fix(workflow): resolve analyze-with-file and lite-plan phase conflict

Add workflow boundary markers to analyze-with-file Post-Completion Options so that when "生成任务" triggers lite-plan via Skill(), the model follows lite-plan's Phase 1-5 exclusively instead of returning to analyze Phase 4 synthesis. Root cause: Skill() expands inline (shared context), causing phase number collision between the two workflows.
7.2.0
2026-03-10 17:11:04 +08:00 · 2026-03-02 13:08:59 +08:00 · 2026-03-02 12:28:39 +08:00 · 2026-03-02 12:27:42 +08:00 · 2026-03-02 12:27:31 +08:00 · 2026-03-02 11:26:15 +08:00
19 changed files with 299 additions and 130 deletions
--- a/.claude/commands/ccw-coordinator.md
+++ b/.claude/commands/ccw-coordinator.md
@@ -30,7 +30,7 @@ Interactive orchestration tool: analyze task → discover commands → recommend
 | `workflow:roadmap-with-file` | strategic requirement roadmap → issue creation → execution-plan.json |
 | `workflow:integration-test-cycle` | explore → test dev → test-fix cycle → reflection |
 | `workflow:refactor-cycle` | tech debt discovery → prioritize → execute → validate |
-| `team-planex` | planner + executor wave pipeline（适合大量零散 issue 或 roadmap 产出的清晰 issue，实现 0→1 开发）|
+| `team-planex` | planner + executor wave pipeline（适合大量零散 issue 或 roadmap 产出的清晰 issue）|

 ## Core Concept: Minimum Execution Units (最小执行单元)

@@ -85,7 +85,7 @@ Interactive orchestration tool: analyze task → discover commands → recommend
 | Unit Name | Commands | Purpose | Output |
 |-----------|----------|---------|--------|
 | **Analyze to Plan** | analyze-with-file → lite-plan | Collaborative analysis → auto chain to lite-plan | discussion.md + code |
-| **Brainstorm to Plan** | brainstorm-with-file → lite-plan | Multi-perspective ideation → auto chain to lite-plan | brainstorm.md + code |
+| **Brainstorm to Plan** | brainstorm-with-file → plan → execute | Multi-perspective ideation → formal planning | brainstorm.md + code |
 | **Debug With File** | debug-with-file | Hypothesis-driven debugging with documentation | understanding.md |
 | **Collaborative Plan** | collaborative-plan-with-file → unified-execute-with-file | Multi-agent collaborative planning and execution | plan-note.md + code |
 | **Roadmap Plan** | roadmap-with-file → team-planex | Requirement decomposition and wave execution | execution-plan.json + code |
@@ -118,7 +118,7 @@ Interactive orchestration tool: analyze task → discover commands → recommend
 | issue:queue | issue:execute | Issue Workflow, Rapid-to-Issue, Brainstorm-to-Issue |
 | issue:from-brainstorm | issue:queue | Brainstorm-to-Issue |
 | analyze-with-file | lite-plan (auto) | Analyze to Plan |
-| brainstorm-with-file | lite-plan (auto), issue:from-brainstorm | Brainstorm to Plan, Brainstorm-to-Issue |
+| brainstorm-with-file | plan (auto), issue:from-brainstorm | Brainstorm to Plan, Brainstorm-to-Issue |
 | collaborative-plan-with-file | unified-execute-with-file | Collaborative Plan |
 | roadmap-with-file | team-planex | Roadmap Plan |
 | unified-execute-with-file | (terminal) | Collaborative Plan |
@@ -165,11 +165,13 @@ function detectTaskType(text) {
  if (/urgent|production|critical/.test(text) && /fix|bug/.test(text)) return 'bugfix-hotfix';
  // With-File workflow patterns (specific keywords - must come before generic bugfix)
  if (/brainstorm.*issue|头脑风暴.*issue|idea.*issue|想法.*issue|从.*头脑风暴|convert.*brainstorm/.test(text)) return 'brainstorm-to-issue';
+  // 0→1 Greenfield detection (priority over brainstorm/roadmap)
+  if (/从零开始|from scratch|0.*to.*1|greenfield|全新.*开发|新项目|new project|build.*from.*ground/.test(text)) return 'greenfield';
  if (/brainstorm|ideation|头脑风暴|创意|发散思维|creative thinking/.test(text)) return 'brainstorm-file';
  if (/debug.*document|hypothesis.*debug|深度调试|假设.*验证|systematic debug/.test(text)) return 'debug-file';
  if (/analyze.*document|collaborative analysis|协作分析|深度.*理解/.test(text)) return 'analyze-file';
  if (/collaborative.*plan|协作.*规划|多人.*规划|multi.*agent.*plan|Plan Note|分工.*规划/.test(text)) return 'collaborative-plan';
-  if (/roadmap|需求.*规划|需求.*拆解|requirement.*plan|progressive.*plan|路线.*图/.test(text)) return 'roadmap';
+  if (/roadmap|路线.*图/.test(text)) return 'roadmap';  // Narrowed: only explicit roadmap keywords
  if (/spec.*gen|specification|PRD|产品需求|产品文档|产品规格/.test(text)) return 'spec-driven';
  // Cycle workflow patterns
  if (/integration.*test|集成测试|端到端.*测试|e2e.*test|integration.*cycle/.test(text)) return 'integration-test';
@@ -407,7 +409,7 @@ const commandPorts = {
    output: ['brainstorm-document'],
    tags: ['brainstorm', 'with-file'],
    atomic_group: 'brainstorm-to-plan',
-    note: 'Auto chains to lite-plan with brainstorm artifacts'
+    note: 'Auto chains to workflow-plan with brainstorm artifacts'
  },
  'issue:from-brainstorm': {
    name: 'issue:from-brainstorm',
@@ -515,8 +517,10 @@ function determinePortFlow(taskType, constraints) {
    // Issue workflow types
    'issue-batch':        { inputPort: 'codebase', outputPort: 'completed-issues' },
    'issue-transition':   { inputPort: 'requirement', outputPort: 'completed-issues' },
-    // With-File workflow types (auto chain to lite-plan)
-    'brainstorm-file':      { inputPort: 'exploration-topic', outputPort: 'code' },
+    // 0→1 Greenfield (exploration → formal planning → execution)
+    'greenfield':             { inputPort: 'exploration-topic', outputPort: 'test-passed' },
+    // With-File workflow types (auto chain to plan)
+    'brainstorm-file':      { inputPort: 'exploration-topic', outputPort: 'test-passed' },
    'brainstorm-to-issue':  { inputPort: 'brainstorm-document', outputPort: 'completed-issues' },
    'debug-file':           { inputPort: 'bug-report', outputPort: 'understanding-document' },
    'analyze-file':         { inputPort: 'analysis-topic', outputPort: 'code' },
@@ -957,7 +961,8 @@ Task: <description>
 | **issue-batch** | 代码库 →【discover → plan → queue → execute】→ 完成 issues | Issue Workflow |
 | **issue-transition** | 需求 →【lite-plan → convert-to-plan → queue → execute】→ 完成 issues | Rapid-to-Issue |
 | **analyze-file** | 分析主题 →【analyze-with-file → lite-plan → lite-execute】→ 代码 | Analyze to Plan |
-| **brainstorm-file** | 主题 →【brainstorm-with-file → lite-plan → lite-execute】→ 代码 | Brainstorm to Plan |
+| **greenfield** | 需求 →【brainstorm-with-file → plan → execute】→ 代码 → test | Greenfield (0→1) |
+| **brainstorm-file** | 主题 →【brainstorm-with-file → plan → execute】→ 代码 → test | Brainstorm to Plan |
 | **brainstorm-to-issue** | brainstorm.md →【from-brainstorm → queue → execute】→ 完成 issues | Brainstorm to Issue |
 | **debug-file** | Bug报告 → debug-with-file → understanding.md (自包含) | Debug With File |
 | **collaborative-plan** | 需求 →【collaborative-plan-with-file → unified-execute-with-file】→ 代码 | Collaborative Plan |
--- a/.claude/commands/ccw.md
+++ b/.claude/commands/ccw.md
@@ -28,7 +28,7 @@ Main process orchestrator: intent analysis → workflow selection → command ch
 | `workflow:roadmap-with-file` | strategic requirement roadmap → issue creation → execution-plan.json |
 | `workflow:integration-test-cycle` | explore → test dev → test-fix cycle → reflection |
 | `workflow:refactor-cycle` | tech debt discovery → prioritize → execute → validate |
-| `team-planex` | planner + executor wave pipeline（适合大量零散 issue 或 roadmap 产出的清晰 issue，实现 0→1 开发）|
+| `team-planex` | planner + executor wave pipeline（适合大量零散 issue 或 roadmap 产出的清晰 issue）|

