zugzbot 1.0.21 → 1.0.22

package/.opencode/oh-my-design/README.ko.md

@@ -1,112 +0,0 @@
-<p align="center">
-  <img src=".github/assets/logo-bg.png" width="480" alt="oh-my-design" />
-</p>
-
-<h1 align="center">oh-my-design</h1>
-
-<p align="center">
-  <strong>AI 코딩 에이전트를 위한 스킬 기반 디자인 부트스트랩 — 명령어 한 번.</strong> 246개 실제 기업 디자인 시스템. 설치에 AI 호출 없음. 그 다음은 에이전트에게 말만 하면 됩니다.
-</p>
-
-<p align="center">
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/v/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm version" /></a>
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/dm/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm downloads" /></a>
-  <a href="LICENSE"><img src="https://img.shields.io/github/license/kwakseongjae/oh-my-design?style=flat-square" alt="License" /></a>
-  <a href="https://github.com/kwakseongjae/oh-my-design/stargazers"><img src="https://img.shields.io/github/stars/kwakseongjae/oh-my-design?style=social" alt="GitHub Stars" /></a>
-  <img src="https://img.shields.io/badge/references-246-7c5cfc?style=flat-square" alt="246 References" />
-  <img src="https://img.shields.io/badge/CLI%20commands-1-blue?style=flat-square" alt="One CLI command" />
-</p>
-
-<p align="center">
-  한국어 | <a href="README.md">English</a> | <a href="README.ja.md">日本語</a> | <a href="README.zh-TW.md">繁體中文</a>
-</p>
-
----
-
-## oh-my-design이란?
-
-**oh-my-design (OmD)** 은 AI 코딩 에이전트를 위한 디자인 시스템입니다. Claude Code / Codex / OpenCode / Cursor를 당신의 브랜드를 기억하는 시니어 프로덕트 디자이너로 만듭니다. 한 번 설치하면, 이후엔 원하는 것을 설명하기만 하면 됩니다 — 컴포넌트, 화면, 카피, 에셋, 차트 — 에이전트가 프로젝트의 디자인 시스템을 적용해 결과물을 만들어냅니다. `DESIGN.md`가 브랜드 스펙이고([Google Stitch](https://stitch.withgoogle.com/docs/design-md/overview/) 토큰 + 브랜드 철학 레이어: Voice / Narrative / Principles / Personas / States / Motion), 246개 실제 기업의 DESIGN.md 파일이 패키지에 함께 들어 있습니다. **API 키 불필요. 외부 인프라 불필요. 모든 것이 기존 CLI 세션 안에서 동작합니다.**
-
-## 설치
-
-```bash
-npx oh-my-design-cli install-skills
-```
-
-설치 후 에이전트를 재시작하세요 (Claude Code는 Cmd+Q 후 재실행) — 새 스킬 + 에이전트가 로드됩니다.
-
-이것이 당신이 실행할 유일한 CLI 명령어입니다. 나머지는 전부 에이전트에게 자연어로 말하면 됩니다.
-
-설치 시 **어디에** 설치할지 묻습니다: **Project** (`./.claude/skills` — 이 프로젝트만, 기본값) 또는 **Global** (`~/.claude/skills` — 모든 프로젝트; 스킬 + 서브에이전트, 전역 hooks/settings는 건드리지 않음). `--global` 플래그로 프롬프트를 건너뛸 수 있습니다.
-
-## 첫 60초
-
-이게 핵심입니다: 프롬프트 한 줄이 에이전트가 모든 미래 세션에서 기억하는 `DESIGN.md`가 됩니다.
-
-1. 설치(위) 후 **에이전트 재시작** (Cmd+Q, 재실행) — 새 스킬이 로드됩니다.
-
-2. 프로젝트에서 첫 프롬프트를 입력하세요 — 그대로 복사해도 됩니다:
-
-   > Set up our design system — Toss-style, for a family meal-tracking app.
-
-   에이전트가 **`omd:init`** 을 실행합니다: 246개 실제 기업 카탈로그에서 레퍼런스를 추천하고, 확인을 받은 뒤 프로젝트 루트에 **`DESIGN.md`** 를 작성합니다. (`omd:sync`가 `CLAUDE.md` / `AGENTS.md` / Cursor shim까지 연결해 모든 에이전트가 자동으로 읽게 합니다.)
