catlog22/Claude-Code-Workflow
1
0
Fork 0
mirror of https://github.com/catlog22/Claude-Code-Workflow.git synced 2026-02-11 02:33:51 +08:00
Code Issues Packages Projects Releases Wiki Activity
531 Commits 20 Branches 102 Tags
eeffa8a9e8740a52df048b8c63071f3d4bbd2b78
Commit Graph

232 Commits

Author SHA1 Message Date
catlog22
eeffa8a9e8 Enhance UI Design Workflows with Intelligent Path Detection and Source Selection 
- Updated explore-auto.md to include a new phase for intelligent path detection and source selection, allowing the system to identify existing file paths from prompts and images.
- Introduced conditional code import and completeness assessment based on detected design sources (code only, visual only, hybrid).
- Modified phase execution flow to accommodate new checks for style, animation, and layout completeness based on the design source.
- Added error handling for missing design elements and user prompts for visual supplementation.
- Enhanced imitate-auto.md with intelligent path detection and a new phase for code import and completeness assessment, ensuring a hybrid approach when applicable.
- Implemented detailed reporting for each phase, including missing elements and recommendations for improvement.
- Created a comprehensive import-from-code.md file outlining the workflow for importing design systems from code files, detailing the execution process and error handling.
 2025-11-06 21:00:49 +08:00
catlog22
920b179440 docs: 更新所有命令描述并重新生成索引文件 
- 更新所有69个命令文件的description字段，基于实际功能重新生成详细描述
- 重新生成5个索引文件（all-commands, by-category, by-use-case, essential-commands, command-relationships）
- 移动analyze_commands.py到scripts/目录并完善功能
- 移除临时备份文件

命令描述改进示例：
- workflow:plan: 增加了工具和代理的详细说明（Gemini, action-planning-agent）
- cli:execute: 说明了YOLO权限和多种执行模式
- memory:update-related: 详细说明了批处理策略和工具回退链

索引文件改进：
- usage_scenario从2种扩展到10种（更精细分类）
- command-relationships覆盖所有69个命令
- 区分built-in（内置调用）和sequential（用户顺序执行）关系

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-06 15:11:31 +08:00
catlog22
854464b221 refactor: 重组CLI模板系统，采用优先级前缀命名 
主要变更：
- 模板重命名：采用优先级前缀（01-通用, 02-专用, 03-领域特定）
- 目录调整：bug-diagnosis从development移至analysis
- 引用更新：5个命令文件中21处模板引用更新为新路径
- 路径统一：所有引用统一使用完整路径格式

模板变更详情：
- analysis/：8个模板（01-trace-code-execution, 01-diagnose-bug-root-cause等）
- development/：5个模板（02-implement-feature, 02-refactor-codebase等）
- planning/：5个模板（01-plan-architecture-design, 02-breakdown-task-steps等）
- memory/：1个模板（02-document-module-structure）

命令文件更新：
- cli/mode/bug-diagnosis.md（6处引用）
- cli/mode/code-analysis.md（6处引用）
- cli/mode/plan.md（6处引用）
- task/execute.md（1处引用）
- workflow/tools/test-task-generate.md（2处引用）

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-06 10:57:17 +08:00
catlog22
dd04433079 docs: 在任务生成命令中添加focus_paths路径清晰度准则 
在三个任务生成命令中添加核心准则：
- task-generate.md
- task-generate-tdd.md
- task-generate-agent.md

路径要求：
- 优先使用绝对路径（如 D:\project\src\module）
- 或使用从项目根目录的清晰相对路径（如 ./src/module）
- 避免模糊的相对路径（如 analysis, interfaces）

更新示例JSON展示混合使用两种路径方式

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-05 21:51:58 +08:00
catlog22
c36ff8fcec docs: 继续标准化工作流会话和版本命令文档格式 
移除剩余命令文档中的表情符号和视觉装饰：
- version.md - 版本检查和更新命令
- workflow/session/list.md - 会话列表命令
- workflow/session/resume.md - 会话恢复命令
- workflow/session/start.md - 会话启动命令

