agestra 4.9.0 → 4.10.1

package/.claude-plugin/marketplace.json

@@ -11,8 +11,8 @@
     {
       "name": "agestra",
       "source": "./",
-      "description": "Orchestrate Ollama, Gemini, and Codex for multi-AI debates, cross-validation, and GraphRAG memory",
-      "version": "4.9.0",
+      "description": "Orchestrate Ollama, Gemini, and Codex for multi-AI debates, code review, and cross-validation",
+      "version": "4.10.1",
       "author": {
         "name": "mua-vtuber"
       },
@@ -32,8 +32,8 @@
       "tags": [
         "orchestration",
         "debate",
-        "memory",
-        "graphrag"
+        "code-review",
+        "cross-validation"
       ]
     }
   ]

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "agestra",
-  "version": "4.9.0",
-  "description": "Claude Code plugin — orchestrate Ollama, Gemini, and Codex for multi-AI debates, cross-validation, and GraphRAG memory",
+  "version": "4.10.1",
+  "description": "Claude Code plugin — orchestrate Ollama, Gemini, and Codex for multi-AI debates, code review, and cross-validation",
   "mcpServers": {
     "agestra": {
       "command": "node",

package/README.ja.md

@@ -84,7 +84,7 @@ Gemini ではリポジトリ直下の [GEMINI.md](GEMINI.md) と、[`.gemini/com
 | `/agestra setup` | 初期 AI プロバイダー選択とセットアップ |
 | `/agestra implement [task]` | Claude only または Multi-AI モードで実装を進める |
 
-外部プロバイダーが利用可能な場合、テキストコマンド（review、design、idea）は直接ディベートモード（끝장토론）に入り、マルチ AI クロスバリデーションを行います。プロバイダーが検出されない場合、Claude が自動的に単独で作業します。
+外部プロバイダーが利用可能な場合、テキストコマンド（review、design、idea）は直接ディベートモードに入り、マルチ AI クロスバリデーションを行います。プロバイダーが検出されない場合、Claude が自動的に単独で作業します。
 
 ## エージェント
 
@@ -137,7 +137,7 @@ Turborepo モノレポで、9 パッケージ構成です:
 
 ### 作業モード
 
-**Text work**（レビュー、設計、アイデア）: プロバイダーあり → 끝장토론（ディベート）; なし → Claude only
+**Text work**（レビュー、設計、アイデア）: プロバイダーあり → ディベートモード; なし → Claude only
 
 **Implementation work**（team-lead orchestration）:
 - **Claude만으로** — Claude がプロジェクト/グローバルエージェントを使って直接実装します。
@@ -240,16 +240,18 @@ Turborepo モノレポで、9 パッケージ構成です:
 
 ## 設定
 
-### providers.config.json (任意)
+### providers.config.json
 
-Agestra は起動時にプロバイダーを自動検出します。手動で制御したい場合は、プロジェクトルートに `providers.config.json` を作成してください:
+`/agestra setup` が生成します。単一のプラグイン範囲の場所に保存されます（解決順: `AGESTRA_CONFIG_PATH` 環境変数 → `$CLAUDE_PLUGIN_ROOT/providers.config.json` → `~/.agestra/providers.config.json`）。リポジトリには置かない方針で、gitignore 済みです。
 
 | 項目 | 説明 |
 |------|------|
-| `defaultProvider` | 未指定時に使うプロバイダー ID |
+| `selectionPolicy` | `"default-only"`（現状サポートされる唯一の値） |
+| `locale` | モデレーターの UI ロケール (`ko`/`en`/`ja`/`zh`) |
 | `providers[].id` | 一意の識別子 |
-| `providers[].type` | `ollama`, `gemini-cli`, `codex-cli` |
-| `providers[].enabled` | 起動時に読み込むか |
+| `providers[].type` | `ollama`, `gemini-cli`, `codex-cli`, `claude-cli` |
+| `providers[].enabled` | 起動時に登録するか — `false` は明示的オプトアウト |
+| `providers[].executionPolicy` | `read-only` または `workspace-write` |
 | `providers[].config` | タイプ別設定（host、timeout など） |
 
 ### ランタイムデータ

package/README.ko.md

@@ -281,16 +281,18 @@ Turborepo 모노레포, 9개 패키지:
 
