores urgency: \u003Cstrong>LOW\u003C\u002Fstrong> (light touch), \u003Cstrong>MEDIUM\u003C\u002Fstrong> (significant changes), \u003Cstrong>HIGH\u003C\u002Fstrong> (harness\u002Fprotocol files touched).\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 🔍 Pre-Implementation Intelligence\n\nBefore a single line of code is written, the system can catch issues through specialized analysis:\n\n\u003Cbr>\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🎭\u003C\u002Fh1>\n\u003Cstrong>5-Persona Pre-Implementation Debate\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-predict\u003C\u002Fcode> — Architect, Security, Performance, UX, and Devil's Advocate debate your plan. Produces a \u003Cstrong>GO \u002F CAUTION \u002F STOP\u003C\u002Fstrong> verdict before you write a line of code.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🎲\u003C\u002Fh1>\n\u003Cstrong>12-Dimension Edge Case Generator\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-scenario\u003C\u002Fcode> — Decomposes any feature across 12 dimensions (user types, input extremes, timing, scale, state, env, errors, auth, data, integrations, compliance, business logic). Outputs usable as test specs.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔐\u003C\u002Fh1>\n\u003Cstrong>STRIDE + OWASP Security Audit\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-security\u003C\u002Fcode> — Dual-methodology security audit with dependency auditing, secret detection, and \u003Cstrong>auto-fix mode\u003C\u002Fstrong> that sorts by severity and fixes Critical first with regression guards.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## 🤖 Autonomous Agent Capabilities\n\n\u003Cbr>\n\n\u003Ctable>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔄\u003C\u002Fh1>\n\u003Cstrong>Autonomous Metric Optimization\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-autoresearch\u003C\u002Fcode> — Set a goal, walk away. Iterative git-backed loop: make ONE atomic change → commit → measure → keep or revert. Stuck detection after 5 consecutive discards triggers strategy shifts.\n\u003C\u002Ftd>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>👥\u003C\u002Fh1>\n\u003Cstrong>Parallel Agent Teams\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-team\u003C\u002Fcode> — Multiple agents working \u003Cstrong>simultaneously\u003C\u002Fstrong> with git worktree isolation. Research in parallel, execute in parallel, review in parallel, debug adversarially.\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003Ctr>\n\u003Ctd width=\"50%\" valign=\"top\">\n\u003Ch1>🔬\u003C\u002Fh1>\n\u003Cstrong>Evidence-Before-Hypothesis Debugging\u003C\u002Fstrong>\u003Cbr>\u003Cbr>\n\u003Ccode>vc-debugger\u003C\u002Fcode> — Gathers evidence first → forms 2-3 competing hypotheses → systematically tests each → documents elimination path. \u003Cstrong>Never guesses — proves.\u003C\u002Fstrong>\n\u003C\u002Ftd>\n\u003C\u002Ftr>\n\u003C\u002Ftable>\n\n---\n\n## ✅ Quality Pipeline — Built Into Execution\n\nThe execute-agent doesn't just write code and call it done. It auto-chains through a **quality pipeline**:\n\n\u003Cbr>\n\n```mermaid\n%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '16px', 'lineColor': '#8888AA'}} }%%\nflowchart TD\n    E[\"⚡ Execute-Agent\\nImplements the plan\"]\n    SR[\"🔎 Self-Review\\nLine-by-line check\\nagainst plan\"]\n    T[\"🧪 Tester\\nDiff-aware — only\\nruns affected tests\"]\n    CR[\"🔍 Code Reviewer\\nEdge case scout\\n+ adversarial review\"]\n    CS[\"✨ Code Simplifier\\nClarity refactoring\"]\n    GM[\"📦 Git Manager\\nLogical commit splitting\\nfrom touched_files\"]\n\n    E --> SR\n    SR --> T\n    T --> CR\n    CR --> CS\n    CS --> GM\n\n    style E fill:#C62828,stroke:#B71C1C,color:#FFFFFF\n    