保持与其他命令文档的格式一致性

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-05 21:49:03 +08:00
catlog22
967e3805b7 docs: 标准化命令文档格式，移除表情符号并添加命令模板规范 
更新所有命令文档以提高可读性和一致性：
- 移除所有表情符号（⚠️, ✅, ❌, ▸等），使用纯文本替代
- 统一标题格式，改进章节结构
- 简化状态指示器和格式标记
- 添加三个新的命令模板规范文档

新增文档：
- COMMAND_FLOW_STANDARD.md - 标准命令流程规范
- COMMAND_TEMPLATE_EXECUTOR.md - 执行器命令模板
- COMMAND_TEMPLATE_ORCHESTRATOR.md - 编排器命令模板

影响范围：
- CLI命令（cli-init, codex-execute, discuss-plan, execute）
- 内存管理命令（skill-memory, tech-research, workflow-skill-memory）
- 任务管理命令（breakdown, create, execute, replan）
- 工作流命令（所有workflow相关命令）

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-05 21:48:43 +08:00
catlog22
a38a216cc9 docs: 移除完成工作流会话的SKILL包更新选项，简化文档并优化流程描述 2025-11-05 19:59:57 +08:00
catlog22
3b351693fc refactor: 重构UI设计工作流的交互模式和字段架构 
主要变更：

1. animation-extract.md 重构交互流程
   - Phase 2 交互从Agent移至主流程（参考artifacts.md模式）
   - 使用【问题N - 标签】格式，提供方向性指导而非具体示例
   - Phase 3 Agent改为纯合成任务（无用户交互）
   - 保留独立的animation-tokens.json输出

2. style-extract.md 字段扩展优化
   - 保留字段扩展：typography.combinations、opacity、component_styles、border_radius、shadows
   - 移除animations字段（由专门的animation-tokens.json提供）
   - 完善design-tokens.json格式文档

3. generate.md 支持新字段
   - 保留animation-tokens.json可选加载
   - 支持typography.combinations、opacity、component_styles
   - 生成对应CSS类（组件样式类、排版预设类）

4. batch-generate.md 字段支持
   - 解析新的design-tokens字段结构
   - 使用组件预设和排版预设

架构改进：
- 清晰分离：design-tokens.json（基础设计系统）+ animation-tokens.json（动画配置）
- 主流程交互：用户决策在主流程完成，Agent只负责合成
- 字段完整性：保留所有设计系统扩展字段

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-11-05 19:50:47 +08:00
catlog22
15bf9cdbed Refactor workflow session archival and SKILL package updates 
- Updated command invocation for SKILL memory generator to use session ID instead of incremental mode.
- Enhanced documentation on the processing of archived sessions and intelligent aggregation by the agent.
- Added templates for generating conflict patterns, lessons learned, SKILL index, and sessions timeline.
- Established clear update strategies for incremental and full modes across all new templates.
- Improved structure and formatting rules for better clarity and usability in generated documents.
 2025-11-04 21:46:37 +08:00
catlog22
779581ec3b Add workflow-skill-memory command and skill aggregation prompt 
- Implemented the workflow-skill-memory command for generating SKILL packages from archived workflow sessions.
- Defined a 4-phase execution process for reading sessions, extracting data, organizing information, and generating SKILL files.
- Created a detailed prompt for skill aggregation, outlining tasks for analyzing archived sessions, aggregating lessons learned, conflict patterns, and implementation summaries.
- Established output formats for aggregated lessons, conflict patterns, and implementation summaries to ensure structured and actionable insights.
 2025-11-04 21:34:36 +08:00
catlog22
6326fbf2fb refactor: consolidate agent architecture and archive legacy templates 
- Remove general-purpose agent in favor of universal-executor
- Enhance workflow session completion with better state management
- Improve context-gather with advanced filtering and validation
- Archive legacy prompt templates for reference