 ## 설정
 
-### providers.config.json (선택)
+### providers.config.json
 
-Agestra는 시작 시 공급자를 자동 감지합니다. 수동 제어가 필요하면 프로젝트 루트에 `providers.config.json`을 생성하세요:
+`/agestra setup`이 생성합니다. 단일 플러그인 범위 경로에 저장됩니다 (해석 우선순위: `AGESTRA_CONFIG_PATH` 환경변수 → `$CLAUDE_PLUGIN_ROOT/providers.config.json` → `~/.agestra/providers.config.json`). 프로젝트 저장소에 두지 않으며 gitignore에도 등록되어 있습니다.
 
 | 필드 | 설명 |
 |------|------|
-| `defaultProvider` | 미지정 시 사용할 공급자 ID |
+| `selectionPolicy` | `"default-only"` (현재 지원 값) |
+| `locale` | 중재자 내러티브 로케일 (`ko`/`en`/`ja`/`zh`) |
 | `providers[].id` | 고유 식별자 |
-| `providers[].type` | `ollama`, `gemini-cli`, `codex-cli` |
-| `providers[].enabled` | 시작 시 로드 여부 |
+| `providers[].type` | `ollama`, `gemini-cli`, `codex-cli`, `claude-cli` |
+| `providers[].enabled` | 시작 시 등록 여부 — `false`면 강제 제외 |
+| `providers[].executionPolicy` | `read-only` 또는 `workspace-write` |
 | `providers[].config` | 타입별 설정 (host, timeout 등) |
 
 ### 런타임 데이터

package/README.md

@@ -125,7 +125,7 @@ All three hosts drive the same MCP server and shared workflow specs from `comman
 | `/agestra setup` | Initial AI provider selection and setup |
 | `/agestra implement [task]` | Execute implementation in Claude-only or Multi-AI mode |
 
-When external providers are available, text commands (review, design, idea) go directly to debate mode (끝장토론) for multi-AI cross-validation. When no providers are detected, Claude works alone automatically.
+When external providers are available, text commands (review, design, idea) go directly to debate mode for multi-AI cross-validation. When no providers are detected, Claude works alone automatically.
 
 ## Agents
 
@@ -178,7 +178,7 @@ Turborepo monorepo with 9 packages:
 
 ### Work Modes
 
-**Text work** (review, design, idea): providers available → 끝장토론 (debate); no providers → Claude only
+**Text work** (review, design, idea): providers available → debate mode; no providers → Claude only
 
 **Implementation work** (team-lead orchestration):
 - **Claude만으로** — Claude implements directly with project/global agents.
@@ -281,16 +281,18 @@ Turborepo monorepo with 9 packages:
 
 ## Configuration
 
-### providers.config.json (Optional)
+### providers.config.json
 
-Agestra auto-detects providers at startup. For manual control, create `providers.config.json` in the project root:
+Created by `/agestra setup`. The file lives at a single plugin-scoped location (resolution order: `AGESTRA_CONFIG_PATH` env var → `$CLAUDE_PLUGIN_ROOT/providers.config.json` → `~/.agestra/providers.config.json`). It is not meant to sit in the project repo and is gitignored accordingly.
 
 | Field | Description |
 |-------|-------------|
-| `defaultProvider` | Provider ID when none specified |
+| `selectionPolicy` | `"default-only"` (only supported value today) |
+| `locale` | UI locale for moderator narration (`ko`/`en`/`ja`/`zh`) |
 | `providers[].id` | Unique identifier |
-| `providers[].type` | `ollama`, `gemini-cli`, or `codex-cli` |
-| `providers[].enabled` | Load at startup |
+| `providers[].type` | `ollama`, `gemini-cli`, `codex-cli`, or `claude-cli` |
+| `providers[].enabled` | Whether to register this provider at startup — hard opt-out when `false` |
+| `providers[].executionPolicy` | `read-only` or `workspace-write` |
 | `providers[].config` | Type-specific settings (host, timeout, etc.) |
 
 ### Runtime Data

package/README.zh.md

@@ -84,7 +84,7 @@ Gemini 会结合仓库根目录下的 [GEMINI.md](GEMINI.md) 与 [`.gemini/comma
 | `/agestra setup` | 初始 AI 提供方选择与设置 |
 | `/agestra implement [task]` | 以 Claude only 或 Multi-AI 模式执行实现 |
 