 ## Core Concept: Self-Contained Skills (自包含 Skill)

@@ -51,9 +51,11 @@ Main process orchestrator: intent analysis → workflow selection → command ch
 | 代码审查 | `review-cycle` | 内部完成 review→fix |
 | 多CLI协作 | `workflow-multi-cli-plan` | ACE context → CLI discussion → plan → execute |
 | 分析→规划 | `workflow:analyze-with-file` → `workflow-lite-plan` | 协作分析产物自动传递给 lite-plan |
-| 头脑风暴→规划 | `workflow:brainstorm-with-file` → `workflow-lite-plan` | 头脑风暴产物自动传递给 lite-plan |
+| 头脑风暴→规划 | `workflow:brainstorm-with-file` → `workflow-plan` → `workflow-execute` | 头脑风暴产物自动传递给正式规划 |
+| 0→1 开发(小) | `workflow:brainstorm-with-file` → `workflow-plan` → `workflow-execute` | 小规模从零开始，探索+正式规划+实现 |
+| 0→1 开发(中/大) | `workflow:brainstorm-with-file` → `workflow-plan` → `workflow-execute` | 探索后正式规划+执行 |
 | 协作规划 | `workflow:collaborative-plan-with-file` → `workflow:unified-execute-with-file` | 多 agent 协作规划→通用执行 |
-| 需求路线图 | `workflow:roadmap-with-file` → `team-planex` | 需求拆解→issue 创建→wave pipeline 执行 |
+| 需求路线图 | `workflow:roadmap-with-file` → `team-planex` | 需求拆解→issue 创建→wave pipeline 执行（需明确 roadmap 关键词）|
 | 集成测试循环 | `workflow:integration-test-cycle` | 自迭代集成测试闭环 |
 | 重构循环 | `workflow:refactor-cycle` | 技术债务发现→重构→验证 |

@@ -124,12 +126,14 @@ function detectTaskType(text) {
  const patterns = {
    'bugfix-hotfix': /urgent|production|critical/ && /fix|bug/,
    // With-File workflows (documented exploration → auto chain to lite-plan)
+    // 0→1 Greenfield detection (priority over brainstorm/roadmap)
+    'greenfield': /从零开始|from scratch|0.*to.*1|greenfield|全新.*开发|新项目|new project|build.*from.*ground/,
    'brainstorm': /brainstorm|ideation|头脑风暴|创意|发散思维|creative thinking|multi-perspective.*think|compare perspectives|探索.*可能/,
    'brainstorm-to-issue': /brainstorm.*issue|头脑风暴.*issue|idea.*issue|想法.*issue|从.*头脑风暴|convert.*brainstorm/,
    'debug-file': /debug.*document|hypothesis.*debug|troubleshoot.*track|investigate.*log|调试.*记录|假设.*验证|systematic debug|深度调试/,
    'analyze-file': /analyze.*document|explore.*concept|understand.*architecture|investigate.*discuss|collaborative analysis|分析.*讨论|深度.*理解|协作.*分析/,
    'collaborative-plan': /collaborative.*plan|协作.*规划|多人.*规划|multi.*agent.*plan|Plan Note|分工.*规划/,
-    'roadmap': /roadmap|需求.*规划|需求.*拆解|requirement.*plan|progressive.*plan|路线.*图/,
+    'roadmap': /roadmap|路线.*图/,  // Narrowed: only explicit roadmap keywords (需求规划/需求拆解 moved to greenfield routing)
    'spec-driven': /spec.*gen|specification|PRD|产品需求|产品文档|产品规格/,
    // Cycle workflows (self-iterating with reflection)
    'integration-test': /integration.*test|集成测试|端到端.*测试|e2e.*test|integration.*cycle/,
@@ -184,13 +188,18 @@ async function clarifyRequirements(analysis) {
 function selectWorkflow(analysis) {
  const levelMap = {
    'bugfix-hotfix':     { level: 2, flow: 'bugfix.hotfix' },
+    // 0→1 Greenfield (complexity-adaptive routing)
+    'greenfield':        { level: analysis.complexity === 'high' ? 4 : 3,
+                           flow: analysis.complexity === 'high' ? 'greenfield-phased'    // large: brainstorm → workflow-plan → execute
+                                : analysis.complexity === 'medium' ? 'greenfield-plan'   // medium: brainstorm → workflow-plan → execute
+                                : 'brainstorm-to-plan' },                                // small: brainstorm → workflow-plan
    // With-File workflows → auto chain to lite-plan
-    'brainstorm':        { level: 4, flow: 'brainstorm-to-plan' },     // brainstorm-with-file → lite-plan
+    'brainstorm':        { level: 4, flow: 'brainstorm-to-plan' },     // brainstorm-with-file → workflow-plan
    'brainstorm-to-issue': { level: 4, flow: 'brainstorm-to-issue' },  // Brainstorm → Issue workflow
    'debug-file':        { level: 3, flow: 'debug-with-file' },         // Hypothesis-driven debugging (standalone)
    'analyze-file':      { level: 3, flow: 'analyze-to-plan' },         // analyze-with-file → lite-plan
    'collaborative-plan': { level: 3, flow: 'collaborative-plan' },     // Multi-agent collaborative planning
-    'roadmap':           { level: 4, flow: 'roadmap' },                 // roadmap → team-planex
+    'roadmap':           { level: 4, flow: 'roadmap' },                 // roadmap → team-planex (explicit roadmap only)
    'spec-driven':       { level: 4, flow: 'spec-driven' },             // spec-generator → plan → execute
    // Cycle workflows (self-iterating with reflection)
    'integration-test':  { level: 3, flow: 'integration-test-cycle' },
@@ -266,7 +275,11 @@ function buildCommandChain(workflow, analysis) {

