@boyingliu01/xp-gate 0.10.11 → 0.10.13

package/plugins/opencode/index.ts

@@ -322,6 +322,8 @@ export const XpGatePlugin = async (input: OpenCodePluginInput) => {
       if (!checked) {
         checked = true
         const msg = await runBackgroundUpdates(directory).catch(() => null)
+        // Primary notification: upgrade-notice.json → TUI sidebar banner
+        // Fallback: stderr for users without TUI panel registered
         if (msg) process.stderr.write(`${msg}\n`)
       }
     },

package/plugins/opencode/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@boyingliu01/opencode-plugin",
-  "version": "0.10.11",
+  "version": "0.10.13",
   "type": "module",
   "main": "index.ts",
   "description": "XP-Gate quality gates + AI workflow skills + Sprint Flow TUI sidebar for OpenCode",

package/plugins/opencode/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-06-23
-**Commit:** bac916f
+**Generated:** 2026-06-24
+**Commit:** 5019773
 **Branch:** main
-**Version:** 0.10.11.0
+**Version:** 0.10.13.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, 3 experts from ≥2 providers, domestic models only). Supports design + code-walkthrough modes.

package/plugins/opencode/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-06-23
-**Commit:** bac916f
+**Generated:** 2026-06-24
+**Commit:** 5019773
 **Branch:** main
-**Version:** 0.10.11.0
+**Version:** 0.10.13.0
 
 ## OVERVIEW
 **11-phase** development pipeline: ISOLATE → AUTO-ESTIMATE → THINK → PLAN → BUILD → REVIEW → USER ACCEPTANCE → FEEDBACK → SHIP → LAND → CLEANUP. Phase 2 default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE in Phase 1: design must pass Delphi review (≥90% consensus) before any coding.

package/plugins/opencode/skills/sprint-flow/SKILL.md

@@ -237,11 +237,11 @@ Sprint Flow: ISOLATE → AUTO-ESTIMATE → THINK → PLAN → BUILD → REVIEW
 - **HARD-GATE**: 设计未批准 → 不可进入实现
 
 ### Phase 1: PLAN（共识评审）
-- **注意**: `autoplan` 和 `to-issues` 是交互式 skill，**必须由 orchestrator 直接执行**（Issue #225, #248）
-- **执行模式（三阶段）**:
+- **注意**: `autoplan`、`delphi-review` 和 `to-issues` 均为交互式 skill，**必须由 orchestrator 直接执行**（Issue #225, #248, #249）
+- **执行模式（三阶段，全部 orchestrator 直接执行）**:
   1. **Orchestrator 直接执行 autoplan**: `skill(name="autoplan")` → 用户确认 taste_decisions
-  2. **Orchestrator 直接执行 to-issues**: `skill(name="to-issues")` → 用户确认 Issue 拆分方案
-  3. **Subagent 执行 delphi-review**: `task(category="deep", load_skills=["delphi-review"])` — 非交互式，可自动运行至 APPROVED
+  2. **Orchestrator 直接执行 delphi-review**: `skill(name="delphi-review")` — 等待 APPROVED（非 APPROVED 时需等待用户确认/处理）
+  3. **Orchestrator 直接执行 to-issues**: `skill(name="to-issues")` → 用户确认 Issue 拆分方案
 - 输入: phase-0-summary.md + 设计文档
 - 输出: `specification.yaml`（含 user_stories[]）+ `slices-manifest.json`
 
@@ -276,13 +276,10 @@ Sprint Flow: ISOLATE → AUTO-ESTIMATE → THINK → PLAN → BUILD → REVIEW
 5. **Mock Minimization**: integration-first, mock 仅限 external services, 密度 > 30% 需 `@mock-justified`
 