style SR fill:#AD1457,stroke:#880E4F,color:#FFFFFF\n    style T fill:#6A1B9A,stroke:#4A148C,color:#FFFFFF\n    style CR fill:#283593,stroke:#1A237E,color:#FFFFFF\n    style CS fill:#00695C,stroke:#004D40,color:#FFFFFF\n    style GM fill:#37474F,stroke:#263238,color:#FFFFFF\n```\n\n\u003Cbr>\n\n| Step | What it does |\n|---|---|\n| 🔎 **Self-review** | Checks every checklist item against plan for deviations, documents them |\n| 🧪 **Tester** | Maps changed files to test files, auto-escalates to full suite when >70% mapped |\n| 🔍 **Code reviewer** | Dispatches edge case scout BEFORE review, checks N+1 queries, auth paths, data leaks |\n| ✨ **Simplifier** | Clarity refactoring after review passes — no behavior changes |\n| 📦 **Git manager** | Receives `touched_files` list, splits into logical conventional commits, refuses unknown files |\n\n---\n\n## 📋 The Plan Lifecycle — Spec-Driven, Not Vibes-Driven\n\nEvery non-trivial feature follows a **plan lifecycle** — a written spec that is created, reviewed, executed against, and archived as project history.\n\n\u003Cbr>\n\n```mermaid\n%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '16px', 'lineColor': '#8888AA'}} }%%\nflowchart TD\n    A[\"🆕 Feature Request\"]\n    B[\"📝 Plan Created\\nin active\u002F\"]\n    C{\"👀 User Reviews\\nthe Plan\"}\n    D[\"⚡ Execute Against Plan\"]\n    E[\"📦 Plan Archived\\nto completed\u002F\"]\n    F[\"🧠 Learnings Written\\nto all-context.md\"]\n    G[\"🔄 Next Feature\\nStarts Smarter\"]\n\n    A --> B\n    B --> C\n    C -->|\"✅ Approved\"| D\n    C -->|\"✏️ Needs Changes\"| B\n    D --> E\n    E --> F\n    F --> G\n    G -.->|\"context compounds\"| A\n\n    style A fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n    style B fill:#E65100,stroke:#BF360C,color:#FFFFFF\n    style C fill:#F57F17,stroke:#F9A825,color:#000000\n    style D fill:#C62828,stroke:#B71C1C,color:#FFFFFF\n    style E fill:#6A1B9A,stroke:#4A148C,color:#FFFFFF\n    style F fill:#00695C,stroke:#004D40,color:#FFFFFF\n    style G fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n```\n\n\u003Cbr>\n\n> 💡 Six months from now, when someone asks *\"why did we build auth this way?\"*, the answer is in `completed\u002F`. Not lost in a Slack thread.\n\n\u003Cbr>\n\n**Where plans live on disk:**\n\n```\nprocess\u002F\n├── general-plans\u002F\n│   ├── active\u002F                  # 📋 Plans currently being worked on\n│   │   └── webhooks_PLAN_28-05-26.md\n│   ├── completed\u002F               # ✅ Archived plans (searchable history)\n│   ├── backlog\u002F                 # 📌 Deferred work\n│   ├── reports\u002F                 # 📄 Cross-cutting reports\n│   └── references\u002F              # 📚 Research outputs\n└── features\u002F\n    └── billing\u002F                 # 🏷️ Feature-scoped (5+ artifacts)\n        ├── active\u002F\n        ├── completed\u002F\n        ├── backlog\u002F\n        ├── reports\u002F\n        └── references\u002F\n```\n\n---\n\n## 🏗️ Phase Programs — Large Projects That Don't Fall Apart\n\nNormal features use one plan. **Large multi-phase projects** use a phase program — an umbrella plan plus individual phase plans, each with its own validation gate.\n\n\u003Cbr>\n\n```mermaid\n%%{init: {'theme': 'base', 'themeVariables': {'fontSize': '16px', 'lineColor': '#8888AA'}} }%%\nflowchart TD\n    UP[\"🎯 Umbrella Plan\\nOverall program goal\"]\n    P1[\"📋 Phase 1 Plan\"]\n    P2[\"📋 Phase 2 Plan\"]\n    P3[\"📋 Phase 3 Plan\"]\n\n    R1[\"🔍 Re-Research\"]\n    E1[\"⚡ Execute\"]\n    V1[\"✅ Validate\"]\n    RP1[\"📄 Durable Report\"]\n\n    R2[\"🔍 Re-Research\"]\n    E2[\"⚡ Execute\"]\n    V2[\"✅ Validate\"]\n    RP2[\"📄 Durable Report\"]\n\n    UP --> P1\n    UP --> P2\n    UP --> P3\n\n    P1 --> R1\n    R1 -->|\"approval\"| E1\n    E1 --> V1\n    V1 --> RP1\n    RP1 -.