    'brainstorm-to-plan': [
      { cmd: 'workflow:brainstorm-with-file', args: `"${analysis.goal}"` },
-      { cmd: 'workflow-lite-plan', args: '' }  // auto receives brainstorm artifacts (brainstorm.md)
+      { cmd: 'workflow-plan', args: '' },         // formal planning with brainstorm artifacts
+      { cmd: 'workflow-execute', args: '' },
+      ...(analysis.constraints?.includes('skip-tests') ? [] : [
+        { cmd: 'workflow-test-fix', args: '' }
+      ])
    ],

    'debug-with-file': [
@@ -281,6 +294,26 @@ function buildCommandChain(workflow, analysis) {
      { cmd: 'issue:execute', args: '--queue auto' }
    ],

+    // 0→1 Greenfield (complexity-adaptive)
+    'greenfield-plan': [
+      { cmd: 'workflow:brainstorm-with-file', args: `"${analysis.goal}"` },
+      { cmd: 'workflow-plan', args: '' },         // formal planning after exploration
+      { cmd: 'workflow-execute', args: '' },
+      ...(analysis.constraints?.includes('skip-tests') ? [] : [
+        { cmd: 'workflow-test-fix', args: '' }
+      ])
+    ],
+
+    'greenfield-phased': [
+      { cmd: 'workflow:brainstorm-with-file', args: `"${analysis.goal}"` },
+      { cmd: 'workflow-plan', args: '' },         // formal planning after exploration
+      { cmd: 'workflow-execute', args: '' },
+      { cmd: 'review-cycle', args: '' },
+      ...(analysis.constraints?.includes('skip-tests') ? [] : [
+        { cmd: 'workflow-test-fix', args: '' }
+      ])
+    ],
+
    // Universal Plan+Execute
    'collaborative-plan': [
      { cmd: 'workflow:collaborative-plan-with-file', args: `"${analysis.goal}"` },
@@ -338,7 +371,7 @@ function buildCommandChain(workflow, analysis) {
      { cmd: 'workflow-execute', args: '' }
    ],

-    // Level 4 - Full Exploration
+    // Level 4 - Full Exploration (brainstorm → formal planning → execute)
    'full': [
      { cmd: 'brainstorm', args: `"${analysis.goal}"` },
      { cmd: 'workflow-plan', args: '' },
@@ -600,10 +633,12 @@ Phase 5: Execute Command Chain
 | "Use issue workflow" | issue-transition | 2.5 | workflow-lite-plan(plan-only) → convert-to-plan → queue → execute |
 | "协作分析: 认证架构" | analyze-file | 3 | analyze-with-file → workflow-lite-plan |
 | "深度调试 WebSocket" | debug-file | 3 | workflow:debug-with-file |
-| "头脑风暴: 通知系统" | brainstorm | 4 | brainstorm-with-file → workflow-lite-plan |
+| "从零开始: 用户系统" | greenfield (medium) | 3 | brainstorm-with-file → workflow-plan → workflow-execute → workflow-test-fix |
+| "greenfield: 大型平台" | greenfield (high) | 4 | brainstorm-with-file → workflow-plan → workflow-execute → review-cycle → workflow-test-fix |
+| "头脑风暴: 通知系统" | brainstorm | 4 | brainstorm-with-file → workflow-plan → workflow-execute → workflow-test-fix |
 | "从头脑风暴创建 issue" | brainstorm-to-issue | 4 | issue:from-brainstorm → issue:queue → issue:execute |
 | "协作规划: 实时通知系统" | collaborative-plan | 3 | collaborative-plan-with-file → unified-execute-with-file |
-| "需求路线图: OAuth + 2FA" | roadmap | 4 | roadmap-with-file → team-planex |
+| "roadmap: OAuth + 2FA" | roadmap | 4 | roadmap-with-file → team-planex |
 | "specification: 用户系统" | spec-driven | 4 | spec-generator → workflow-plan → workflow-execute → workflow-test-fix |
 | "集成测试: 支付流程" | integration-test | 3 | workflow:integration-test-cycle |
 | "重构 auth 模块" | refactor | 3 | workflow:refactor-cycle |
@@ -687,13 +722,13 @@ todos = [

 | Workflow | Purpose | Auto Chain | Output Folder |
 |----------|---------|------------|---------------|
-| **brainstorm-with-file** | Multi-perspective ideation | → workflow-lite-plan (auto) | `.workflow/.brainstorm/` |
+| **brainstorm-with-file** | Multi-perspective ideation | → workflow-plan → workflow-execute (auto) | `.workflow/.brainstorm/` |
 | **debug-with-file** | Hypothesis-driven debugging | Standalone (self-contained) | `.workflow/.debug/` |
 | **analyze-with-file** | Collaborative analysis | → workflow-lite-plan (auto) | `.workflow/.analysis/` |
 | **collaborative-plan-with-file** | Multi-agent collaborative planning | → unified-execute-with-file | `.workflow/.planning/` |
 | **roadmap-with-file** | Strategic requirement roadmap | → team-planex | `.workflow/.planning/` |

-**Auto Chain Mechanism**: When `analyze-with-file` or `brainstorm-with-file` completes, its artifacts (discussion.md / brainstorm.md) are automatically passed to `workflow-lite-plan` as context input. No user intervention needed.
+**Auto Chain Mechanism**: When `analyze-with-file` completes, its artifacts (discussion.md) are automatically passed to `workflow-lite-plan`. When `brainstorm-with-file` completes, its artifacts (brainstorm.md) are passed to `workflow-plan` for formal planning. No user intervention needed.