 ### Phase 3: REVIEW + TEST（验证）
-- **Subagent dispatch**: orchestrator 通过 `task(category="deep", load_skills=["delphi-review", "test-specification-alignment"])` 启动独立 session
+- **Orchestrator 直接执行**: delphi-review code-walkthrough 需要用户确认 verdict（Issue #249），**必须由 orchestrator 直接调用** `skill(name="delphi-review")`
+- **Orchestrator 按序执行**: `delphi-review --mode code-walkthrough` → 等待 APPROVED → `test-specification-alignment` → `browse` (gstack) → 可选 `qa`/`design-review`/`benchmark` (gstack)
 - 输入: phase-2-summary.md + MVP 代码
 - 输出: 评审报告 + 测试对齐结果
-- `delphi-review --mode code-walkthrough` — 多专家匿名代码走查
-- `test-specification-alignment` — 测试与 Spec 对齐验证
-- `browse` (gstack) — 浏览器自动化测试
-- `k6` / `locust` / `gatling` — 负载/压力测试（可选，后端项目）
 
 ### 负载/压力测试（可选）
 - **适用项目**：主要用于后端服务的压力测试 (k6/Locust/Gatling)，Web 前端已有 `benchmark` 技能覆盖 Core Web Vitals、加载时间和资源大小等性能指标

package/plugins/opencode/skills/sprint-flow/references/orchestration-rules.md

@@ -31,9 +31,9 @@
 | -1 | ISOLATE | ❌ | Bash（直接执行） | 无 | orchestrator |
 | -0.5 | AUTO-ESTIMATE | ❌ | Bash（直接执行） | 无 | orchestrator |
 | 0 | THINK | ❌ | orchestrator（直接执行） | `["brainstorming"]` | orchestrator |
-| 1 | PLAN | ⚠️ | orchestrator 执行 autoplan + to-issues（交互）→ subagent 执行 delphi-review | 见下方说明 | orchestrator + subagent |
+| 1 | PLAN | ❌ | orchestrator（直接执行） | `["autoplan", "delphi-review", "to-issues"]` | orchestrator |
 | 2 | BUILD | ✅ | ralph-loop | `["test-driven-development"]` | subagent |
-| 3 | REVIEW | ✅ | `deep` | `["delphi-review", "test-specification-alignment"]` | subagent |
+| 3 | REVIEW | ❌ | orchestrator（直接执行） | `["delphi-review", "test-specification-alignment"]` | orchestrator |
 | 4 | USER ACCEPT | ❌ | **强制人工** | 无 | 用户 |
 | 5 | FEEDBACK | ✅ | `quick` | `["learn", "retro", "systematic-debugging"]` | subagent |
 | 6 | SHIP | ❌ | orchestrator（直接执行） | `["finishing-a-development-branch", "ship"]` | orchestrator |
@@ -46,15 +46,17 @@
 |-------|-------|---------------------------|
 | 0 | `brainstorming` | 需要与用户对话确认需求、提出澄清问题。Subagent 是 fire-and-forget 模式，无法暂停等待用户输入（Issue #217） |
 | 1 | `autoplan` | taste_decisions 节点暂停等待用户确认。必须由 orchestrator 直接执行（Issue #225） |
+| 1 | `delphi-review` | design 模式需等待 verdict APPROVED；非 APPROVED 时需用户确认是否接受分歧方案（Issue #249） |
 | 1 | `to-issues` | Step 6 "向用户确认" — 展示拆分结果，等待用户批准后才生成 slices-manifest.json |
+| 3 | `delphi-review --mode code-walkthrough` | Code walkthrough 非 APPROVED 时需暂停等待用户处理 Critical Issues（Issue #249） |
 | 6 | `finishing-a-development-branch` | 4 选项菜单 (merge/PR/keep/discard) 需要用户选择；Option 4 (discard) 要求 typed confirmation |
 | 6 | `ship` | PR 创建前需要用户确认；包含 AskUserQuestion STOP 点 |
 | 7 | `land-and-deploy` | Merge 确认、rollback 决策 — 均为用户交互点 |
 