-
-   **그 `DESIGN.md`가 활성화 지점입니다 — 이제 에이전트가 당신의 브랜드를 기억합니다.**
-
-3. 이제 그 위에서 빌드하세요:
-
-   > Design the home screen.
-
-   에이전트가 `DESIGN.md`를 읽고 브랜드에 맞는 UI를 만듭니다. 디자인 시스템을 다시 설명할 일은 없습니다.
-
-Toss가 아니어도 됩니다 — `Stripe-style`, `Linear-clone B2B SaaS`, `Karrot-style marketplace` 전부 가능. 전체 카탈로그: [oh-my-design.kr/design-systems](https://oh-my-design.kr/design-systems).
-
-## 지원 에이전트
-
-| 에이전트 | 채널 | 설치되는 것 |
-|---|---|---|
-| **Claude Code** | `--agent claude-code` (기본) | 풀 번들 — `.claude/` 아래 스킬, 16 서브에이전트, hooks, data |
-| **Codex** | `--agent codex` | `.agents/skills/` 스킬 번들 (공식 discovery 경로) |
-| **OpenCode** | `--agent opencode` | `.opencode/skills/` 스킬 번들 |
-| **Cursor** | `--agent cursor` | 정식 rules 채널 — `.cursor/rules/omd-design.mdc` shim + 공용 `.claude/data` 카탈로그 (스킬/훅 미설치) |
-
-기본 설치는 감지된 모든 에이전트를 대상으로 합니다; 단일 채널만 원하면 `--agent <name>`.
-
-## 패키지 구성
-
-**17 스킬 · 16 서브에이전트 · 246 검증된 레퍼런스 · 활성화 hooks** — 위 명령어 하나로 전부 설치됩니다.
-
-- **스킬** — core flow (`omd:init` / `omd:apply` / `omd:harness` / `omd:sync` / `omd:remember` / `omd:learn` / `omd:taste` — "내 취향 보여줘" 한마디로 루프가 배운 것·대기 중·보류된 것을 한 뷰로), 라이브 캡처 + 에셋 (`omd:reference-capture` / `omd:asset-fetch` / `omd:experiment-gallery`), v0.2 agent layer (`omd:orchestrator` / `omd:kr-writer` / `omd:locale-adapter` / `omd:designer-review` / `omd:final-qa` / `omd:codex-image`), 그리고 터미널에서 claude.ai/design을 구동하는 단독 스킬 `claude-design`.
-- **서브에이전트** — `omd-master` + 15 스페셜리스트 (UX 리서치, UI 생성, 에셋 큐레이션, 마이크로카피, a11y 감사, 페르소나 테스트, 비평, …).
-- **레퍼런스** — 246개 실제 기업 `DESIGN.md`, 전부 라이브 소스 대조 검증. 모든 레퍼런스는 `oh-my-design.kr/design-systems/<id>.md`에서 raw markdown으로도 제공되어 에이전트가 직접 fetch할 수 있습니다.
-- **Hooks** — UserPromptSubmit / SessionStart / PostToolUse 활성화 — 슬래시 명령 없이 자연어만으로 스킬이 발동합니다.
-
-스킬·에이전트별 상세 레퍼런스: **[oh-my-design.kr/docs](https://oh-my-design.kr/docs)**.
-
-MCP를 선호한다면 **[oh-my-design-mcp](./packages/mcp/)** — 같은 카탈로그를 Claude Desktop, Cursor, Cline, Continue, Codex용 MCP 리소스/툴로 노출합니다. [`packages/mcp/README.md`](./packages/mcp/README.md) 참고.
-
-## 업그레이드
-
-```bash
-npx oh-my-design-cli@latest install-skills
-```
-
-Idempotent. `<!-- omd:installed-skill -->` 마커가 있는 관리 파일은 in-place 갱신, 사용자가 수정한 파일은 그대로 둡니다 (`skipped-drift`) — 덮어쓰려면 `--force`. 재실행 후 에이전트 재시작.
-
-```bash
-npx oh-my-design-cli --version       # 프로젝트에 깔린 버전
-npm view oh-my-design-cli version    # npm registry 최신
-```
-
-릴리스별 변경 사항 — re-install 이상이 필요한 변경 포함 — 은 [CHANGELOG.md](./CHANGELOG.md)에 있습니다.