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-27 15:00:00 +08:00
catlog22
367040037a refactor: migrate prompt templates to standardized structure and enhance CLI command documentation 
Template Migration:
- Move templates from .claude/prompt-templates/ to .claude/workflows/cli-templates/prompts/
- Rename and reorganize: bug-fix.md → development/bug-diagnosis.txt
- Rename and reorganize: code-analysis.md → analysis/code-execution-tracing.txt
- Rename and reorganize: plan.md → planning/architecture-planning.txt

CLI Command Enhancements:
- Add clear tool selection hierarchy (gemini primary, qwen fallback, codex alternative)
- Enhance analyze.md, chat.md with tool descriptions and agent context
- Enhance mode/code-analysis.md, mode/bug-diagnosis.md, mode/plan.md with Task() wrapper
- Add all necessary codex parameters (--skip-git-repo-check -s danger-full-access)
- Simplify descriptions while preserving core functionality

Agent Updates:
- Streamline cli-execution-agent.md (600→250 lines, -60%)
- Add complete templates reference for standalone usage
- Remove dependency on intelligent-tools-strategy.md

Reference Updates:
- Update test-task-generate.md template path references
- Delete duplicate bug-index.md
- All template paths now use ~/.claude/workflows/cli-templates/prompts/ format

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-27 10:26:35 +08:00
catlog22
2b52eae3f8 refactor: enhance test-context-gather and test-context-search-agent for improved context collection and coverage analysis 2025-10-25 19:13:37 +08:00
catlog22
bb6f74f44b refactor: enhance conflict resolution command with modification suggestions for custom handling 2025-10-25 17:03:05 +08:00
catlog22
986eb31c03 refactor: enhance context-search-agent execution mode and session information structure 2025-10-25 15:47:30 +08:00
catlog22
4f0edb27ff refactor: update subagent type to context-search-agent in load and context-gather commands 2025-10-25 15:32:33 +08:00
catlog22
3e83f77304 refactor: remove redundant context-package validation and logging from context-gather command 2025-10-25 15:29:36 +08:00
catlog22
18d369e871 refactor: update context-gather command to utilize context-search-agent for project context collection 2025-10-25 15:26:30 +08:00
catlog22
c363b5dd0e refactor: remove redundant references to workflow architecture in documentation 2025-10-25 14:48:55 +08:00
catlog22
692a68da6f Refactor workflow tools and user interaction methods 
- Updated synthesis tool to enhance user interaction with multi-select options and improved question presentation in Chinese.
- Revised conflict resolution tool to allow batch processing of conflicts, increasing the limit from 4 to 10 per round and changing user interaction from AskUserQuestion to text output.
- Added context_package_path to task generation tools for better context management.
- Improved task generation schema to include context_package_path for enhanced context delivery.
- Updated CLI templates to reflect changes in task JSON schema, ensuring context_package_path is included.
 2025-10-25 14:43:55 +08:00
catlog22
89f22ec3cf refactor: update file naming conventions and restrictions for analysis outputs in multiple agents 2025-10-25 13:15:14 +08:00
catlog22
b7db6c86bd refactor(synthesis): remove CLI concept enhancement references and streamline analysis agent execution 2025-10-24 22:50:56 +08:00
catlog22
71138a95e1 refactor: add Windows path format guidelines to multiple agent documents 2025-10-24 22:46:49 +08:00
catlog22
ecccae1664 refactor(auto-parallel): add memory check for selected roles to optimize file reading 2025-10-24 22:39:24 +08:00
catlog22
642d25f161 refactor(artifacts): add option limit for role selection in AskUserQuestion 2025-10-24 22:35:36 +08:00
catlog22
20d53bbd8e refactor(artifacts, auto-parallel): enhance task tracking and execution phases with detailed user interaction and metadata management 2025-10-24 21:46:14 +08:00
catlog22
9a63512256 refactor(artifacts): clarify role selection process and emphasize user interaction requirements 2025-10-24 20:57:36 +08:00
catlog22
080c8be87f refactor(auto-parallel): delegate role selection to artifacts command for interactive user experience 2025-10-24 20:32:24 +08:00
catlog22
a208af22af refactor(brainstorm): enhance artifacts role selection with intelligent recommendations 
- Remove hardcoded keyword-to-role mappings
- Implement intelligent role recommendation based on topic analysis
- Recommend count+2 roles for user multiSelect (default: 3+2=5 options)
- Add --count parameter support for flexible role selection
- Limit role questions to 3-4 per role (AskUserQuestion constraint)
- Add batching for Phase 4 conflict questions (max 4 per round)
- Provide complete list of 9 available roles with Chinese names
- Add clear rationale requirement for each role recommendation