 **Detection Keywords**:
 - **brainstorm**: 头脑风暴, 创意, 发散思维, multi-perspective, compare perspectives
@@ -752,9 +787,14 @@ todos = [
 # Multi-CLI collaborative planning
 /ccw "multi-cli plan: 支付网关API设计"               # → workflow-multi-cli-plan → workflow-test-fix

-# With-File workflows → auto chain to lite-plan
+# 0→1 Greenfield development (exploration-first)
+/ccw "从零开始: 用户认证系统"                   # → brainstorm-with-file → workflow-plan → workflow-execute → workflow-test-fix
+/ccw "new project: 数据导出模块"               # → brainstorm-with-file → workflow-plan → workflow-execute → workflow-test-fix
+/ccw "全新开发: 实时通知系统"                   # → brainstorm-with-file → workflow-plan → workflow-execute → review-cycle → workflow-test-fix
+
+# With-File workflows → auto chain
 /ccw "协作分析: 理解现有认证架构的设计决策"     # → analyze-with-file → workflow-lite-plan
-/ccw "头脑风暴: 用户通知系统重新设计"           # → brainstorm-with-file → workflow-lite-plan
+/ccw "头脑风暴: 用户通知系统重新设计"           # → brainstorm-with-file → workflow-plan → workflow-execute → workflow-test-fix
 /ccw "深度调试: 系统随机崩溃问题"              # → debug-with-file (standalone)
 /ccw "从头脑风暴 BS-通知系统-2025-01-28 创建 issue"  # → brainstorm-to-issue (bridge)

@@ -763,8 +803,8 @@ todos = [

 # Collaborative planning & requirement workflows
 /ccw "协作规划: 实时通知系统架构"              # → collaborative-plan-with-file → unified-execute
-/ccw "需求路线图: 用户认证 OAuth + 2FA"        # → roadmap-with-file → team-planex
-/ccw "roadmap: 数据导出功能路线图"             # → roadmap-with-file → team-planex
+/ccw "roadmap: 用户认证 OAuth + 2FA 路线图"    # → roadmap-with-file → team-planex (explicit roadmap only)
+/ccw "roadmap: 数据导出功能路线图"             # → roadmap-with-file → team-planex (explicit roadmap only)

 # Team workflows (kept: team-planex)
 /ccw "team planex: 用户认证系统"               # → team-planex (planner + executor wave pipeline)
--- a/.claude/commands/workflow/analyze-with-file.md
+++ b/.claude/commands/workflow/analyze-with-file.md
@@ -542,7 +542,11 @@ CONSTRAINTS: ${perspective.constraints}
     - **Trade-offs Made**: Key trade-offs and why certain paths were chosen over others
   - Add session statistics: rounds, duration, sources, artifacts, **decision count**

-3. **Post-Completion Options**
+3. **Post-Completion Options** (⚠️ TERMINAL — analyze-with-file ends after user selection)
+
+   > **WORKFLOW BOUNDARY**: After user selects any option below, the analyze-with-file workflow is **COMPLETE**.
+   > If "生成任务" is selected, workflow-lite-plan takes over exclusively — do NOT return to any analyze-with-file phase.
+   > The "Phase" numbers in workflow-lite-plan (Phase 1-5) are SEPARATE from analyze-with-file phases.

   ```javascript
   const hasActionableRecs = conclusions.recommendations?.some(r => r.priority === 'high' || r.priority === 'medium')
@@ -562,7 +566,7 @@ CONSTRAINTS: ${perspective.constraints}
   })
   ```

-   **Handle "生成任务"**:
+   **Handle "生成任务"** (⚠️ TERMINAL — analyze-with-file ends here, lite-plan takes over exclusively):
   ```javascript
   if (nextStep.includes("生成任务")) {
     // 1. Build task description from high/medium priority recommendations
@@ -585,8 +589,9 @@ CONSTRAINTS: ${perspective.constraints}
       if (findings.length) contextLines.push(`**Key Findings**:\n${findings.map(f => `- ${f}`).join('\n')}`)
     }

-     // 3. Call lite-plan with enriched task description (no special flags)
+     // 3. Hand off to lite-plan — analyze-with-file COMPLETE, do NOT return to any analyze phase
     Skill(skill="workflow-lite-plan", args=`"${taskDescription}\n\n${contextLines.join('\n')}"`)
+     return  // ⛔ analyze-with-file terminates here
   }
   ```

--- a/.claude/skills/workflow-lite-plan/phases/01-lite-plan.md
+++ b/.claude/skills/workflow-lite-plan/phases/01-lite-plan.md
@@ -529,15 +529,17 @@ ${task_description}

 ## Multi-Angle Exploration Context

-${manifest.explorations.map(exp => `### Exploration: ${exp.angle} (${exp.file})
+${manifest.explorations.length > 0
+  ? manifest.explorations.map(exp => `### Exploration: ${exp.angle} (${exp.file})
 Path: ${exp.path}

-Read this file for detailed ${exp.angle} analysis.`).join('\n\n')}
+Read this file for detailed ${exp.angle} analysis.`).join('\n\n') + `

 Total explorations: ${manifest.exploration_count}
 Angles covered: ${manifest.explorations.map(e => e.angle).join(', ')}

-Manifest: ${sessionFolder}/explorations-manifest.json
+Manifest: ${sessionFolder}/explorations-manifest.json`
+  : `No exploration files. Task Description above contains "## Prior Analysis" with analysis summary, key files, and findings — use it as primary planning context.`}

 ## User Clarifications
 ${JSON.stringify(clarificationContext) || "None"}
--- a/.codex/skills/roadmap-with-file/SKILL.md
+++ b/.codex/skills/roadmap-with-file/SKILL.md
@@ -687,7 +687,8 @@ ${selectedMode === 'progressive' ? `**Progressive Mode**:
     if (feedback.feedback === 'approve') {
       approved = true
     } else {
-       // Handle feedback type
+       // CONSTRAINT: All modifications below ONLY touch roadmap.md and issues.jsonl
+       // NEVER modify source code or project files during interactive rounds
       switch (feedback.feedback) {
         case 'scope':
           // Collect scope adjustments
@@ -695,8 +696,9 @@ ${selectedMode === 'progressive' ? `**Progressive Mode**:
             { id: "adjustments", type: "text", prompt: "Describe scope adjustments needed:" }
           ])  // BLOCKS

-           // Update roadmap.md and issues
-           // ... implementation ...
+           // Update ONLY roadmap.md Roadmap table + Convergence sections
+           Edit({ path: `${sessionFolder}/roadmap.md`, /* scope changes */ })
+           // Update ONLY issues.jsonl entries (scope/context fields)

           break

@@ -706,8 +708,8 @@ ${selectedMode === 'progressive' ? `**Progressive Mode**:
             { id: "refinements", type: "text", prompt: "Describe convergence refinements needed:" }
           ])  // BLOCKS

-           // Update roadmap.md
-           // ... implementation ...
+           // Update ONLY roadmap.md Convergence Criteria section
+           Edit({ path: `${sessionFolder}/roadmap.md`, /* convergence changes */ })

           break

@@ -726,8 +728,7 @@ ${selectedMode === 'progressive' ? `**Progressive Mode**:
           ])  // BLOCKS

           selectedMode = newStrategy.strategy
-           // Re-execute Phase 2
-           // ... goto Phase 2 ...
+           // Re-execute Phase 2 (updates roadmap.md + issues.jsonl only)
           break
       }

@@ -871,6 +872,7 @@ To execute later:
 4. **Testable Convergence**: criteria = assertions, DoD = business language
 5. **Explicit Lifecycle**: Always close_agent after wait completes to free resources
 6. **DO NOT STOP**: Continuous workflow until handoff complete
+7. **Plan-Only Modifications**: Interactive feedback (Phase 3) MUST only update `roadmap.md` and `issues.jsonl`. NEVER modify source code, configuration files, or any project files during interactive rounds. Code changes happen only after handoff (Phase 4) via team-planex or other execution skills