-当外部提供方可用时，文本命令（review、design、idea）直接进入辩论模式（끝장토론），进行多 AI 交叉验证。当未检测到提供方时，Claude 自动独立工作。
+当外部提供方可用时，文本命令（review、design、idea）直接进入辩论模式，进行多 AI 交叉验证。当未检测到提供方时，Claude 自动独立工作。
 
 ## 代理
 
@@ -137,7 +137,7 @@ Gemini 会结合仓库根目录下的 [GEMINI.md](GEMINI.md) 与 [`.gemini/comma
 
 ### 工作模式
 
-**文本工作**（review、design、idea）：有提供方 → 끝장토론（辩论）；无提供方 → Claude only
+**文本工作**（review、design、idea）：有提供方 → 辩论模式；无提供方 → Claude only
 
 **实现工作**（team-lead orchestration）：
 - **Claude만으로** — Claude 直接结合项目/全局代理完成实现。
@@ -240,16 +240,18 @@ Gemini 会结合仓库根目录下的 [GEMINI.md](GEMINI.md) 与 [`.gemini/comma
 
 ## 配置
 
-### providers.config.json（可选）
+### providers.config.json
 
-Agestra 会在启动时自动检测提供方。如果需要手动控制，可在项目根目录创建 `providers.config.json`：
+由 `/agestra setup` 生成。文件存放于单一的插件范围位置（解析顺序：`AGESTRA_CONFIG_PATH` 环境变量 → `$CLAUDE_PLUGIN_ROOT/providers.config.json` → `~/.agestra/providers.config.json`）。不随仓库一起提交，已加入 gitignore。
 
 | 字段 | 说明 |
 |------|------|
-| `defaultProvider` | 未指定时使用的提供方 ID |
+| `selectionPolicy` | `"default-only"`（目前唯一支持的取值） |
+| `locale` | 主持人叙述的 UI 语言 (`ko`/`en`/`ja`/`zh`) |
 | `providers[].id` | 唯一标识符 |
-| `providers[].type` | `ollama`、`gemini-cli` 或 `codex-cli` |
-| `providers[].enabled` | 是否在启动时加载 |
+| `providers[].type` | `ollama`、`gemini-cli`、`codex-cli`、`claude-cli` |
+| `providers[].enabled` | 启动时是否注册 — `false` 为强制跳过 |
+| `providers[].executionPolicy` | `read-only` 或 `workspace-write` |
 | `providers[].config` | 类型相关配置（host、timeout 等） |
 
 ### 运行时数据

package/agents/agestra-moderator.md

@@ -250,57 +250,135 @@ Invoked when multiple AIs have independently analyzed the same target and their
 
 ### Mode: Document Review Round (Debate Phase 2)
 
-Invoked after Independent Aggregation has produced an initial document. The document is iteratively reviewed by all AIs until consensus.
+Invoked after Independent Aggregation has produced an initial (Round 0) working document. The document is iteratively reviewed until every participant's `<round-votes>` block reports no `disagree`.
 
-**Input:** Current document + list of participating providers.
+**Input:** Current working doc + ordered participant list.
 
-**Process (per round, no maximum — until all agree):**
+**Turn order within a round:** external providers alphabetical first, Claude last. Providers are invoked **sequentially, not in parallel** — each one must read every earlier Round `N` review before writing its own, so late-turn providers can endorse-and-move-on instead of re-stating. This is the single biggest lever against round-after-round churn.
 
-1. Send the current document to each AI for review:
-   - **Claude:** Spawn the appropriate specialist agent → analyze document → produce feedback.
-   - **External providers:** Call `agent_debate_turn` with the document as prompt context, requesting feedback on each section.
+**Per round `N` procedure:**
 
-2. Collect all feedback.
+1. For each participant in turn:
 
-3. For each section of the document:
-   - Count agree/disagree from each AI.
-   - If disagreement: extract the specific objection and proposed revision.
+   a. Build the prompt envelope:
+   - Working doc (full content).
+   - Round 0 aggregated doc (for anchoring the idea IDs).
+   - Every Round `N` review already written this round, by ID (fetched with `workspace_read`).
+   - The `<round-votes>` contract below.
 