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 19:35:55 +08:00
catlog22
7f82d0da86 refactor(conflict-resolution): transform to interactive resolution with in-place modifications 
BREAKING CHANGE: Remove CONFLICT_RESOLUTION.md generation in favor of interactive user confirmation and direct file modifications

Core Changes:
- Replace markdown report generation with structured JSON output for programmatic processing
- Add interactive conflict resolution via AskUserQuestion (max 4 conflicts, 2-4 strategies each)
- Apply modifications directly to guidance-specification.md and role analyses (*.md) using Edit tool
- Update context-package.json with conflict_risk status (resolved/none/low)
- Remove Phase 3 output validation (no file generation needed)

Modified Files:
- conflict-resolution.md: Complete rewrite of agent prompt and execution flow
  - Step 4: JSON output instead of markdown generation
  - Phase 3: User confirmation via AskUserQuestion
  - Phase 4: Apply modifications using Edit tool
  - Success criteria updated for in-place modifications
- plan.md: Update Phase 3 data flow and TodoWrite pattern
  - Data flow now shows "Apply modifications via Edit tool"
  - Todo description changed to "Resolve conflicts and apply fixes"
- task-generate-agent.md: Update conflict resolution context description
  - No longer references CONFLICT_RESOLUTION.md file
  - Notes conflicts resolved in guidance-specification.md and role analyses
- task-generate.md: Comprehensive cleanup of all CONFLICT_RESOLUTION.md references
  - Remove CONFLICT_RESOLUTION.md from artifact catalog
  - Update load_planning_context step to read guidance-specification.md
  - Update task implementation logic_flow
  - Update artifact priority and integration sections
  - Update directory structure documentation

Benefits:
- Seamless workflow: conflicts detected → user confirms → applied automatically
- No intermediate files to manage
- User interaction at decision point (not after-the-fact)
- Resolved conflicts integrated directly into source artifacts
- Clear conflict_risk status tracking in context-package.json

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 17:31:19 +08:00
catlog22
2b3541941e fix(conflict-resolution): fix markdown formatting errors 
Fix formatting issues:
- Line 104-117: Replace escaped backticks (\`\`\`) with proper markdown code fence (```)
- Line 158: Remove unnecessary backslash escapes from inline code
- Line 182: Fix stray backslash before backtick in prompt template

These formatting errors were causing markdown rendering issues.

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 17:07:10 +08:00
catlog22
04373ee368 feat(brainstorm): enforce Chinese language for all user questions 
Add mandatory Chinese language rule for AskUserQuestion in artifacts.md and synthesis.md:
- Phase 3 Rules: Questions MUST be asked in Chinese (用中文提问)
- Question Generation Guidelines: ALL questions MUST be in Chinese (所有问题必须用中文)
- Phase 4 (synthesis.md): ALL AskUserQuestion calls MUST use Chinese

Rationale: Improve user understanding and interaction experience for Chinese-speaking users