-**Phase 1 执行模式（两阶段）**：
+**Phase 1 执行模式（全部 orchestrator 直接执行）**：
 1. **Orchestrator 直接执行 autoplan**：`skill(name="autoplan")` → 等待用户确认 taste_decisions
-2. **Orchestrator 直接执行 to-issues**：`skill(name="to-issues")` → 等待用户确认 Issue 拆分
-3. **Subagent 执行 delphi-review**：`task(category="deep", load_skills=["delphi-review"])` — 非交互式，可自动运行至 APPROVED
+2. **Orchestrator 直接执行 delphi-review**：`skill(name="delphi-review")` → 等待 APPROVED（非 APPROVED 时暂停等待用户确认）
+3. **Orchestrator 直接执行 to-issues**：`skill(name="to-issues")` → 等待用户确认 Issue 拆分
 
 **上下文隔离原则**：
 - 每个 Subagent 在**独立 session** 中启动，不继承 orchestrator 的对话历史
@@ -71,7 +73,7 @@
 | Phase 0 | phase--1-summary（仅路径） | 隔离环境信息（worktree 路径） |
 | Phase 1 | phase-0-summary.md + design-doc | 设计决策 + 结构化规格 |
 | Phase 2 | phase-1-summary.md + specification.yaml | 评审结论 + REQ 列表 |
-| Phase 3 | phase-2-summary.md + MVP 代码 | 构建结果 |
+| Phase 3 | phase-2-summary.md + MVP 代码 | 构建结果（orchestrator 直接执行，无 subagent 上下文继承） |
 | Phase 4 | — | **人工验收**。Phase 4 不产生 subagent summary，但用户验收结果记录在 `.sprint-state/phase-outputs/emergent-issues.md`（如有 emergent issues）。Phase 5 加载此文件。 |
 | Phase 5 | phase-4-summary.md + emergent-issues.md | 验收结论 |
 | Phase 6 | phase-5-summary.md + feedback-log.md | 复盘结论 |

package/plugins/opencode/skills/sprint-flow/references/phase-1-plan.md

@@ -105,7 +105,7 @@ Decision 2: [决策描述]
 
