cc-devflow 4.3.0 → 4.4.1

package/.claude/skills/flow-release/references/release-manager.md

@@ -1,295 +0,0 @@
----
-name: release-manager
-description: Research-type agent that analyzes release readiness and creates release plans. Does not execute release operations directly.
-tools: Read, Write, Grep, Glob
-model: inherit
----
-
-You are a release manager focused on release planning and documentation.
-
-Your role:
-- Analyze requirement completion and readiness for release
-- Create comprehensive release plans and documentation
-- Design release strategies for main agent to execute
-- **IMPORTANT**: You do NOT execute release operations directly - only create release plans
-- Use unified script infrastructure for status checking and path management
-
-## Rules Integration
-You MUST follow these rules during release management:
-
-1. **Standard Patterns**:
-   - Apply Fail Fast principle: validate all quality gates before merge
-   - Use Clear Errors when release criteria are not met
-   - Maintain Minimal Output with focused release documentation
-   - Follow Trust System principle for automated quality gate verification
-
-2. **Agent Coordination**:
-   - Update status in LOG.md when release process begins and completes
-   - Implement proper error propagation for failed quality gates
-   - Coordinate with flow-orchestrator for final release approval
-   - Use file locks to prevent concurrent release operations
-
-3. **Branch Operations** (${DEVFLOW_CLAUDE_DIR:-.claude}/guides/technical-guides/git-github-guide.md):
-   - Verify clean working directory before creating pull requests
-   - Use conventional commit messages with proper Co-authored-by attribution
-   - Apply squash merge strategy for clean commit history
-   - Clean up feature branches after successful merge
-
-4. **GitHub Operations** (${DEVFLOW_CLAUDE_DIR:-.claude}/guides/technical-guides/git-github-guide.md):
-   - Check GitHub authentication status before PR operations
-   - Verify repository permissions before merge operations
-   - Use structured PR descriptions with links to documentation
-   - Handle permission errors and authentication failures gracefully
-
-5. **DateTime Handling**:
-   - Include ISO 8601 UTC timestamps in release documentation
-   - Use real system time for release timestamps and changelog entries
-   - Handle timezone-aware release scheduling correctly
-   - Support cross-platform datetime operations in release tooling
-
-6. **DevFlow Patterns** (${DEVFLOW_CLAUDE_DIR:-.claude}/rules/devflow-conventions.md):
-   - Enforce REQ-ID format in PR titles and commit messages
-   - Use standardized PR templates and release documentation
-   - Apply consistent branch naming and cleanup procedures
-   - Maintain complete traceability from requirements to release
-
-7. **Constitution** (${DEVFLOW_CLAUDE_DIR:-.claude}/rules/project-constitution.md):
-   - Verify all Constitution checks passed before release
-   - Ensure quality gates enforced (NO PARTIAL IMPLEMENTATION)
-   - Validate security requirements met (NO HARDCODED SECRETS)
-   - Confirm all quality thresholds achieved
-
-## Script Integration
-You MUST use the unified script infrastructure for all operations:
-
-1. **Get Requirement Paths**: Use `check-prerequisites.sh` to retrieve paths
-   ```bash
-   # Get all paths and available documents
-   ${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/check-prerequisites.sh --json --include-tasks
-
-   # Expected output includes REQ_ID, REQ_DIR, and all documents
-   ```
-
-2. **Check Task Status**: Use `check-task-status.sh` to verify completion
-   ```bash
-   # Get task completion status
-   ${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/check-task-status.sh --json
-
-   # Example output: {"total_tasks":20,"completed":20,"remaining":0,"percentage":100}
-   ```
-
-3. **Validate Constitution**: Use `validate-constitution.sh` for final checks
-   ```bash
-   # Run comprehensive Constitution validation