-
-## 링크
-
-- **카탈로그** — [oh-my-design.kr/design-systems](https://oh-my-design.kr/design-systems) (모든 레퍼런스 + 에이전트용 raw `.md` twin)
-- **컬렉션** — [oh-my-design.kr/collections](https://oh-my-design.kr/collections) (유스케이스별 큐레이션)
-- **문서** — [oh-my-design.kr/docs](https://oh-my-design.kr/docs) (설치 옵션, 스킬, 에이전트, FAQ)
-- **체인지로그** — [CHANGELOG.md](CHANGELOG.md) · 0.1.x에서 마이그레이션: [MIGRATION.md](MIGRATION.md)
-
-## 라이선스
-
-MIT — [LICENSE](LICENSE) 참고. 레퍼런스는 각 기업의 자산이며, 교육적 참조 목적으로만 재구성되었습니다.

package/.opencode/oh-my-design/README.md

@@ -1,112 +0,0 @@
-<p align="center">
-  <img src=".github/assets/logo-bg.png" width="480" alt="oh-my-design" />
-</p>
-
-<h1 align="center">oh-my-design</h1>
-
-<p align="center">
-  <strong>One-command bootstrap for skill-driven design with your AI coding agent.</strong> 246 real company design systems. Zero AI calls in the install. Then you just talk to your agent.
-</p>
-
-<p align="center">
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/v/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm version" /></a>
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/dm/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm downloads" /></a>
-  <a href="LICENSE"><img src="https://img.shields.io/github/license/kwakseongjae/oh-my-design?style=flat-square" alt="License" /></a>
-  <a href="https://github.com/kwakseongjae/oh-my-design/stargazers"><img src="https://img.shields.io/github/stars/kwakseongjae/oh-my-design?style=social" alt="GitHub Stars" /></a>
-  <img src="https://img.shields.io/badge/references-246-7c5cfc?style=flat-square" alt="246 References" />
-  <img src="https://img.shields.io/badge/CLI%20commands-1-blue?style=flat-square" alt="One CLI command" />
-</p>
-
-<p align="center">
-  English | <a href="README.ko.md">한국어</a> | <a href="README.ja.md">日本語</a> | <a href="README.zh-TW.md">繁體中文</a>
-</p>
-
----
-
-## What is oh-my-design?
-
-**oh-my-design (OmD)** is a design system for AI coding agents. It turns Claude Code / Codex / OpenCode / Cursor into a senior product designer with a working memory of your brand. You install once. After that, you just describe what you want — components, screens, copy, assets, charts — and the agent applies your project's design system and ships. `DESIGN.md` is the brand spec ([Google Stitch](https://stitch.withgoogle.com/docs/design-md/overview/) tokens + a brand-philosophy layer: Voice / Narrative / Principles / Personas / States / Motion), and 246 real-company DESIGN.md files ship in this package. **No API keys. No external infra. Everything runs inside your existing CLI session.**
-
-## Install
-
-```bash
-npx oh-my-design-cli install-skills
-```
-
-Then restart your agent (Cmd+Q in Claude Code, then relaunch) so the new skills + agents are loaded.
-
-That is the only CLI command you will run. Everything else is natural language to your agent.
-
-The installer asks **where** to install: **Project** (`./.claude/skills` — this project only, the default) or **Global** (`~/.claude/skills` — every project; skills + sub-agents, leaves your global hooks/settings untouched). Skip the prompt with `--global` for the user-level install.
-
-## Your first 60 seconds
-
-This is the whole point: one prompt turns into a `DESIGN.md` your agent remembers across every future session.
-
-1. Install (above), then **restart your agent** (Cmd+Q, relaunch) so it loads the new skills.
-
-2. In your project, type your first prompt — copy this exactly:
-
-   > Set up our design system — Toss-style, for a family meal-tracking app.
-
-   Your agent runs **`omd:init`**: it recommends a reference from the 246 real-company catalog, asks you to confirm, and writes **`DESIGN.md`** to your project root. (`omd:sync` then wires up the `CLAUDE.md` / `AGENTS.md` / Cursor shims so every agent reads it automatically.)