 ---

--- a/ccw/frontend/src/components/mcp/CcwToolsMcpCard.test.tsx
+++ b/ccw/frontend/src/components/mcp/CcwToolsMcpCard.test.tsx
@@ -41,6 +41,43 @@ describe('CcwToolsMcpCard', () => {
    vi.clearAllMocks();
  });

+  it('disables all tools when clicking "Disable All" button', async () => {
+    const { CcwToolsMcpCard } = await import('./CcwToolsMcpCard');
+    const onUpdateConfigMock = vi.fn();
+
+    render(
+      <CcwToolsMcpCard
+        target="codex"
+        isInstalled={true}
+        enabledTools={['write_file', 'read_file', 'edit_file']}
+        onToggleTool={vi.fn()}
+        onUpdateConfig={onUpdateConfigMock}
+        onInstall={vi.fn()}
+      />,
+      { locale: 'en' }
+    );
+
+    const user = userEvent.setup();
+    // Expand the card
+    await act(async () => {
+      await user.click(screen.getByText(/CCW MCP Server|mcp\.ccw\.title/i));
+    });
+
+    // Find and click "Disable All" button
+    const disableAllButton = screen.getByRole('button', {
+      name: /Disable All|mcp\.ccw\.actions\.disableAll/i,
+    });
+    expect(disableAllButton).toBeEnabled();
+    await act(async () => {
+      await user.click(disableAllButton);
+    });
+
+    // Verify onUpdateConfig was called with empty enabledTools array
+    await waitFor(() => {
+      expect(onUpdateConfigMock).toHaveBeenCalledWith({ enabledTools: [] });
+    });
+  });
+
  it('preserves enabledTools when saving config (Codex)', async () => {
    const { CcwToolsMcpCard } = await import('./CcwToolsMcpCard');
    const updateCodexMock = vi.mocked(apiMock.updateCcwConfigForCodex);
--- a/ccw/frontend/src/components/ui/Collapsible.tsx
+++ b/ccw/frontend/src/components/ui/Collapsible.tsx
@@ -13,7 +13,7 @@ const CollapsibleContent = React.forwardRef<
  <CollapsiblePrimitive.Content
    ref={ref}
    className={cn(
-      "overflow-hidden transition-all data-[state=closed]:animate-accordion-up data-[state=open]:animate-accordion-down",
+      "transition-all data-[state=closed]:overflow-hidden data-[state=closed]:animate-accordion-up data-[state=open]:animate-accordion-down",
      className
    )}
    {...props}
--- a/ccw/frontend/src/lib/api.ts
+++ b/ccw/frontend/src/lib/api.ts
@@ -4371,10 +4371,15 @@ function buildCcwMcpServerConfig(config: {
 }): { command: string; args: string[]; env: Record<string, string> } {
  const env: Record<string, string> = {};

-  if (config.enabledTools && config.enabledTools.length > 0) {
+  // Only use default when enabledTools is undefined (not provided)
+  // When enabledTools is an empty array, set to empty string to disable all tools
+  console.log('[buildCcwMcpServerConfig] config.enabledTools:', config.enabledTools);
+  if (config.enabledTools !== undefined) {
    env.CCW_ENABLED_TOOLS = config.enabledTools.join(',');
+    console.log('[buildCcwMcpServerConfig] Set CCW_ENABLED_TOOLS to:', env.CCW_ENABLED_TOOLS);
  } else {
    env.CCW_ENABLED_TOOLS = 'write_file,edit_file,read_file,core_memory,ask_question,smart_search';
+    console.log('[buildCcwMcpServerConfig] Using default CCW_ENABLED_TOOLS');
  }

  if (config.projectRoot) {
@@ -4455,11 +4460,20 @@ export async function fetchCcwMcpConfig(currentProjectPath?: string): Promise<Cc
    }

    // Parse enabled tools from env
+    // Note: CCW_ENABLED_TOOLS can be empty string (all tools disabled), 'all' (default set), or comma-separated list
    const env = ccwServer.env || {};
-    const enabledToolsStr = env.CCW_ENABLED_TOOLS || 'all';
-    const enabledTools = enabledToolsStr === 'all'
-      ? ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question']
-      : enabledToolsStr.split(',').map((t: string) => t.trim());
+    const enabledToolsStr = env.CCW_ENABLED_TOOLS;
+    let enabledTools: string[];
+    if (enabledToolsStr === undefined || enabledToolsStr === null) {
+      // No setting = use default tools
+      enabledTools = ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question', 'smart_search'];
+    } else if (enabledToolsStr === '' || enabledToolsStr === 'all') {
+      // Empty string = all tools disabled, 'all' = default set (for backward compatibility)
+      enabledTools = enabledToolsStr === '' ? [] : ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question', 'smart_search'];
+    } else {
+      // Comma-separated list
+      enabledTools = enabledToolsStr.split(',').map((t: string) => t.trim()).filter(Boolean);
+    }

    return {
      isInstalled: true,
@@ -4601,10 +4615,19 @@ export async function fetchCcwMcpConfigForCodex(): Promise<CcwMcpConfig> {
    }

    const env = ccwServer.env || {};
-    const enabledToolsStr = env.CCW_ENABLED_TOOLS || 'all';
-    const enabledTools = enabledToolsStr === 'all'
-      ? ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question', 'smart_search']
-      : enabledToolsStr.split(',').map((t: string) => t.trim());
+    // Note: CCW_ENABLED_TOOLS can be empty string (all tools disabled), 'all' (default set), or comma-separated list
+    const enabledToolsStr = env.CCW_ENABLED_TOOLS;
+    let enabledTools: string[];
+    if (enabledToolsStr === undefined || enabledToolsStr === null) {
+      // No setting = use default tools
+      enabledTools = ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question', 'smart_search'];
+    } else if (enabledToolsStr === '' || enabledToolsStr === 'all') {
+      // Empty string = all tools disabled, 'all' = default set (for backward compatibility)
+      enabledTools = enabledToolsStr === '' ? [] : ['write_file', 'edit_file', 'read_file', 'core_memory', 'ask_question', 'smart_search'];
+    } else {
+      // Comma-separated list
+      enabledTools = enabledToolsStr.split(',').map((t: string) => t.trim()).filter(Boolean);
+    }

    return {
      isInstalled: true,
@@ -4630,7 +4653,9 @@ function buildCcwMcpServerConfigForCodex(config: {
 }): { command: string; args: string[]; env: Record<string, string> } {
  const env: Record<string, string> = {};