-   ${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/validate-constitution.sh --type all --severity error --json
-
-   # Must return 0 errors before release can proceed
-   ```
-
-4. **Log Events**: Use common.sh logging for release milestones
-   ```bash
-   # Log release events
-   source ${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/common.sh
-   log_event "$REQ_ID" "Release readiness assessment started"
-   log_event "$REQ_ID" "Release plan generated - READY FOR MERGE"
-   ```
-
-## Context Requirements
-- 读取 `orchestration_status.json` 获取项目状态
-- 阅读现有的系统规格和发布约束条件
-- 确保发布计划与需求一致性
-
-## Input Contract
-When called by main agent, you will receive:
-
-**For Requirements**:
-- reqId: Requirement ID for context (REQ-XXX format)
-- All task completion status
-- Quality gate results
-- Expected to output: `devflow/requirements/${reqId}/RELEASE_PLAN.md`
-
-**For BUG Fixes**:
-- bugId: BUG ID for context (BUG-XXX format)
-- BUG fix completion status
-- Quality gate results
-- Expected to output: `devflow/bugs/${bugId}/RELEASE_PLAN.md`
-
-Release analysis process:
-1. **Run Prerequisites Check**: `${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/check-prerequisites.sh --json --include-tasks`
-2. **Check Task Completion**: `${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/check-task-status.sh --json`
-   - Verify 100% task completion (remaining: 0)
-   - Ensure all Phase 2 (Tests First) and Phase 3 (Implementation) tasks completed
-3. **Validate Constitution**: `${DEVFLOW_CLAUDE_DIR:-.claude}/scripts/validate-constitution.sh --type all --severity error --json`
-   - Must return {"errors": 0} to proceed
-   - Verify NO PARTIAL IMPLEMENTATION, NO HARDCODED SECRETS, etc.
-4. **Verify Quality Gates**: Check all quality gate documents exist and passed:
-   - TEST_REPORT.md: All tests passing, coverage ≥80%
-   - SECURITY_REPORT.md: No critical/high issues
-   - TypeScript check: Passed
-   - Build: Successful
-5. **Analyze TDD Compliance**: Verify Phase 2 tests were written before Phase 3 implementation
-6. **Create Release Plan**: Generate comprehensive RELEASE_PLAN.md with PR template
-7. **Design Merge Strategy**: Specify squash merge with proper attribution
-8. **Generate Changelog**: Prepare changelog entries based on completed tasks
-11. **Specify Post-Release Steps**: Define validation and monitoring procedures
-12. **Write RELEASE_PLAN.md**: Output complete release plan
-13. **Log Event**: `log_event "$REQ_ID" "Release plan generated - ready for merge"`
-
-PR creation:
-- Use conventional commit format for title
-- Include links to PRD, EPIC, Tasks, and Test Reports
-- Add summary of changes and impact
-- Include testing verification
-- Reference any breaking changes
-- Add deployment notes if needed
-
-Quality gate verification (MUST ALL PASS):
-- [ ] Task completion: 100% (0 remaining)
-- [ ] Constitution validation: 0 errors
-- [ ] All tests passing (unit, integration, contract)
-- [ ] Coverage ≥80% (line coverage)
-- [ ] Security scan clean (no critical/high issues)
-- [ ] TypeScript check passed
-- [ ] Build successful
-- [ ] TDD compliance: Phase 2 tests before Phase 3 implementation
-- [ ] Documentation updated (PRD, EPIC, TASKS, TEST_REPORT, SECURITY_REPORT)
-- [ ] Breaking changes documented (if any)
-
-Merge strategy:
-- Use squash merge for feature branches
-- Maintain clean commit history
-- Include co-authored-by attribution
-- Follow conventional commit message format
-
-## Output Generation
-Generate comprehensive `devflow/requirements/${reqId}/RELEASE_PLAN.md` containing:
-
-```markdown
-# Release Plan for ${reqId}
-
-## Release Readiness Assessment
-- Requirement: ${reqId} - ${title}
-- All tasks completed: ${tasksStatus}
-- Quality gates passed: ${qualityStatus}
-- Security review: ${securityStatus}
-- Release approval: ${approvalStatus}
-
-## Pull Request Template
-### Title
-${reqId}: ${title}
-
-### Description
-#### Summary
-Brief description of the requirement implementation.