-4. Revise disputed sections incorporating feedback:
-   - If a revision is supported by evidence or reasoning, apply it.
-   - If revisions contradict each other, present both positions in the document.
+   b. Send the review request. For Claude, spawn the appropriate specialist agent (`agestra-ideator`, `agestra-reviewer`, etc.). For external providers, use `ai_chat` with `save_as_document.kind = "debate"`.
 
-5. Track consensus status per section:
-   ```json
-   { "section": "Security", "status": "agreed", "round": 2 }
-   { "section": "Performance", "status": "disputed", "round": 2,
-     "positions": { "claude": "optimize later", "gemini": "optimize now" } }
+   c. File the reply as a dedicated round review doc (kind `"debate"`, title `Round {N} Review — {provider}`).
+
+2. **`<round-votes>` contract** (every reviewer appends this at the end of its review):
+
+   ```
+   <round-votes>
+     <item id="IDEA-01" stance="agree"/>
+     <item id="IDEA-02" stance="agree-with-note" note="priority should be MEDIUM not HIGH"/>
+     <item id="IDEA-03" stance="disagree" reason="conflicts with stated constraint X"/>
+     <item id="IDEA-04" stance="defer-to-codex"/>
+   </round-votes>
    ```
 
-6. **Consensus check:**
-   - All AIs agree on all sections → consensus reached. Proceed to final document.
-   - Disagreements remain → next round with the revised document.
-   - **Every 10 rounds:** Ask the user via AskUserQuestion whether to continue or stop with the current state. If the user stops, conclude with split positions documented.
+   Closed stance set: `agree | agree-with-note | disagree | defer-to-<providerId>`. One `<item>` per idea in the working doc. `reason` is required for `disagree`; `note` is required for `agree-with-note`; `defer-to-<providerId>` must reference a provider that already wrote a review earlier in the same round.
+
+3. **Round wrap-up** (moderator, after all reviewers finish):
+   - Absorb every `agree-with-note` the moderator judges well-reasoned; cite the source review in the working doc.
+   - Mark every unresolved `disagree` as Disputed, preserving both positions with proposer + reason verbatim.
+   - Record every `defer-to` edge for the final endorsement graph.
+
+4. **Consensus check:**
+   - Round `N` has zero `disagree` items AND every `agree-with-note` was either applied or explicitly declined-with-reason → consensus reached. Write the synthesis per `<Final_Consensus_Format>`.
+   - Otherwise → Round `N+1` with the revised working doc.
+   - **Every 10 rounds:** call `AskUserQuestion` offering `continue`, `stop with current state`, `escalate split positions`.
+
+**Why sequential:** parallel rounds produce the "everyone independently says the same thing" pathology the idea workflow suffered from before. Reading prior reviews lets a late-turn participant simply `stance="defer-to-{earlier}"` instead of re-writing three paragraphs that say the same thing.
+
+</Workflow_Document_Review_Round>
+
+<Final_Consensus_Format>
+
+### Mode: Final Consensus document (idea/review workflows)
 
-7. Return: revised document + consensus map.
+The idea/review final doc is for **humans skimming after the fact**, not for machine parsing. Write it for readability first: expanded prose, explicit per-idea reasoning, clear visual hierarchy. Avoid the metadata-header-plus-tables-only pattern — tables are a quick-reference layer, not the whole document.
+
+**Required sections (in order):**
+
+#### 1. 한눈에 보기 (Executive Summary)
+
+3–5 sentences in plain prose. What was decided, how many rounds it took, which dimensions were disputed, and what the reader should do with the doc. No tables here.
+
+#### 2. 참여자 및 라운드 요약
+
+Plain list (not a table):
+- **참여자:** `claude/ideator`, `codex`, … — one line per participant, with a short note if any dropped out mid-flow and why.
+- **라운드:** `{N}/{max}` with a one-sentence arc ("Round 1: big-picture disagreement on tool scope. Round 2: narrowed to implementation priorities. Round 3: consensus.").
+- **사용자 제약:** every binding constraint the user gave during Phase 1, verbatim.
+
+#### 3. 최종 결정 (Accepted Ideas)
+
+One **subsection per accepted idea** (not a single table row). For each idea:
 