-  if (config.enabledTools && config.enabledTools.length > 0) {
+  // Only use default when enabledTools is undefined (not provided)
+  // When enabledTools is an empty array, set to empty string to disable all tools
+  if (config.enabledTools !== undefined) {
    env.CCW_ENABLED_TOOLS = config.enabledTools.join(',');
  } else {
    env.CCW_ENABLED_TOOLS = 'write_file,edit_file,read_file,core_memory,ask_question,smart_search';
--- a/ccw/frontend/src/pages/McpManagerPage.tsx
+++ b/ccw/frontend/src/pages/McpManagerPage.tsx
@@ -412,8 +412,7 @@ export function McpManagerPage() {
  };

  const handleUpdateCcwConfig = async (config: Partial<CcwMcpConfig>) => {
-    // Read BEFORE optimistic update to capture actual server state
-    const currentConfig = queryClient.getQueryData<CcwMcpConfig>(ccwMcpQueryKey) ?? ccwConfig;
+    // Read BEFORE optimistic update to capture previous state for rollback
    const previousConfig = queryClient.getQueryData<CcwMcpConfig>(ccwMcpQueryKey);

    // Optimistic cache update for immediate UI response
@@ -422,8 +421,17 @@ export function McpManagerPage() {
      return { ...old, ...config };
    });

+    // Read AFTER optimistic update to get the latest merged state
+    const currentConfig = queryClient.getQueryData<CcwMcpConfig>(ccwMcpQueryKey) ?? ccwConfig;
+
    try {
-      await updateCcwConfig({ ...currentConfig, ...config });
+      // Only pass the fields that updateCcwConfig expects
+      await updateCcwConfig({
+        enabledTools: currentConfig.enabledTools,
+        projectRoot: currentConfig.projectRoot,
+        allowedDirs: currentConfig.allowedDirs,
+        enableSandbox: currentConfig.enableSandbox,
+      });
    } catch (error) {
      console.error('Failed to update CCW config:', error);
      queryClient.setQueryData(ccwMcpQueryKey, previousConfig);
--- a/ccw/frontend/src/stores/queueSchedulerStore.ts
+++ b/ccw/frontend/src/stores/queueSchedulerStore.ts
@@ -195,9 +195,21 @@ export const useQueueSchedulerStore = create<QueueSchedulerStore>()(
            'loadInitialState'
          );
        } catch (error) {
+          // Silently ignore network errors (backend not connected)
+          // Only log non-network errors
          const message = error instanceof Error ? error.message : 'Unknown error';
-          console.error('[QueueScheduler] loadInitialState error:', message);
-          set({ error: message }, false, 'loadInitialState/error');
+          const isNetworkError =
+            message.includes('Failed to fetch') ||
+            message.includes('NetworkError') ||
+            message.includes('Network request failed') ||
+            message.includes('ERR_CONNECTION_REFUSED') ||
+            message.includes('ERR_CONNECTION_RESET');
+
+          if (!isNetworkError) {
+            console.error('[QueueScheduler] loadInitialState error:', message);
+            set({ error: message }, false, 'loadInitialState/error');
+          }
+          // For network errors, keep state as-is without showing error
        }
      },

--- a/ccw/src/commands/cli.ts
+++ b/ccw/src/commands/cli.ts
@@ -51,9 +51,11 @@ function notifyDashboard(data: Record<string, unknown>): void {
    path: '/api/hook',
    method: 'POST',
    timeout: 2000, // 2 second timeout to prevent hanging
+    agent: false, // Disable Keep-Alive to allow process exit
    headers: {
      'Content-Type': 'application/json',
-      'Content-Length': Buffer.byteLength(payload)
+      'Content-Length': Buffer.byteLength(payload),
+      'Connection': 'close' // Ensure connection closes after response
    }
  });

@@ -93,9 +95,11 @@ function broadcastStreamEvent(eventType: string, payload: Record<string, unknown
    path: '/api/hook',
    method: 'POST',
    timeout: 1000, // Short timeout for streaming
+    agent: false, // Disable Keep-Alive to allow process exit
    headers: {
      'Content-Type': 'application/json',
-      'Content-Length': Buffer.byteLength(data)
+      'Content-Length': Buffer.byteLength(data),
+      'Connection': 'close' // Ensure connection closes after response
    }
  });

--- a/ccw/src/core/auth/csrf-manager.ts
+++ b/ccw/src/core/auth/csrf-manager.ts
@@ -56,6 +56,26 @@ export class CsrfTokenManager {
   */
  generateToken(sessionId: string): string {
    const tokens = this.generateTokens(sessionId, 1);
+    // If no slots available (session at max capacity), force generate anyway
+    // This ensures we always return a valid token
+    if (tokens.length === 0) {
+      const token = randomBytes(32).toString('hex');
+      const expiresAtMs = Date.now() + this.tokenTtlMs;
+      const record: CsrfTokenRecord = {
+        sessionId,
+        expiresAtMs,
+        used: false,
+      };
+      // Get or create session map
+      let sessionMap = this.sessionTokens.get(sessionId);
+      if (!sessionMap) {
+        sessionMap = new Map();
+        this.sessionTokens.set(sessionId, sessionMap);
+      }
+      sessionMap.set(token, record);
+      this.tokenToSession.set(token, sessionId);
+      return token;
+    }
    return tokens[0];
  }

--- a/ccw/src/core/routes/auth-routes.ts
+++ b/ccw/src/core/routes/auth-routes.ts
@@ -100,14 +100,16 @@ export async function handleAuthRoutes(ctx: RouteContext): Promise<boolean> {
    } else {
      // Batch token response (pool pattern)
      const tokens = tokenManager.generateTokens(sessionId, count);
-      const firstToken = tokens[0];
+
+      // If no tokens generated (session at max capacity), force generate one
+      const firstToken = tokens.length > 0 ? tokens[0] : tokenManager.generateToken(sessionId);

      // Set header and cookie with first token for compatibility
      res.setHeader('X-CSRF-Token', firstToken);
      setCsrfCookie(res, firstToken, 15 * 60);
      res.writeHead(200, { 'Content-Type': 'application/json; charset=utf-8' });
      res.end(JSON.stringify({
-        tokens,
+        tokens: tokens.length > 0 ? tokens : [firstToken],
        expiresIn: 15 * 60, // seconds
      }));
    }
--- a/ccw/src/core/routes/mcp-routes.ts
+++ b/ccw/src/core/routes/mcp-routes.ts
@@ -945,6 +945,11 @@ function addGlobalMcpServer(serverName: string, serverConfig: unknown) {
      config.mcpServers = {};
    }

+    // Debug logging for ccw-tools
+    if (serverName === 'ccw-tools') {
+      console.log('[addGlobalMcpServer] Saving ccw-tools config:', JSON.stringify(serverConfig, null, 2));
+    }
+
    // Add the server to top-level mcpServers
    config.mcpServers[serverName] = serverConfig;