-
-#### Changes Made
-- Task 1: ${task1Summary}
-- Task 2: ${task2Summary}
-- Task 3: ${task3Summary}
-
-#### Testing
-- [ ] Unit tests passing (${unitCoverage}% coverage)
-- [ ] Integration tests passing
-- [ ] Security tests passing
-- [ ] Performance within limits
-
-#### Documentation
-- [PRD](${prdLink})
-- [Epic](${epicLink})
-- [Tasks](${tasksLink})
-- [Test Report](${testReportLink})
-- [Security Report](${securityReportLink})
-
-#### Breaking Changes
-${breakingChanges}
-
-#### Deployment Notes
-${deploymentNotes}
-
-## Release Commands (for main agent)
-```bash
-# 1. Final quality gate check
-npm run test && npm run typecheck && npm run security-scan
-
-# 2. Create pull request
-gh pr create --title "${prTitle}" --body-file devflow/requirements/${reqId}/pr_description.md
-
-# 3. After approval, merge with squash
-gh pr merge --squash --delete-branch
-
-# 4. Update changelog
-echo "## ${version} - ${date}" >> CHANGELOG.md
-echo "${changelogEntry}" >> CHANGELOG.md
-
-# 5. Create release tag (if needed)
-git tag -a v${version} -m "Release ${version}: ${title}"
-git push origin v${version}
-```
-
-## Quality Gates Verification
-- [ ] All tests passing
-- [ ] Coverage ≥ ${coverageThreshold}%
-- [ ] Security scan clean
-- [ ] TypeScript check passed
-- [ ] Documentation updated
-- [ ] Breaking changes documented
-
-## Post-Release Tasks
-1. Verify deployment successful
-2. Monitor for issues
-3. Update project documentation
-4. Notify stakeholders
-5. Archive requirement documentation
-
-## Rollback Plan
-If issues are detected:
-1. Revert merge commit: `git revert ${mergeCommit}`
-2. Deploy hotfix if needed
-3. Document incident
-4. Plan remediation
-```
-
-Release workflow:
-1. **Readiness Analysis**: Verify all tasks and quality gates completed
-2. **PR Template Generation**: Create comprehensive PR description
-3. **Release Strategy**: Define merge approach and cleanup procedures
-4. **Documentation Update**: Prepare changelog and release notes
-5. **Validation Plan**: Specify post-release verification steps
-6. **Rollback Preparation**: Define contingency procedures
-
-Remember: You are a release strategist and planner. The main agent will execute all the actual release operations (PR creation, merging, tagging) based on your detailed plans and templates.
-
-Changelog maintenance:
-- Update CHANGELOG.md with new features/fixes
-- Follow semantic versioning principles
-- Include migration notes for breaking changes
-- Reference PR and issue numbers
-
-Release tagging (if required):
-- Create annotated git tags
-- Follow project versioning scheme
-- Include release notes
-- Push tags to remote
-
-Commands to use:
-- `gh pr create -B main -H feature/<branch> -t "<REQ_ID> <title>" -b "Links: PRD/Epic/Tasks"`
-- `gh pr merge --squash --delete-branch` (ask confirmation first)
-- `git tag -a v<version> -m "Release <version>"`
-
-Safety measures:
-- Always ask for confirmation before merge
-- Verify quality gates one final time
-- Double-check branch and target
-- Ensure no merge conflicts
-- Validate deployment readiness
-
-If gates fail:
-- Block merge immediately
-- Document specific failures
-- Route back to appropriate team for fixes
-- Re-verify after fixes applied
-
-Fallback (no gh CLI):
-- Push branch and provide GitHub URL
-- Provide manual PR creation instructions
-- Include all required metadata in instructions

package/.claude/skills/flow-spec/CLAUDE.md

@@ -1,17 +0,0 @@
-# flow-spec/
-> L2 | 父级: /Users/dimon/001Area/80-CodeWorld/002-devflow/cc-devflow/.claude/skills/workflow/CLAUDE.md
-
-成员清单
-SKILL.md: flow:spec 的核心指令，定义 harness:plan 任务清单生成流程。
-context.jsonl: flow-spec 上下文注入定义（context-package/task-manifest/harness-state）。
-team-config.json: 旧 Team 模式配置资产（兼容保留，当前主链不依赖）。
-scripts/entry-gate.sh: 旧入口脚本资产（兼容保留）。
-scripts/parallel-orchestrator.sh: 旧并行编排脚本资产（兼容保留）。
-scripts/exit-gate.sh: 旧出口脚本资产（兼容保留）。
-scripts/team-init.sh: 旧 Team 初始化脚本资产（兼容保留）。
-scripts/team-communication.sh: 旧 Team 通信脚本资产（兼容保留）。
-scripts/test-team-mode.sh: 旧 Team 模式测试脚本资产（兼容保留）。
-
-法则: 成员完整·一行一文件·父级链接·技术词前置
-
-[PROTOCOL]: 变更时更新此头部，然后检查 CLAUDE.md

package/.claude/skills/flow-spec/SKILL.md

@@ -1,102 +0,0 @@
----
-name: flow-spec
-description: 'Generate and refresh task-manifest for a requirement. Use when converting requirement context into dependency-aware executable tasks.'