-**Final document format:**
 ```markdown
-## Final Document
+### ✓ {idea title}  `IDEA-XX`
 
-### [Section — Consensus ✓]
-[content all parties agreed on]
+**요약.** 1–2문장으로 이 아이디어가 무엇이고 왜 채택됐는지.
 
-### [Section — Consensus ✓ (Round 3)]
-[content agreed after revision in round 3]
+**근거.** 구체 파일/패턴/사용자 피드백 인용. 제안자가 제시한 Evidence를 그대로.
 
-### [Section — No Consensus ✗]
-**Majority position:** [content]
-**Dissenting view ([provider]):** [alternative position]
-**Recommendation:** [moderator's neutral framing of the trade-off]
+**노력 · 우선순위.** {effort} · {priority} — 협의 결과면 그렇게 표기 ("Round 2 protocol: 노력 S→M, codex 근거 채택").
+
+**동의 현황.**
+- 🟢 agree: claude/ideator, codex
+- 🟡 agree-with-note: gemini — "MEDIUM이 더 현실적, HIGH는 낙관적"
+- 모두가 agree이면 "전원 agree — 이견 없음." 한 줄로 대체.
 ```
 
-</Workflow_Document_Review_Round>
+Emoji indicators (🟢 agree / 🟡 note / 🔴 disagree) are load-bearing: they're the "at-a-glance 동의 어디서 하는지" signal the user asked for. Use them every time.
+
+#### 4. 분쟁 항목 (Disputed Ideas)
+
+One subsection per item where Round `N-final` still had any `disagree`. Format:
+
+```markdown
+### ✗ {idea title}  `IDEA-YY`  — Disputed
+
+**쟁점.** 1–2문장으로 무엇이 걸려 있는지.
+
+**입장 대립.**
+- **{provider-A}** — (agree/disagree/revise). 주장: "…". 근거: "…"
+- **{provider-B}** — …
+
+**중재자 노트.** 왜 합의가 안 됐는지, 어떤 추가 정보가 있으면 결정 가능한지. 본인 의견은 주입하지 않는다.
+```
+
+#### 5. 범위 외 / 보류
+
+한 줄씩. 왜 보류했는지 포함.
+
+#### 6. 전체 동의 매트릭스 (at-a-glance reference)
+
+마지막에 한 번만 나오는 축약 표 — 3~5번 섹션의 중복이 아니라 **스캔용**. 열은 아이디어 ID, 행은 프로바이더, 셀은 🟢/🟡/🔴만.
+
+```markdown
+| Idea | claude/ideator | codex | gemini |
+|------|----------------|-------|--------|
+| IDEA-01 Auto-focus on Create | 🟢 | 🟢 | 🟢 |
+| IDEA-02 숫자 키 타입 선택 | 🟢 | 🟡 | 🟢 |
+| IDEA-03 Tab-to-Next-Field | 🟢 | 🟢 | 🔴 |
+```
+
+#### 7. 참조 문서
+
+Round 0 aggregated doc, 각 라운드의 provider별 review doc, 원본 individual docs — 전부 ID 리스트. 독자가 전체 전사본으로 내려갈 수 있게.
+
+**금지 사항:**
+- "참여자: A, B, C" 한 줄로 동의 내역을 뭉뚱그리지 말 것. 누가 무엇에 동의했는지 아이디어별로 분리.
+- 표 하나로 모든 근거를 퉁치지 말 것. 표는 축약이지 본문이 아니다.
+- 메타데이터만 잔뜩 적고 본문이 얇은 상태로 닫지 말 것. 각 아이디어의 "요약 + 근거" 본문이 항상 표보다 먼저.
+
+</Final_Consensus_Format>
 