->|\"learnings feed\\nnext phase\"| R2\n\n    P2 --> R2\n    R2 -->|\"approval\"| E2\n    E2 --> V2\n    V2 --> RP2\n\n    style UP fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n    style P1 fill:#E65100,stroke:#BF360C,color:#FFFFFF\n    style P2 fill:#E65100,stroke:#BF360C,color:#FFFFFF\n    style P3 fill:#E65100,stroke:#BF360C,color:#FFFFFF\n    style R1 fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n    style E1 fill:#C62828,stroke:#B71C1C,color:#FFFFFF\n    style V1 fill:#2E7D32,stroke:#1B5E20,color:#FFFFFF\n    style RP1 fill:#6A1B9A,stroke:#4A148C,color:#FFFFFF\n    style R2 fill:#1565C0,stroke:#0D47A1,color:#FFFFFF\n    style E2 fill:#C62828,stroke:#B71C1C,color:#FFFFFF\n    style V2 fill:#2E7D32,stroke:#1B5E20,color:#FFFFFF\n    style RP2 fill:#6A1B9A,stroke:#4A148C,color:#FFFFFF\n```\n\n\u003Cbr>\n\n**Key features:**\n\n| | Feature | Why it matters |\n|---|---|---|\n| 🔄 | **Re-research at every phase** | Checks for code drift, reads latest reports, updates assumptions |\n| ✅ | **Validation gates** | Phase isn't `VERIFIED` until evidence proves it. Honest status: `PLANNED` → `CODE DONE` → `TESTING` → `VERIFIED` or `BLOCKED` |\n| 📄 | **Durable reports** | Every phase writes results to disk. Progress survives context compaction |\n| 🧠 | **Learnings feed forward** | Phase 1 discoveries update Phase 2's plan before execution |\n| 🏗️ | **Foundation vs expansion** | Explicitly splits \"prove the architecture\" from \"implement everything\" |\n| 🚧 | **Honest blocker handling** | Blocked phases stay `BLOCKED` with evidence. No forcing green status |\n\n---\n\n## 🧠 Context Groups — Organized Knowledge, Not One Giant File\n\nProject knowledge is organized into **context groups** — durable knowledge domains, each with an `all-{group}.md` router that tells agents what to read and when.\n\n\u003Cbr>\n\n```\nprocess\u002Fcontext\u002F\n├── all-context.md              # 🧭 Root router — architecture, stack, patterns, conventions\n├── tests\u002F\n│   └── all-tests.md            # 🧪 Test runners, commands, debugging procedures\n├── container\u002F\n│   └── all-container.md        # 🐳 Docker, deployment, infra procedures\n├── uxui\u002F\n│   └── all-uxui.md             # 🎨 Components, design tokens, patterns\n├── infra\u002F\n│   └── all-infra.md            # 🖥️ Worker nodes, provisioning, DNS\n├── skills\u002F\n│   └── all-skills.md           # ⚡ Skill runtime, app architecture\n├── workflows\u002F\n│   └── all-workflows.md        # 🔄 Workflow runtime, deployment\n└── {your-domain}\u002F\n    └── all-{domain}.md         # 📚 Any knowledge domain with 3+ durable docs\n```\n\n\u003Cbr>\n\n| | How it works |\n|---|---|\n| 🧭 **Router pattern** | Agents read only what's relevant to their task, not everything |\n| 📏 **Auto-promotion** | Topics with 3+ docs or 800+ lines get their own context group |\n| 🔄 **Living docs** | Updated by `update-process-agent` after every non-trivial feature |\n| 🧪 **Auditable** | `vc-audit-context` verifies routing and consistency |\n\n---\n\n## 📁 Feature Folders — Self-Organizing Project Memory\n\nWhen a topic accumulates 5+ artifacts, it gets its own **feature folder** — a complete lifecycle container.