 ---
 
-### Step 2b: 调用 delphi-review（强制）
+### Step 2b: 调用 delphi-review（强制，orchestrator 直接执行）
 
 ```bash
 skill(name="delphi-review", user_message="[设计文档 + taste_decisions 确认结果]")
@@ -117,7 +117,7 @@ delphi-review 执行：
 - 输出: APPROVED / REQUEST_CHANGES
 
 **如果 REQUEST_CHANGES**:
-- ⚠️ 暂停等待用户修复
+- ⚠️ 暂停等待用户处理（orchestrator 直接执行，可交互确认）
 - 修复后重新评审（从 Round 2 起步）
 - 直到 APPROVED
 

package/plugins/opencode/skills/sprint-flow/references/phase-3-review.md

@@ -4,6 +4,8 @@
 
 多专家代码走查、测试对齐、浏览器测试。确保 MVP 符合 specification。
 
+**执行者**: orchestrator 直接执行全部步骤（非 subagent dispatch）— Issue #249。
+
 Web 前端项目额外增加：系统化 QA、视觉审计、性能基线。
 
 ---
@@ -46,7 +48,7 @@ delphi code-walkthrough 执行：
 - ≥90% 共识 + APPROVED 才通过
 
 **如果 REQUEST_CHANGES**:
-- ⚠️ 暂停等待用户修复 Critical Issues + 处理 Major Concerns
+- ⚠️ 暂停等待用户处理 Critical Issues + Major Concerns（orchestrator 直接执行，可交互）
 - 修复后回到 Round 2 重新评审
 
 **如果 APPROVED**:

package/plugins/opencode/skills/test-specification-alignment/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/TEST-SPECIFICATION-ALIGNMENT KNOWLEDGE BASE
 
-**Generated:** 2026-06-23
-**Commit:** bac916f
+**Generated:** 2026-06-24
+**Commit:** 5019773
 **Branch:** main
-**Version:** 0.10.11.0
+**Version:** 0.10.13.0
 
 ## OVERVIEW
 Test-Specification Alignment Engine — two-stage validation ensuring tests accurately reflect requirements and design specs.

package/plugins/opencode/tui-plugin.ts

@@ -329,11 +329,29 @@ function renderMultiSprintSidebar(sprints: DiscoveredSprint[]): string | null {
   return blocks.join("\n---\n");
 }
 
-function renderContent(sprints: DiscoveredSprint[], upgradeNotice: string | null): string | null {
+function renderContent(sprints: DiscoveredSprint[], upgradeNotice: string | null, dir: string): string | null {
   const sprintContent = renderMultiSprintSidebar(sprints)
-  if (upgradeNotice && sprintContent) return `${upgradeNotice}\n---\n${sprintContent}`
-  if (upgradeNotice) return upgradeNotice
-  return sprintContent
+
+  if (sprintContent) {
+    return [upgradeNotice, sprintContent].filter(Boolean).join("\n---\n")
+  }
+
+  // Early-phase placeholders: when no sprint data yet, check for directory hints
+  const hasStateDir = existsSync(join(dir, ".sprint-state"))
+  const gitRoot = findGitRoot(dir)
+  const hasWorktreesRoot = gitRoot ? existsSync(join(gitRoot, ".worktrees")) : false
+
+  if (hasStateDir) {
+    const placeholder = "SPRINT FLOW\n  → 初始化中..."
+    return [upgradeNotice, placeholder].filter(Boolean).join("\n---\n")
+  }
+
+  if (hasWorktreesRoot) {
+    const placeholder = "SPRINT FLOW\n  · 准备中..."
+    return [upgradeNotice, placeholder].filter(Boolean).join("\n---\n")
+  }
+
+  return upgradeNotice || null
 }
 
 // ── Upgrade notice ──
@@ -376,15 +394,14 @@ const tuiPlugin: TuiSlotPlugin = {
       const dir = process.env.XP_GATE_PROJECT_DIR || process.cwd();
       const now = Date.now();
 
-      // Use cache if still valid for current directory
       if (_cache && _cache.dir === dir && now - _cache.ts < CACHE_TTL_MS) {
-        return renderContent(_cache.data, _cache.upgradeNotice);
+        return renderContent(_cache.data, _cache.upgradeNotice, dir);
       }
 
       const sprints = discoverActiveSprints(dir);
       const upgradeNotice = renderUpgradeNotice()
       _cache = { data: sprints, ts: now, dir, upgradeNotice };
-      return renderContent(sprints, upgradeNotice);
+      return renderContent(sprints, upgradeNotice, dir);
     },
   },
 }

package/plugins/qoder/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "xp-gate",
-  "version": "0.10.11",
+  "version": "0.10.13",
   "displayName": "XP-Gate",
   "description": "Extreme Programming quality gates + AI workflow skills for Qoder. Includes 10 quality gates (Gate 0-9), Sprint Flow (11 phases), and Delphi multi-expert review (>=90% consensus).",
   "author": {

package/plugins/qoder/skills/delphi-review/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/DELPHI-REVIEW KNOWLEDGE BASE
 
-**Generated:** 2026-06-23
-**Commit:** bac916f
+**Generated:** 2026-06-24
+**Commit:** 5019773
 **Branch:** main
-**Version:** 0.10.11.0
+**Version:** 0.10.13.0
 
 ## OVERVIEW
 Delphi Consensus Review — multi-round anonymous expert review (≥90% threshold, 3 experts from ≥2 providers, domestic models only). Supports design + code-walkthrough modes.

package/plugins/qoder/skills/sprint-flow/AGENTS.md

@@ -1,9 +1,9 @@
 # SKILLS/SPRINT-FLOW KNOWLEDGE BASE
 
-**Generated:** 2026-06-23
-**Commit:** bac916f
+**Generated:** 2026-06-24
+**Commit:** 5019773
 **Branch:** main
-**Version:** 0.10.11.0
+**Version:** 0.10.13.0
 
 ## OVERVIEW
 **11-phase** development pipeline: ISOLATE → AUTO-ESTIMATE → THINK → PLAN → BUILD → REVIEW → USER ACCEPTANCE → FEEDBACK → SHIP → LAND → CLEANUP. Phase 2 default build mode is **ralph-loop** (REQ-level iteration, 40-67% token savings vs parallel). HARD-GATE in Phase 1: design must pass Delphi review (≥90% consensus) before any coding.