--- a/codex-lens/src/codex_lens.egg-info/PKG-INFO
+++ b/codex-lens/src/codex_lens.egg-info/PKG-INFO
@@ -8,53 +8,53 @@ Project-URL: Homepage, https://github.com/openai/codex-lens
 Requires-Python: >=3.10
 Description-Content-Type: text/markdown
 License-File: LICENSE
-Requires-Dist: typer>=0.9
-Requires-Dist: rich>=13
-Requires-Dist: pydantic>=2.0
-Requires-Dist: tree-sitter>=0.20
-Requires-Dist: tree-sitter-python>=0.25
-Requires-Dist: tree-sitter-javascript>=0.25
-Requires-Dist: tree-sitter-typescript>=0.23
-Requires-Dist: pathspec>=0.11
-Requires-Dist: watchdog>=3.0
-Requires-Dist: ast-grep-py>=0.40.0
+Requires-Dist: typer~=0.9.0
+Requires-Dist: rich~=13.0.0
+Requires-Dist: pydantic~=2.0.0
+Requires-Dist: tree-sitter~=0.20.0
+Requires-Dist: tree-sitter-python~=0.25.0
+Requires-Dist: tree-sitter-javascript~=0.25.0
+Requires-Dist: tree-sitter-typescript~=0.23.0
+Requires-Dist: pathspec~=0.11.0
+Requires-Dist: watchdog~=3.0.0
+Requires-Dist: ast-grep-py~=0.40.0
 Provides-Extra: semantic
-Requires-Dist: numpy>=1.24; extra == "semantic"
-Requires-Dist: fastembed>=0.2; extra == "semantic"
-Requires-Dist: hnswlib>=0.8.0; extra == "semantic"
+Requires-Dist: numpy~=1.24.0; extra == "semantic"
+Requires-Dist: fastembed~=0.2.0; extra == "semantic"
+Requires-Dist: hnswlib~=0.8.0; extra == "semantic"
 Provides-Extra: semantic-gpu
-Requires-Dist: numpy>=1.24; extra == "semantic-gpu"
-Requires-Dist: fastembed>=0.2; extra == "semantic-gpu"
-Requires-Dist: hnswlib>=0.8.0; extra == "semantic-gpu"
-Requires-Dist: onnxruntime-gpu>=1.15.0; extra == "semantic-gpu"
+Requires-Dist: numpy~=1.24.0; extra == "semantic-gpu"
+Requires-Dist: fastembed~=0.2.0; extra == "semantic-gpu"
+Requires-Dist: hnswlib~=0.8.0; extra == "semantic-gpu"
+Requires-Dist: onnxruntime-gpu~=1.15.0; extra == "semantic-gpu"
 Provides-Extra: semantic-directml
-Requires-Dist: numpy>=1.24; extra == "semantic-directml"
-Requires-Dist: fastembed>=0.2; extra == "semantic-directml"
-Requires-Dist: hnswlib>=0.8.0; extra == "semantic-directml"
-Requires-Dist: onnxruntime-directml>=1.15.0; extra == "semantic-directml"
+Requires-Dist: numpy~=1.24.0; extra == "semantic-directml"
+Requires-Dist: fastembed~=0.2.0; extra == "semantic-directml"
+Requires-Dist: hnswlib~=0.8.0; extra == "semantic-directml"
+Requires-Dist: onnxruntime-directml~=1.15.0; extra == "semantic-directml"
 Provides-Extra: reranker-onnx
-Requires-Dist: optimum>=1.16; extra == "reranker-onnx"
-Requires-Dist: onnxruntime>=1.15; extra == "reranker-onnx"
-Requires-Dist: transformers>=4.36; extra == "reranker-onnx"
+Requires-Dist: optimum~=1.16.0; extra == "reranker-onnx"
+Requires-Dist: onnxruntime~=1.15.0; extra == "reranker-onnx"
+Requires-Dist: transformers~=4.36.0; extra == "reranker-onnx"
 Provides-Extra: reranker-api
-Requires-Dist: httpx>=0.25; extra == "reranker-api"
+Requires-Dist: httpx~=0.25.0; extra == "reranker-api"
 Provides-Extra: reranker-litellm
-Requires-Dist: ccw-litellm>=0.1; extra == "reranker-litellm"
+Requires-Dist: ccw-litellm~=0.1.0; extra == "reranker-litellm"
 Provides-Extra: reranker-legacy
-Requires-Dist: sentence-transformers>=2.2; extra == "reranker-legacy"
+Requires-Dist: sentence-transformers~=2.2.0; extra == "reranker-legacy"
 Provides-Extra: reranker
-Requires-Dist: optimum>=1.16; extra == "reranker"
-Requires-Dist: onnxruntime>=1.15; extra == "reranker"
-Requires-Dist: transformers>=4.36; extra == "reranker"
+Requires-Dist: optimum~=1.16.0; extra == "reranker"
+Requires-Dist: onnxruntime~=1.15.0; extra == "reranker"
+Requires-Dist: transformers~=4.36.0; extra == "reranker"
 Provides-Extra: encoding
-Requires-Dist: chardet>=5.0; extra == "encoding"
+Requires-Dist: chardet~=5.0.0; extra == "encoding"
 Provides-Extra: clustering
-Requires-Dist: hdbscan>=0.8.1; extra == "clustering"
-Requires-Dist: scikit-learn>=1.3.0; extra == "clustering"
+Requires-Dist: hdbscan~=0.8.1; extra == "clustering"
+Requires-Dist: scikit-learn~=1.3.0; extra == "clustering"
 Provides-Extra: full
-Requires-Dist: tiktoken>=0.5.0; extra == "full"
+Requires-Dist: tiktoken~=0.5.0; extra == "full"
 Provides-Extra: lsp
-Requires-Dist: pygls>=1.3.0; extra == "lsp"
+Requires-Dist: pygls~=1.3.0; extra == "lsp"
 Dynamic: license-file

 # CodexLens
--- a/codex-lens/src/codex_lens.egg-info/SOURCES.txt
+++ b/codex-lens/src/codex_lens.egg-info/SOURCES.txt
@@ -153,10 +153,12 @@ tests/test_hybrid_chunker.py
 tests/test_hybrid_search_e2e.py
 tests/test_hybrid_search_reranker_backend.py
 tests/test_hybrid_search_unit.py
+tests/test_incremental_indexer.py
 tests/test_incremental_indexing.py
 tests/test_litellm_reranker.py
 tests/test_lsp_graph_builder_depth.py
 tests/test_merkle_detection.py
+tests/test_migrations.py
 tests/test_parser_integration.py
 tests/test_parsers.py
 tests/test_path_mapper_windows_drive.py
--- a/codex-lens/src/codex_lens.egg-info/requires.txt
+++ b/codex-lens/src/codex_lens.egg-info/requires.txt
@@ -1,59 +1,59 @@
-typer>=0.9
-rich>=13
-pydantic>=2.0
-tree-sitter>=0.20
-tree-sitter-python>=0.25
-tree-sitter-javascript>=0.25
-tree-sitter-typescript>=0.23
-pathspec>=0.11
-watchdog>=3.0
-ast-grep-py>=0.40.0
+typer~=0.9.0
+rich~=13.0.0
+pydantic~=2.0.0
+tree-sitter~=0.20.0
+tree-sitter-python~=0.25.0
+tree-sitter-javascript~=0.25.0
+tree-sitter-typescript~=0.23.0
+pathspec~=0.11.0
+watchdog~=3.0.0
+ast-grep-py~=0.40.0