-
-   **That `DESIGN.md` is your activation — your agent now remembers your brand.**
-
-3. Now build against it:
-
-   > Design the home screen.
-
-   The agent reads `DESIGN.md` and ships brand-correct UI. No re-explaining your design system, ever again.
-
-Don't want Toss? Any brand works — `Stripe-style`, `Linear-clone B2B SaaS`, `Karrot-style marketplace`. Browse the full catalog at [oh-my-design.kr/design-systems](https://oh-my-design.kr/design-systems).
-
-## Supported agents
-
-| Agent | Channel | What gets installed |
-|---|---|---|
-| **Claude Code** | `--agent claude-code` (default) | Full bundle — skills, 16 sub-agents, hooks, data under `.claude/` |
-| **Codex** | `--agent codex` | Skill bundle at `.agents/skills/` (official discovery path) |
-| **OpenCode** | `--agent opencode` | Skill bundle at `.opencode/skills/` |
-| **Cursor** | `--agent cursor` | First-class rules channel — `.cursor/rules/omd-design.mdc` shim + the shared `.claude/data` catalog (no skills/hooks) |
-
-The default install targets every detected agent; pass `--agent <name>` to install a single channel.
-
-## What's inside
-
-**17 skills · 16 sub-agents · 246 verified references · activation hooks** — installed by the one command above.
-
-- **Skills** — core flow (`omd:init` / `omd:apply` / `omd:harness` / `omd:sync` / `omd:remember` / `omd:learn` / `omd:taste` — say "what are my preferences" to see everything the loop has learned, pending, or snoozed), live capture + assets (`omd:reference-capture` / `omd:asset-fetch` / `omd:experiment-gallery`), the v0.2 agent layer (`omd:orchestrator` / `omd:kr-writer` / `omd:locale-adapter` / `omd:designer-review` / `omd:final-qa` / `omd:codex-image`), plus the standalone `claude-design` skill that drives claude.ai/design from your terminal.
-- **Sub-agents** — `omd-master` + 15 specialists (UX research, UI generation, asset curation, microcopy, a11y audit, persona testing, critique, …).
-- **References** — 246 real-company `DESIGN.md` files, each verified against live sources. Every reference is also served as raw markdown at `oh-my-design.kr/design-systems/<id>.md`, so agents can fetch it directly.
-- **Hooks** — UserPromptSubmit / SessionStart / PostToolUse activation so the skills trigger on natural language, not just slash commands.
-
-Full skill-by-skill and agent-by-agent reference: **[oh-my-design.kr/docs](https://oh-my-design.kr/docs)**.
-
-Prefer MCP? **[oh-my-design-mcp](./packages/mcp/)** exposes the same catalog as MCP resources/tools for Claude Desktop, Cursor, Cline, Continue, and Codex — see [`packages/mcp/README.md`](./packages/mcp/README.md).
-
-## Upgrading
-
-```bash
-npx oh-my-design-cli@latest install-skills
-```
-
-Idempotent. Managed files (those carrying the `<!-- omd:installed-skill -->` marker) are refreshed in place; files you edited are left untouched (`skipped-drift`) unless you pass `--force`. Restart your agent after re-running.
-
-```bash
-npx oh-my-design-cli --version       # what your project currently uses
-npm view oh-my-design-cli version    # latest on the registry
-```
-
-What changed each release — including anything that needs more than a re-install — is in [CHANGELOG.md](./CHANGELOG.md).
-
-## Links
-
-- **Catalog** — [oh-my-design.kr/design-systems](https://oh-my-design.kr/design-systems) (every reference, with raw `.md` twins for agents)
-- **Collections** — [oh-my-design.kr/collections](https://oh-my-design.kr/collections) (curated sets by use case)
-- **Docs** — [oh-my-design.kr/docs](https://oh-my-design.kr/docs) (install options, skills, agents, FAQ)
-- **Changelog** — [CHANGELOG.md](CHANGELOG.md) · migrating from 0.1.x: [MIGRATION.md](MIGRATION.md)
-
-## License
-
-MIT — see [LICENSE](LICENSE). References belong to their respective companies; reproduced for educational reference.