\n\n\u003Cbr>\n\n```\nprocess\u002Ffeatures\u002F{feature}\u002F\n├── active\u002F       # 📋 Plans currently being worked on\n├── completed\u002F    # ✅ Archived plans (searchable decision history)\n├── backlog\u002F      # 📌 Deferred work (agents check before duplicating)\n├── reports\u002F      # 📄 Execution reports, post-mortems, validation results\n└── references\u002F   # 📚 Research outputs that inform future decisions\n```\n\n\u003Cbr>\n\n| | What happens |\n|---|---|\n| 🆕 | New work starts in `active\u002F` → reports accumulate → plan archives to `completed\u002F` |\n| 📌 | Deferred work goes to `backlog\u002F` — agents check it before creating duplicate plans |\n| 📦 | Feature promotion happens automatically when general artifacts hit 5+ |\n| 🔍 | Every feature has complete, self-contained history — plans, decisions, reports, research |\n\n---\n\n## 🤖 What's Inside\n\n\u003Cbr>\n\n### 12 Agents\n\n\u003Cdetails>\n\u003Csummary>Click to expand agent list (12 agents)\u003C\u002Fsummary>\n\n\u003Cbr>\n\n**Core workflow agents** — one per RIPER-5 phase:\n\n| Agent | Role |\n|-------|------|\n| 🔍 `vc-research-agent` | Codebase + web research, read-only. Contradiction tracking built in |\n| 💡 `vc-innovate-agent` | Brainstorm 2-3 approaches. Must produce decision summary before PLAN |\n| 📋 `vc-plan-agent` | Write spec with anti-rationalization guards. \"I already know how\" is not a plan |\n| ⚡ `vc-execute-agent` | Implement per plan. 50% check-in, deviation protocol, self-review |\n| ⏩ `vc-fast-mode-agent` | Compressed RESEARCH→INNOVATE→PLAN with mandatory safety pause |\n| 🧠 `vc-update-process-agent` | 7-phase mandatory checklist including stale artifact scanning |\n\n\u003Cbr>\n\n**Specialist agents** — called during EXECUTE or standalone:\n\n| Agent | Role |\n|-------|------|\n| 🐛 `vc-debugger` | Evidence-before-hypothesis. Competing hypotheses, elimination chains |\n| 🧪 `vc-tester` | Diff-aware. Only runs affected tests. Auto-escalates on config changes |\n| 🔎 `vc-code-reviewer` | Edge case scout BEFORE review. N+1 detection, auth path validation |\n| ✨ `vc-code-simplifier` | Clarity refactoring without behavior change |\n| 🎨 `vc-ui-ux-designer` | Design-aware frontend. Can spawn research subagent mid-execution |\n| 📦 `vc-git-manager` | Logical commit splitting from `touched_files`. Refuses unknown files |\n\n\u003C\u002Fdetails>\n\n\u003Cbr>\n\n### 31 Skills (auto-discovered)\n\n\u003Cdetails>\n\u003Csummary>Click to expand skill list (31 skills)\u003C\u002Fsummary>\n\n\u003Cbr>\n\n**🔧 Contract skills** — `vc-generate-plan` · `vc-generate-context` · `vc-audit-context` · `vc-audit-plans` · `vc-audit-vc` · `vc-setup` · `vc-update` · `vc-publish`\n\n**🧠 Planning** — `vc-predict` (5-persona debate) · `vc-scenario` (12-dimension edge cases) · `vc-sequential-thinking` · `vc-problem-solving`\n\n**🐛 Debug & security** — `vc-debug` · `vc-security` (STRIDE + OWASP + auto-fix) · `vc-autoresearch` (autonomous optimization)\n\n**📚 Research** — `vc-docs-seeker` · `vc-scout` · `vc-docs` · `vc-repomix` · `vc-xia` (repo comparison)\n\n**🎨 Frontend** — `vc-frontend-design` · `vc-chrome-devtools` · `vc-agent-browser` · `vc-web-testing`\n\n**⚙️ Utilities** — `vc-context-engineering` · `vc-mcp-management` · `vc-preview` · `vc-team` (parallel agents) · `vc-tech-graph` · `vc-watzup` (session handoff) · `vc-merge-worktree`\n\n\u003C\u002Fdetails>\n\n> 💡 Some skills (like `vc-xia`) were inspired by [ClaudeKit](https:\u002F\u002Fclaudekit.cc\u002F?ref=OEOM7R7G) by [@mrgoonie](https:\u002F\u002Fgithub.com\u002Fmrgoonie). We focused on fewer, deeper skills instead of 80+.\n\n\u003Cbr>\n\n### 🪝 7 Hooks\n\n| Hook | What it does |\n|------|-------------|\n| 🔐 **Privacy guardrails** | Blocks `.env`, credentials, SSH keys. Requires explicit approval |\n| 🚫 **Scout blocker** | Prevents agent from wandering into `node_modules\u002F`, `dist\u002F`. Gitignore-syntax `.ckignore` |\n| 🧠 **Session init** | Detects stack, injects env vars, recovers approval gates after compaction |\n| 💉 **Subagent context** | Injects ~200 token compact context block into every subagent |\n| ✨ **Edit quality** | After 5+ edits, nudges to run code-simplifier (non-blocking, throttled) |\n| 