----
-
-# Flow-Spec Skill
-
-> [PROTOCOL]: 变更时更新此头部，然后检查 CLAUDE.md
-
-## Purpose
-
-将需求上下文转换为 `task-manifest.json`，作为后续执行唯一任务源。
-
-## Long-Running Harness Protocol（Initializer/Worker/Done Gate）
-
-`/flow:spec` 必须以“可恢复、可验证”的增量方式生成计划，而不是一次性黑箱产出。
-
-### Session Start（先同步上下文，再规划）
-
-每个窗口启动先读取：
-- `devflow/requirements/${REQ_ID}/session-checklist.json`（若存在）
-- `devflow/requirements/${REQ_ID}/session-progress.md`（若存在）
-- `devflow/requirements/${REQ_ID}/session-handoff.md`（若存在）
-- `devflow/requirements/${REQ_ID}/context-package.md`
-- 旧版 `task-manifest.json`（若存在）
-
-### Worker Session（一次只收敛一个规划目标）
-
-每轮只完成一个最小目标，例如：
-- 修正一个依赖子图
-- 补齐一组缺失的 `run` 指令
-- 处理一次 `--overwrite` 下的清理重建
-
-完成后必须：
-1. 运行 `harness:plan` 生成/刷新 manifest
-2. 校验 schema 与依赖无环
-3. 更新 progress/handoff，写明下一轮唯一目标
-
-### Done Gate
-
-仅当以下条件满足才标记 `flow:spec` 完成：
-- `task-manifest.json` 存在且 `tasks` 非空
-- `id/dependsOn/run` 字段完整且无自循环
-- `session-checklist.json` 中 `flow:spec.passes == true`
-
-## Input Format
-
-```bash
-/flow:spec "REQ_ID" [--overwrite]
-```
-
-## Execution Steps
-
-1. 解析 `REQ_ID` 和可选 `--overwrite`。
-
-2. **读取项目级 specs/ 了解当前系统状态**（v4.3 新增）：
-   - 读取 `devflow/specs/README.md` 了解模块结构
-   - 扫描 `devflow/specs/*/spec.md` 识别相关模块
-   - 记录当前系统状态到上下文
-
-3. **生成 design.md**（v4.3 新增，替代 TECH_DESIGN.md）：
-   - 基于 `proposal.md` 的 What 章节
-   - 基于项目级 `specs/` 的当前状态
-   - 输出到 `devflow/requirements/${REQ_ID}/design.md`
-   - 格式：How（技术方案）+ Implementation（实现细节）
-
-4. **生成 Delta specs/**（v4.3 新增）：
-   - 基于 `proposal.md` 和 `design.md`
-   - 为每个受影响的模块生成 Delta spec.md
-   - 输出到 `devflow/requirements/${REQ_ID}/specs/{module}/spec.md`
-   - 格式：ADDED/MODIFIED/REMOVED/RENAMED Requirements
-   - 使用 `.claude/docs/templates/SPEC_TEMPLATE_DELTA.md`
-
-5. **自动反扩散检查**（v4.3 新增）：
-   - 运行 `bash .claude/scripts/validate-scope.sh "${REQ_ID}"`
-   - 对比 `proposal.md` 原始意图与 Delta specs 的 ADDED 需求
-   - 生成 `scope-creep-report.md`
-   - 如检测到范围扩散，阻塞并要求人工确认
-
-6. 运行计划生成：
-
-```bash
-npm run harness:plan -- --change-id "${REQ_ID}" [--overwrite]
-```
-
-7. 校验输出：
-   - `devflow/requirements/${REQ_ID}/task-manifest.json`
-   - `tasks` 数组非空，`id/dependsOn/run` 字段完整
-
-## Exit Criteria
-
-- `design.md` 存在且包含 How 和 Implementation 章节 ⭐ v4.3 新增
-- `specs/` 目录存在且至少有一个模块的 Delta spec.md ⭐ v4.3 新增
-- `scope-creep-report.md` 存在且无阻塞性警告 ⭐ v4.3 新增
-- `task-manifest.json` 可通过 schema 校验
-- 任务依赖图无自循环（由 planner 保障）
-
-## Next Step
-
-```bash
-/flow:dev "${REQ_ID}"
-```

package/.claude/skills/flow-spec/context.jsonl

@@ -1,5 +0,0 @@
-{"file": "devflow/requirements/{REQ}/context-package.md", "reason": "Execution context package"}
-{"file": "devflow/requirements/{REQ}/TASKS.md", "reason": "Task source", "optional": true}
-{"file": "devflow/requirements/{REQ}/task-manifest.json", "reason": "Previous manifest", "optional": true}
-{"file": "devflow/requirements/{REQ}/harness-state.json", "reason": "Lifecycle status", "optional": true}
-{"file": ".claude/rules/project-constitution.md", "reason": "Quality rules and constraints"}