package/.opencode/oh-my-design/README.zh-TW.md

@@ -1,74 +0,0 @@
-<p align="center">
-  <img src=".github/assets/logo-bg.png" width="480" alt="oh-my-design" />
-</p>
-
-<h1 align="center">oh-my-design</h1>
-
-<p align="center">
-  <strong>為 AI 編碼代理打造的技能驅動設計 — 一個指令完成引導。</strong>246 個真實企業設計系統。安裝過程零 AI 呼叫。之後只要和你的代理對話即可。
-</p>
-
-<p align="center">
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/v/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm version" /></a>
-  <a href="https://www.npmjs.com/package/oh-my-design-cli"><img src="https://img.shields.io/npm/dm/oh-my-design-cli?style=flat-square&color=cb3837" alt="npm downloads" /></a>
-  <a href="LICENSE"><img src="https://img.shields.io/github/license/kwakseongjae/oh-my-design?style=flat-square" alt="License" /></a>
-  <a href="https://github.com/kwakseongjae/oh-my-design/stargazers"><img src="https://img.shields.io/github/stars/kwakseongjae/oh-my-design?style=social" alt="GitHub Stars" /></a>
-  <img src="https://img.shields.io/badge/references-246-7c5cfc?style=flat-square" alt="246 References" />
-  <img src="https://img.shields.io/badge/CLI%20commands-1-blue?style=flat-square" alt="One CLI command" />
-</p>
-
-<p align="center">
-  繁體中文 | <a href="README.md">English</a> | <a href="README.ko.md">한국어</a> | <a href="README.ja.md">日本語</a>
-</p>
-
----
-
-## 什麼是 oh-my-design?
-
-**oh-my-design (OmD)** 是為 AI 編碼代理打造的設計系統。它讓 Claude Code / Codex / OpenCode / Cursor 變成一位記得你品牌的資深產品設計師。安裝一次之後,只要描述你想要的東西 — 元件、畫面、文案、素材、圖表 — 代理就會套用專案的設計系統並交付成果。`DESIGN.md` 是品牌規格([Google Stitch](https://stitch.withgoogle.com/docs/design-md/overview/) 令牌 + 品牌哲學層: Voice / Narrative / Principles / Personas / States / Motion),套件內建 246 個真實企業的 DESIGN.md。**無須 API 金鑰。無須外部基礎設施。一切都在你既有的 CLI 工作階段內執行。**
-
-## 安裝
-
-```bash
-npx oh-my-design-cli install-skills
-```
-
-安裝後請重新啟動你的代理 (Claude Code 為 Cmd+Q 後重新開啟) — 新的 skills + agents 才會載入。
-
-這是你唯一需要執行的 CLI 指令。其餘一切都是對代理說的自然語言。
-
-## 支援的代理
-
-| 代理 | 通道 | 安裝內容 |
-|---|---|---|
-| **Claude Code** | `--agent claude-code` (預設) | 完整套件 — `.claude/` 下的 skills、16 個子代理、hooks、data |
-| **Codex** | `--agent codex` | `.agents/skills/` 技能套件 (官方 discovery 路徑) |
-| **OpenCode** | `--agent opencode` | `.opencode/skills/` 技能套件 |
-| **Cursor** | `--agent cursor` | 正式 rules 通道 — `.cursor/rules/omd-design.mdc` shim + 共用 `.claude/data` 目錄 (不含 skills/hooks) |
-
-預設會安裝到所有偵測到的代理; 只要單一通道請加 `--agent <name>`。
-
-## 套件內容
-
-**17 個 skills · 16 個子代理 · 246 個經驗證的參考 · 活性化 hooks** — 上述一個指令全部安裝完成。
-
-每個參考也以 raw markdown 形式提供於 `oh-my-design.kr/design-systems/<id>.md`,代理可以直接抓取。完整的 skill 與 agent 參考文件: **[oh-my-design.kr/docs](https://oh-my-design.kr/docs)**。
-
-## 升級
-
-```bash
-npx oh-my-design-cli@latest install-skills
-```
-
-Idempotent。帶有 `<!-- omd:installed-skill -->` 標記的受管檔案會就地更新,使用者編輯過的檔案則保持不動 (要覆寫請加 `--force`)。重新執行後請重新啟動代理。
-
-## 連結
-
-- **目錄** — [oh-my-design.kr/design-systems](https://oh-my-design.kr/design-systems)
-- **精選集** — [oh-my-design.kr/collections](https://oh-my-design.kr/collections)
-- **文件** — [oh-my-design.kr/docs](https://oh-my-design.kr/docs)
-- **更新紀錄** — [CHANGELOG.md](CHANGELOG.md)
-
-## 授權
-
-[MIT](LICENSE) — 參考資料屬於各企業所有,僅為教育性參考而重現。

package/.opencode/oh-my-design/agents/AGENT.md

@@ -1,53 +0,0 @@
-# AGENT.md — oh-my-design harness compact context (≤ 60 lines, Ralph convention)
-
-A 60-line operational context card. Loaded into the project at install via `omd install-skills`. Source of truth for what an agent needs to know about *this* project on every iteration.