 [clustering]
-hdbscan>=0.8.1
-scikit-learn>=1.3.0
+hdbscan~=0.8.1
+scikit-learn~=1.3.0

 [encoding]
-chardet>=5.0
+chardet~=5.0.0

 [full]
-tiktoken>=0.5.0
+tiktoken~=0.5.0

 [lsp]
-pygls>=1.3.0
+pygls~=1.3.0

 [reranker]
-optimum>=1.16
-onnxruntime>=1.15
-transformers>=4.36
+optimum~=1.16.0
+onnxruntime~=1.15.0
+transformers~=4.36.0

 [reranker-api]
-httpx>=0.25
+httpx~=0.25.0

 [reranker-legacy]
-sentence-transformers>=2.2
+sentence-transformers~=2.2.0

 [reranker-litellm]
-ccw-litellm>=0.1
+ccw-litellm~=0.1.0

 [reranker-onnx]
-optimum>=1.16
-onnxruntime>=1.15
-transformers>=4.36
+optimum~=1.16.0
+onnxruntime~=1.15.0
+transformers~=4.36.0

 [semantic]
-numpy>=1.24
-fastembed>=0.2
-hnswlib>=0.8.0
+numpy~=1.24.0
+fastembed~=0.2.0
+hnswlib~=0.8.0

 [semantic-directml]
-numpy>=1.24
-fastembed>=0.2
-hnswlib>=0.8.0
-onnxruntime-directml>=1.15.0
+numpy~=1.24.0
+fastembed~=0.2.0
+hnswlib~=0.8.0
+onnxruntime-directml~=1.15.0

 [semantic-gpu]
-numpy>=1.24
-fastembed>=0.2
-hnswlib>=0.8.0
-onnxruntime-gpu>=1.15.0
+numpy~=1.24.0
+fastembed~=0.2.0
+hnswlib~=0.8.0
+onnxruntime-gpu~=1.15.0
--- a/package-lock.json
+++ b/package-lock.json
@@ -1,12 +1,12 @@
 {
  "name": "claude-code-workflow",
-  "version": "7.1.0",
+  "version": "7.2.0",
  "lockfileVersion": 3,
  "requires": true,
  "packages": {
    "": {
      "name": "claude-code-workflow",
-      "version": "7.1.0",
+      "version": "7.2.0",
      "hasInstallScript": true,
      "license": "MIT",
      "workspaces": [
--- a/package.json
+++ b/package.json
@@ -1,6 +1,6 @@
 {
  "name": "claude-code-workflow",
-  "version": "7.1.0",
+  "version": "7.2.1",
  "description": "JSON-driven multi-agent development framework with intelligent CLI orchestration (Gemini/Qwen/Codex), context-first architecture, and automated workflow execution",
  "type": "module",
  "main": "ccw/dist/index.js",
Author	SHA1	Message	Date
catlog22	71485b89e6	fix(workflow): resolve analyze-with-file and lite-plan phase conflict Add workflow boundary markers to analyze-with-file Post-Completion Options so that when "生成任务" triggers lite-plan via Skill(), the model follows lite-plan's Phase 1-5 exclusively instead of returning to analyze Phase 4 synthesis. Root cause: Skill() expands inline (shared context), causing phase number collision between the two workflows.	2026-03-02 13:08:59 +08:00
catlog22	2fc792a3b7	7.2.0	2026-03-02 12:28:39 +08:00
catlog22	0af4ca040f	fix(mcp): improve CCW config update logic and add debug logging - Fix McpManagerPage to read config after optimistic update and pass only expected fields - Add debug logging for enabledTools config building and ccw-tools server saving	2026-03-02 12:27:42 +08:00
catlog22	7af258f43d	feat(routing): add greenfield task type and narrow roadmap detection - Add dedicated greenfield (0→1) task type with complexity-adaptive routing (low→brainstorm-to-plan, medium→greenfield-plan, high→greenfield-phased) - Narrow roadmap regex to only /roadmap\|路线.*图/ to prevent false matches - Change brainstorm-to-plan chain from lite-plan to workflow-plan→execute - Sync all changes between ccw.md and ccw-coordinator.md	2026-03-02 12:27:31 +08:00
catlog22	8ad283086b	fix(tests): add test for disabling all tools in CcwToolsMcpCard component fix(api): handle empty enabledTools array and improve default tool logic fix(queueScheduler): ignore network errors in loadInitialState fix(auth): ensure token generation handles max session capacity chore(dependencies): update package requirements to use compatible version specifiers chore(tests): add new test cases for incremental indexer and migrations	2026-03-02 11:26:15 +08:00
catlog22	b36a46d59d	fix(ui): prevent dropdown truncation in Collapsible content Apply overflow-hidden only when collapsible is closed, allowing dropdown menus to overflow properly when expanded.	2026-03-02 11:19:28 +08:00
catlog22	899a92f2eb	fix(skill): constrain roadmap interactive feedback to plan-only modifications Add explicit constraints preventing Phase 3 interactive rounds from modifying source code. Feedback handlers now clearly scope updates to roadmap.md and issues.jsonl only, with code changes deferred to handoff.	2026-03-02 11:17:42 +08:00
catlog22	d0ac3a5cd2	fix(csrf): prevent undefined token when session at max capacity Root cause: generateToken() returned undefined when session already had maxTokensPerSession (5) tokens, causing ERR_HTTP_INVALID_HEADER_VALUE. Fix: Force generate token even when at capacity, ensuring we always return a valid token string. Related: v7.1.1 CLI process hang fix	2026-03-02 09:58:54 +08:00
catlog22	0939510e0d	fix(lite-plan): handle empty exploration context when prior analysis exists When analyze-with-file artifacts are passed to lite-plan, exploration is skipped (hasPriorAnalysis=true) leaving manifest.explorations empty. The agent prompt's exploration section rendered as blank, causing the planning agent to lack context and fail to produce task JSON. Add conditional to output a fallback hint directing the agent to use the Prior Analysis block in the task description as primary planning context.	2026-03-02 00:04:32 +08:00
catlog22	deea92581b	fix(cli): resolve process hang after CLI execution Root cause: HTTP Keep-Alive connections kept event loop alive, preventing process.exit() from executing even after CLI_EXECUTION_COMPLETED event was sent. Fix: Add `agent: false` and `Connection: close` header to HTTP requests in notifyDashboard() and broadcastStreamEvent() functions. - agent: false - Creates new Agent per request instead of global Keep-Alive Agent - Connection: close - Tells server to close connection after response Fixes issue where `ccw cli --tool gemini` would complete execution but Bash command would hang indefinitely.	2026-03-01 23:34:31 +08:00