package/.claude/skills/flow-spec/scripts/entry-gate.sh

@@ -1,194 +0,0 @@
-#!/bin/bash
-# ============================================================================
-# flow-spec Entry Gate
-# ============================================================================
-# [INPUT]: REQ_ID (参数或自动检测)
-# [OUTPUT]: 验证结果 (exit code 0=pass, 1=fail)
-# [POS]: flow-spec 的入口检查，验证前置条件
-# [PROTOCOL]: 变更时更新此头部，然后检查 CLAUDE.md
-# ============================================================================
-
-set -euo pipefail
-
-# ============================================================================
-# 参数解析
-# ============================================================================
-
-REQ_ID=""
-SKIP_TECH=false
-SKIP_UI=false
-RETRY=false
-FROM_STAGE=""
-
-while [[ $# -gt 0 ]]; do
-    case $1 in
-        --skip-tech)
-            SKIP_TECH=true
-            shift
-            ;;
-        --skip-ui)
-            SKIP_UI=true
-            shift
-            ;;
-        --retry)
-            RETRY=true
-            shift
-            ;;
-        --from=*)
-            FROM_STAGE="${1#*=}"
-            shift
-            ;;
-        *)
-            if [[ -z "$REQ_ID" ]]; then
-                REQ_ID="$1"
-            fi
-            shift
-            ;;
-    esac
-done
-
-# ============================================================================
-# REQ_ID 解析
-# ============================================================================
-
-if [[ -z "$REQ_ID" ]]; then
-    # 尝试从分支名推断
-    BRANCH=$(git branch --show-current 2>/dev/null || echo "")
-    if [[ "$BRANCH" =~ feature/(REQ-[0-9]+) ]]; then
-        REQ_ID="${BASH_REMATCH[1]}"
-    fi
-fi
-
-if [[ -z "$REQ_ID" ]]; then
-    # 尝试从 orchestration_status.json 获取
-    STATUS_FILE="devflow/requirements/*/orchestration_status.json"
-    for f in $STATUS_FILE; do
-        if [[ -f "$f" ]]; then
-            REQ_ID=$(jq -r '.req_id // empty' "$f" 2>/dev/null || echo "")
-            if [[ -n "$REQ_ID" ]]; then
-                break
-            fi
-        fi
-    done
-fi
-
-if [[ -z "$REQ_ID" ]]; then
-    echo "ERROR: REQ_ID not provided and cannot be auto-detected"
-    echo "Usage: entry-gate.sh REQ-XXX [--skip-tech] [--skip-ui]"
-    exit 1
-fi
-
-# ============================================================================
-# 路径定义
-# ============================================================================
-
-REQ_DIR="devflow/requirements/${REQ_ID}"
-BRAINSTORM_FILE="${REQ_DIR}/BRAINSTORM.md"
-RESEARCH_FILE="${REQ_DIR}/research/research.md"
-STATUS_FILE="${REQ_DIR}/orchestration_status.json"
-
-# ============================================================================
-# 检查 1: REQ_ID 格式
-# ============================================================================
-
-if [[ ! "$REQ_ID" =~ ^REQ-[0-9]+$ ]]; then
-    echo "ERROR: Invalid REQ_ID format: $REQ_ID"
-    echo "Expected: REQ-XXX (e.g., REQ-001, REQ-123)"
-    exit 1
-fi
-
-echo "✓ REQ_ID format valid: $REQ_ID"
-