-
-## Identity
-- **Project:** oh-my-design — a CLI + skill bundle that turns AI coding agents into a design harness.
-- **Output artifact:** `DESIGN.md` (Google Stitch 1–9 + OmD 10–15) + harness run dir at `.omd/runs/`.
-- **Target users:** vibe coders, junior designers, solo founders.
-
-## Stack & commands
-- Build: `npm run build` (tsup) · Lint: `npm run lint` (tsc --noEmit) · Test: `npm test` (vitest)
-- Hidden internal helper: `omd harness "<task>" --internal` (slug + run dir; emits single JSON line)
-- User-facing entry: `/omd-harness <task>` (Claude Code or Codex skill)
-
-## Directory map
-- `agents/` — canonical 11 sub-agent definitions (master + 10 specialists; source-of-truth, generates `.claude/agents/` and `.codex/agents/`)
-- `skills/omd-harness/` — entry skill that calls master orchestrator
-- `skills/omd-init/`, `skills/omd-apply/`, `skills/omd-remember/`, `skills/omd-learn/`, `skills/omd-sync/` — existing OmD skills
-- `data/reference-fingerprints.json` — 67-reference voice fingerprint manifest (skill-side semantic match)
-- `data/vocabulary.json`, `data/synonyms.json`, `data/reference-tags.md` — controlled vocab + tagging
-- `references/<id>/DESIGN.md` × 67 — bundled reference design systems
-- `spec/omd-v0.1.md` — OmD spec (15-section DESIGN.md format)
-- `.omd/runs/run-<ts>-<slug>/` — every harness run; permanent archive
-- `skills/omd-lab-02-design-harness/runs/v<N>-...` — Lab #02 versioned experiments
-
-## Core architecture (re-read every Phase 8)
-- **Single Ralph-style loop**, fan-out only at: ux-research (parallel readers), persona-testing (parallel adversarial)
-- **OmD spec is already a harness signature** — §10 → microcopy, §12 → orchestrator constraint, §13 → tester input, §14 → QA checklist
-- **Eval pipeline**: deterministic gate (axe + lighthouse + spec validate) → cross-judge ensemble (3-prompt diversity, no API key) → adversarial persona walkthrough (4 personas, ABANDON token, hard turn budget)
-- **3 mandatory user checkpoints**: Phase 3 (journey), Phase 5 (DESIGN.md.patch), Phase 8 (validation summary)
-- **3-iteration cap**, critic between iterations forces root-cause re-entry (no surface patches)
-
-## Brand non-negotiables (apply to *this* project's own UI)
-- No Inter on white. No purple-on-white. No gradient backgrounds. No unjustified emoji.
-- Korean and Latin script co-equal (this is a bilingual project).
-- Hex + name together for every color: `Toss Blue (#3182f6)`, never just `#3182f6`.
-
-## Numbered-9s (re-read at every iteration boundary)
-- 9 → re-read sub-agent output before acting
-- 99 → user feedback → trace to Phase decision (critic), not surface-patch
-- 999 → never fabricate §11–13 facts
-- 9999 → never invent tokens absent from DESIGN.md
-- 99999 → never auto-skip user checkpoints
-- 999999 → never invent reference ids; 67-only catalog
-- 9999999 → never claim success when output is empty; Read the file
-- 99999999 → never overwrite previous iteration; snapshot to `iteration-<N>/`
-
-## Where to look when stuck
-- `research/harness-design/06-omd-integration-design.md` — full architecture
-- `research/harness-design/07-poc-spec.md` — Scenario A/B/C
-- `skills/omd-lab-02-design-harness/runs/live-*.md` — accumulated live-run insights
-- `research/harness-design/docs-todo.md` — pending docs/blogs queue
-

package/.opencode/oh-my-design/agents/omd-a11y-auditor.md

@@ -1,97 +0,0 @@
----
-name: omd-a11y-auditor
-description: Stage 0 deterministic gate of the eval pipeline. Runs DESIGN.md spec validation, axe-core, lighthouse, and Tier-1 official-DS URL liveness. Pass/fail is binary. Never opinion-based — always tool-output-based.