 <Workflow_Conflict_Resolution>
 

package/commands/design.md

@@ -1,76 +1,85 @@
----
-description: "Explore architecture and design trade-offs before implementation"
-argument-hint: "[idea, feature, or system to design]"
----
-
-You are executing the `/agestra design` command.
-
-**Subject:** $ARGUMENTS
-
-## Step 1: Determine design subject
-
-If `$ARGUMENTS` is empty, present a starting-point choice using AskUserQuestion (in the user's language):
-
-| Option | Description |
-|--------|-------------|
-| **Describe an idea** | User has a specific feature or system in mind — proceed to designer |
-| **Find ideas first** | User doesn't know what to design yet — run `/agestra idea` to discover opportunities, then return here |
-| **Use recent context** | Organize ideas from the current conversation into a design subject |
-
-- If **"Describe an idea"**: ask a follow-up "What would you like to design?" and proceed.
-- If **"Find ideas first"**: run the `agestra:agestra-ideator` agent (or `/agestra idea`) to generate suggestions. After the user selects an idea from the results, continue to Step 2 with that as the subject.
-- If **"Use recent context"**: scan the current conversation for previously discussed ideas, improvements, or features. Summarize them and ask the user which to design.
-
-If `$ARGUMENTS` is provided, use it directly as the subject.
-
-## Step 2: Check environment and select mode
-
-Call `environment_check` to determine which providers and modes are available.
-
-- If **no providers are available**: run the `agestra:agestra-designer` agent directly (Claude only) with the subject as context. The designer will ask questions to understand intent, explore the codebase for existing patterns, propose 2-3 approaches with trade-offs, refine based on feedback, and produce a design document in `docs/plans/`. Skip to presenting the result.
-- If **1+ providers are available**: proceed to 끝장토론 execution below.
-
-## Step 3: Execute 끝장토론
-
-**팀 구성:** `agestra:agestra-moderator` (리더) + `agestra:agestra-designer` (Claude) + 사용 가능한 외부 AI (gemini, codex, ollama 등)
-
-1. Independent work + initial aggregation:
-
-   a. In parallel:
-      - Spawn the `agestra:agestra-designer` agent for Claude's independent architecture exploration.
-        After the agent completes, save Claude's result as a document via `workspace_create_document`:
-        - **title:** `Architecture Design — claude/designer`
-        - **metadata:** `{ "Provider": "claude/designer", "Task": "{subject}", "Mode": "Independent" }`
-        - **content:** The designer agent's full output.
-      - For each available provider, call `ai_chat` with `save_as_document` to let each AI produce its own document directly:
-        - **save_as_document.title:** `Architecture Design — {provider}`
-        - **save_as_document.metadata:** `{ "Task": "{subject}", "Mode": "Independent" }`
-        - **prompt:**
-
-          > Create a complete design document for [subject]. Include: 1) Problem definition — what exactly we're solving and why. 2) Constraints and requirements — technical limits, compatibility needs, performance targets. 3) 2-3 architecture approaches — each with detailed pros/cons, component diagrams, data flow, and effort estimates. 4) Recommended approach — pick one and justify the choice. 5) Implementation plan — step-by-step build sequence with dependencies. 6) Risks and mitigations. Be thorough and detailed — this document will be debated and refined.
-          >
-          > Subject: [the design subject]
-
-   b. Collect all document IDs.
-
-   c. Spawn the `agestra:agestra-moderator` agent in **Independent Aggregation** mode:
-      - Pass the **document ID list**.
-      - Moderator reads each document via `workspace_read`.
-      - Moderator classifies: consensus approaches, unique ideas, disputed trade-offs.
-      - Moderator creates an **aggregated document** via `workspace_create_document`.
-      - The moderator's integrated document becomes the starting document.
-
-2. Document review rounds (no max — until all agree):
-   a. Moderator sends the current document to each AI for review:
-      - Claude: spawn `agestra:agestra-designer` → analyze document → write section-by-section feedback
-      - Other providers: `agent_debate_turn` with the document as prompt, requesting agree/disagree per section
-   b. Moderator collects all feedback.
-   c. Classify: agree/disagree per section per provider.
-   d. Revise document incorporating disagreement feedback.
-   e. If all providers agree on all sections → consensus reached.
-   f. If not → next round with revised document.
-   g. **Every 10 rounds:** Ask the user via AskUserQuestion whether to continue or stop with current state.
-
-3. Present the final document:
-   - Consensus sections: marked as agreed
-   - Disputed sections: show split positions with each provider's rationale
+---
+description: "Explore architecture and design trade-offs before implementation"
+argument-hint: "[idea, feature, or system to design]"
+---
+
+You are executing the `/agestra design` command.
+
+**Subject:** $ARGUMENTS
+
+## Step 0: Setup preflight (MANDATORY)