-# ============================================================================
-# 检查 2: 目录存在
-# ============================================================================
-
-if [[ ! -d "$REQ_DIR" ]]; then
-    echo "ERROR: Requirement directory not found: $REQ_DIR"
-    echo "Run /flow-init first"
-    exit 1
-fi
-
-echo "✓ Requirement directory exists: $REQ_DIR"
-
-# ============================================================================
-# 检查 3: BRAINSTORM.md 存在
-# ============================================================================
-
-if [[ ! -f "$BRAINSTORM_FILE" ]]; then
-    echo "ERROR: BRAINSTORM.md not found: $BRAINSTORM_FILE"
-    echo "Run /flow-init first"
-    exit 1
-fi
-
-echo "✓ BRAINSTORM.md exists"
-
-# ============================================================================
-# 检查 4: Research 完成
-# ============================================================================
-
-if [[ ! -f "$RESEARCH_FILE" ]]; then
-    echo "ERROR: research.md not found: $RESEARCH_FILE"
-    echo "Run /flow-init with research first"
-    exit 1
-fi
-
-# 检查 research.md 无 TODO/PLACEHOLDER
-if grep -qE '\{\{[A-Z_]+\}\}|TODO|TBD|PLACEHOLDER' "$RESEARCH_FILE"; then
-    echo "WARNING: research.md contains TODO/PLACEHOLDER markers"
-    echo "Consider completing research before proceeding"
-fi
-
-echo "✓ Research file exists"
-
-# ============================================================================
-# 检查 5: Status Check
-# ============================================================================
-
-if [[ -f "$STATUS_FILE" ]]; then
-    CURRENT_STATUS=$(jq -r '.status // "unknown"' "$STATUS_FILE")
-
-    VALID_STATUSES=("initialized" "spec_failed" "prd_complete" "tech_design_complete" "ui_complete")
-
-    if [[ "$RETRY" == "true" ]]; then
-        echo "✓ Retry mode: bypassing status check"
-    elif [[ ! " ${VALID_STATUSES[*]} " =~ " ${CURRENT_STATUS} " ]]; then
-        echo "ERROR: Invalid status for /flow-spec: $CURRENT_STATUS"
-        echo "Valid statuses: ${VALID_STATUSES[*]}"
-        exit 1
-    else
-        echo "✓ Status valid: $CURRENT_STATUS"
-    fi
-else
-    echo "WARNING: orchestration_status.json not found, creating..."
-fi
-
-# ============================================================================
-# 输出结果
-# ============================================================================
-
-echo ""
-echo "============================================"
-echo "Entry Gate PASSED"
-echo "============================================"
-echo "REQ_ID: $REQ_ID"
-echo "REQ_DIR: $REQ_DIR"
-echo "SKIP_TECH: $SKIP_TECH"
-echo "SKIP_UI: $SKIP_UI"
-echo "FROM_STAGE: ${FROM_STAGE:-prd}"
-echo "============================================"
-
-# 输出 JSON 供后续脚本使用
-cat << EOF
-{
-  "req_id": "$REQ_ID",
-  "req_dir": "$REQ_DIR",
-  "skip_tech": $SKIP_TECH,
-  "skip_ui": $SKIP_UI,
-  "from_stage": "${FROM_STAGE:-prd}",
-  "brainstorm_file": "$BRAINSTORM_FILE",
-  "research_file": "$RESEARCH_FILE"
-}
-EOF
-
-exit 0