-tools: Read, Bash, Glob, WebFetch, Write
-model: haiku
----
-
-# omd-a11y-auditor
-
-You are the deterministic gate. You run mechanical checks. You don't opinionate. You either find a violation (with exact tool output) or pass.
-
-## Inputs
-
-- `run_dir`: current `.omd/runs/run-<ts>-<slug>/`
-- `design_md_path`: project DESIGN.md
-- `references_path`: `references-cited.md`
-- `output_path`: `eval/deterministic.json`
-
-## Checks
-
-### Check 1 — DESIGN.md spec validation
-
-Read DESIGN.md. Verify:
-
-- YAML frontmatter present with `omd:` and `brand:`
-- Sections 1-5 present and non-empty (REQUIRED)
-- Sections 6-9 present (RECOMMENDED — warning, not failure)
-- Sections 10-15 present (RECOMMENDED — warning, not failure)
-- §11/12/13 do not contain `[FILL IN]` placeholders if user provided facts in Phase 1 (read `brief.md` to determine)
-
-### Check 2 — Tier-1 official-DS URL liveness
-
-For each URL in `references-cited.md` marked Tier-1, run:
-
-```bash
-curl -s -o /dev/null -w "%{http_code}" -L --max-time 10 "<URL>"
-```
-
-200/301/302 = pass. Other = fail. Record actual code.
-
-### Check 3 — axe-core (if Playwright MCP available)
-
-If `mcp__playwright__*` tools available AND wireframes have a renderable preview:
-
-```bash
-# Requires axe-core npm package — check first
-npx --yes @axe-core/cli@latest <URL> --reporter v2 > eval/axe-output.json 2>&1 || echo "axe-core unavailable"
-```
-
-If unavailable, mark this check as `skipped: "axe-core requires Playwright + rendered HTML"`.
-
-### Check 4 — lighthouse (if available)
-
-```bash
-npx --yes lighthouse <URL> --only-categories=accessibility --output=json --output-path=eval/lighthouse.json --chrome-flags="--headless" 2>/dev/null || echo "lighthouse unavailable"
-```
-
-Pass if accessibility score ≥ 90.
-
-### Check 5 — Wireframe state completeness
-
-For each `wireframes/*.md`, verify the "States" table has all 5 rows (Empty / Loading / Error / Success / Skeleton). Missing any = violation.
-
-### Check 6 — Forbidden-phrase scan
-
-`grep` `components/microcopy.json` for forbidden phrases listed in DESIGN.md §10. Any hit = violation.
-
-## Output
-
-Write `eval/deterministic.json`:
-
-```json
-{
-  "run_id": "run-<ts>-<slug>",
-  "audited_at": "<ISO>",
-  "checks": {
-    "design_md_spec": { "status": "pass|fail|warn", "details": "..." },
-    "tier1_urls": { "status": "...", "results": [...] },
-    "axe_core": { "status": "pass|fail|skipped", "violations": [...] },
-    "lighthouse": { "status": "pass|fail|skipped", "score": 96 },
-    "wireframe_states": { "status": "...", "missing": [...] },
-    "forbidden_phrases": { "status": "...", "hits": [...] }
-  },
-  "verdict": "pass | fail",
-  "critical_failures": [],
-  "warnings": []
-}
-```
-
-`verdict = pass` ONLY if every required check is pass and no critical_failures.
-
-## Hard rules
-
-- You do NOT make opinion calls ("this looks good"). Only tool-output-based.
-- You do NOT skip a check silently. If unavailable, mark `skipped` with reason.
-- You do NOT modify any artifact. Read-only on the run except your own output.
-- If a tool errors, capture stderr verbatim into `details`.