+
+Before anything else, call `setup_status`. If it reports `Current config: not found`, **stop this command and run setup first**:
+
+1. Invoke the `agestra:setup` skill (or `/agestra setup` inline) — provider detection, selection, locale, `setup_apply`.
+2. After the config is written, resume this `/agestra design` command **from Step 1**, preserving `$ARGUMENTS`. Do not ask the user to retype.
+
+Agestra uses a single plugin-scoped `providers.config.json`. No config → no sanctioned provider set → setup is the only correct starting point.
+
+## Step 1: Determine design subject
+
+If `$ARGUMENTS` is empty, present a starting-point choice using AskUserQuestion (in the user's language):
+
+| Option | Description |
+|--------|-------------|
+| **Describe an idea** | User has a specific feature or system in mind — proceed to designer |
+| **Find ideas first** | User doesn't know what to design yet — run `/agestra idea` to discover opportunities, then return here |
+| **Use recent context** | Organize ideas from the current conversation into a design subject |
+
+- If **"Describe an idea"**: ask a follow-up "What would you like to design?" and proceed.
+- If **"Find ideas first"**: run the `agestra:agestra-ideator` agent (or `/agestra idea`) to generate suggestions. After the user selects an idea from the results, continue to Step 2 with that as the subject.
+- If **"Use recent context"**: scan the current conversation for previously discussed ideas, improvements, or features. Summarize them and ask the user which to design.
+
+If `$ARGUMENTS` is provided, use it directly as the subject.
+
+## Step 2: Check environment and select mode
+
+Call `environment_check` to determine which providers and modes are available.
+
+- If **no providers are available**: run the `agestra:agestra-designer` agent directly (Claude only) with the subject as context. The designer will ask questions to understand intent, explore the codebase for existing patterns, propose 2-3 approaches with trade-offs, refine based on feedback, and produce a design document in `docs/plans/`. Skip to presenting the result.
+- If **1+ providers are available**: proceed to 끝장토론 execution below.
+
+## Step 3: Execute 끝장토론
+
+**팀 구성:** `agestra:agestra-moderator` (리더) + `agestra:agestra-designer` (Claude) + 사용 가능한 외부 AI (gemini, codex, ollama 등)
+
+1. Independent work + initial aggregation:
+
+   a. In parallel:
+      - Spawn the `agestra:agestra-designer` agent for Claude's independent architecture exploration.
+        After the agent completes, save Claude's result as a document via `workspace_create_document`:
+        - **title:** `Architecture Design — claude/designer`
+        - **metadata:** `{ "Provider": "claude/designer", "Task": "{subject}", "Mode": "Independent" }`
+        - **content:** The designer agent's full output.
+      - For each available provider, call `ai_chat` with `save_as_document` to let each AI produce its own document directly:
+        - **save_as_document.title:** `Architecture Design — {provider}`
+        - **save_as_document.metadata:** `{ "Task": "{subject}", "Mode": "Independent" }`
+        - **prompt:**
+
+          > Create a complete design document for [subject]. Include: 1) Problem definition — what exactly we're solving and why. 2) Constraints and requirements — technical limits, compatibility needs, performance targets. 3) 2-3 architecture approaches — each with detailed pros/cons, component diagrams, data flow, and effort estimates. 4) Recommended approach — pick one and justify the choice. 5) Implementation plan — step-by-step build sequence with dependencies. 6) Risks and mitigations. Be thorough and detailed — this document will be debated and refined.
+          >
+          > Subject: [the design subject]
+
+   b. Collect all document IDs.
+
+   c. Spawn the `agestra:agestra-moderator` agent in **Independent Aggregation** mode:
+      - Pass the **document ID list**.
+      - Moderator reads each document via `workspace_read`.
+      - Moderator classifies: consensus approaches, unique ideas, disputed trade-offs.
+      - Moderator creates an **aggregated document** via `workspace_create_document`.
+      - The moderator's integrated document becomes the starting document.
+
+2. Document review rounds (no max — until all agree):
+   a. Moderator sends the current document to each AI for review:
+      - Claude: spawn `agestra:agestra-designer` → analyze document → write section-by-section feedback
+      - Other providers: `agent_debate_turn` with the document as prompt, requesting agree/disagree per section
+   b. Moderator collects all feedback.
+   c. Classify: agree/disagree per section per provider.
+   d. Revise document incorporating disagreement feedback.
+   e. If all providers agree on all sections → consensus reached.
+   f. If not → next round with revised document.
+   g. **Every 10 rounds:** Ask the user via AskUserQuestion whether to continue or stop with current state.
+
+3. Present the final document:
+   - Consensus sections: marked as agreed
+   - Disputed sections: show split positions with each provider's rationale