📛 **Descriptive naming** | Language-aware file naming conventions on every Write |\n| 📊 **Usage tracking** | Session metrics and token awareness |\n\n\u003Cbr>\n\n**Where everything lives:**\n\n```\nyour-project\u002F\n├── .claude\u002F\n│   ├── agents\u002F              # 🤖 12 agent definitions (.md)\n│   ├── skills\u002F              # ⚡ 31 skill modules (each a directory with SKILL.md)\n│   └── hooks\u002F               # 🪝 7 lifecycle hooks (.cjs)\n├── .codex\u002F\n│   └── agents\u002F              # 🔄 Mirrored for Codex compatibility\n├── .agents\u002F\n│   └── skills -> ..\u002F.claude\u002Fskills   # 🔗 Symlink for Codex discovery\n├── CLAUDE.md                # 📋 Orchestrator config + routing rules\n├── AGENTS.md                # 📖 Agent + skill registry\n└── process\u002F\n    ├── context\u002F             # 🧠 Auto-routed knowledge domains\n    ├── general-plans\u002F       # 📋 Cross-cutting plans + reports\n    ├── features\u002F            # 🏷️ Feature-scoped lifecycle folders\n    └── development-protocols\u002F  # 📜 Shared workflow rules\n```\n\n---\n\n## 🔄 Updating\n\nPull the latest harness improvements:\n\n```\nRun vc-update\n```\n\n> 💡 Shows a dry-run diff, waits for confirmation. Your `process\u002F` directory and project-specific content are **never touched**.\n\n---\n\n## Contributing\n\nWe welcome contributions! See [CONTRIBUTING.md](CONTRIBUTING.md) for guidelines.\n\n\u003Cbr>\n\n**Quick links:**\n\n- 🐛 [Report a bug](https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fissues\u002Fnew?template=1.bug_report.yml)\n- 💡 [Request a feature](https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fissues\u002Fnew?template=2.feature_request.yml)\n- ⚡ [Submit a skill](https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fissues\u002Fnew?template=3.skill_submission.yml)\n- 🌐 [Add a translation](https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fissues\u002Fnew?template=5.translation.yml)\n\n\u003Cbr>\n\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fwithkynam\u002Fvibecode-pro-max-kit\u002Fgraphs\u002Fcontributors\">\n  \u003Cimg src=\"https:\u002F\u002Fcontrib.rocks\u002Fimage?repo=withkynam\u002Fvibecode-pro-max-kit\" alt=\"Contributors\" \u002F>\n\u003C\u002Fa>\n\n\u003Cbr>\n\n### 🙏 Credits\n\nThis project benefited from [ClaudeKit](https:\u002F\u002Fclaudekit.cc\u002F?ref=OEOM7R7G) by [@mrgoonie](https:\u002F\u002Fgithub.com\u002Fmrgoonie) — particularly skills like `ck:xia` which inspired some of ours.\n\nThe difference: vibecode-pro-max-kit focuses on the spec-driven development framework and self-improving context organization, without bloating you with 80+ skills. Fewer tools, more structure.\n\n---\n\n## ⭐ Star History\n\n\u003Ca href=\"https:\u002F\u002Fstar-history.com\u002F#withkynam\u002Fvibecode-pro-max-kit&Date\">\n \u003Cpicture>\n   \u003Csource media=\"(prefers-color-scheme: dark)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=withkynam\u002Fvibecode-pro-max-kit&type=Date&theme=dark\" \u002F>\n   \u003Csource media=\"(prefers-color-scheme: light)\" srcset=\"https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=withkynam\u002Fvibecode-pro-max-kit&type=Date\" \u002F>\n   \u003Cimg alt=\"Star History Chart\" src=\"https:\u002F\u002Fapi.star-history.com\u002Fsvg?repos=withkynam\u002Fvibecode-pro-max-kit&type=Date\" \u002F>\n \u003C\u002Fpicture>\n\u003C\u002Fa>\n\n---\n\n## 📄 License\n\nMIT\n","vibecode-pro-max-kit 是一个专为开发者、产品经理和CEO设计的AI编码助手，能够自动生成产品需求文档、管理待办事项并自动路由上下文。它拥有12个智能代理和32种技能，通过自我优化的知识库来防止上下文丢失，即使在六个月后也能保持良好的记忆状态。该项目基于JavaScript开发，支持Claude Code & Codex等AI技术，适用于任何技术栈。特别适合需要长时间执行大型任务且不希望中断状态的场景，如持续集成\u002F持续部署（CI\u002FCD）流程中，确保代码质量和团队协作效率。",2,"2026-06-11 03:59:49","CREATED_QUERY"]