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 17:05:51 +08:00
catlog22
4dd1ae5a9e refactor(artifacts): enhance question depth and reduce file size 
Major improvements:
- Reduce from 715 to 254 lines (-64.5%) while maintaining core functionality
- Transform from template-based to dynamic question generation
- Phase 1: Add deep topic analysis (root challenges, trade-offs, success metrics)
- Phase 3: Probe implementation depth, trade-offs, and edge cases
- Phase 4: Conflict detection based on actual Phase 3 answers (not static matrix)
- Remove redundant examples, consolidate duplicate content
- Add Anti-Pattern examples and Quality Rules for dynamic generation

Key changes based on Gemini analysis:
- Extract keywords/challenges from topic → Generate task-specific questions
- Map challenges to role expertise → Probe implementation decisions
- Detect actual conflicts in answers → Generate resolution questions

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 17:02:13 +08:00
catlog22
acc792907c docs: refine artifacts command documentation for clarity and structure 2025-10-24 16:35:09 +08:00
catlog22
c3d05826ef docs: update file paths in auto-parallel command documentation for clarity 2025-10-24 16:05:41 +08:00
catlog22
da908d8db4 refactor: remove MCP code-index dependency, replace with ripgrep/find 
Replace all mcp__code-index__ calls with native ripgrep and find commands
across workflow and command files for better performance and portability.

Changes:
- Remove 41 mcp__code-index__ function calls from 12 files
- Replace with ripgrep (rg) for content search
- Replace with find for file discovery
- Remove index refresh dependencies (no longer needed)

Modified files:
- workflow/tools: context-gather, test-context-gather, task-generate-agent,
  task-generate, test-task-generate (core workflow tools)
- workflow: review (security scanning)
- memory: load, update-related, docs (memory management)
- cli/mode: plan, bug-index, code-analysis (CLI modes)

Documentation updates:
- Simplify mcp-tool-strategy.md to only Exa usage (5 lines)
- Streamline context-search-strategy.md to 69 lines
- Standardize codebase-retrieval syntax per intelligent-tools-strategy.md

Benefits:
- Faster search with ripgrep (no index overhead)
- Better cross-platform compatibility
- Simpler configuration (fewer MCP dependencies)
- -232 lines of code removed

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-24 15:45:26 +08:00
catlog22
3068c2ca83 Refactor TDD Workflow: Update phases, introduce conflict resolution, and enhance task generation 
- Revised TDD workflow to reduce phases from 7 to 6, integrating conflict resolution as an optional phase.
- Updated phase descriptions and execution logic to ensure automatic progression through phases based on TodoList status.
- Removed the concept-enhanced command and its associated documentation, streamlining the analysis process.
- Enhanced task generation to prioritize conflict resolution strategies and incorporate context package loading.
- Updated UI design documentation to reflect changes in role analysis and design system references.
- Improved error handling and validation checks across various commands to ensure robustness in execution.
 2025-10-24 15:08:16 +08:00
catlog22
ee7ffdae67 docs: enhance workflow documentation with role analysis and conflict resolution details 2025-10-24 11:56:50 +08:00
catlog22
1f070638b4 docs: update workflow plan to enhance automation and clarify quality gate process 2025-10-24 11:14:34 +08:00
catlog22
57fa379e45 Refactor workflow to replace synthesis-specification.md with role analysis documents 
- Updated references in various workflow commands to utilize role analysis documents instead of synthesis-specification.md.
- Modified CLI templates and command references to reflect the new architecture and document structure.
- Introduced conflict-resolution command to analyze and resolve conflicts between implementation plans and existing codebase.
- Deprecated synthesis role template and provided migration guidance for transitioning to the new role analysis approach.
 2025-10-24 11:08:15 +08:00
catlog22
047d809e23 docs: rename general-purpose agent to universal-executor and update documentation 
Major updates in this release:

1. Agent Renaming (13 files, 21 references):
   - Renamed general-purpose → universal-executor to avoid naming conflicts
   - Updated all references in commands and workflows
   - Maintained backward compatibility in documentation

2. README Updates (4 files):
   - Removed /workflow:session:start step (auto-created by /workflow:plan)
   - Simplified workflow from 4 steps to 3 steps
   - Updated version to v4.6.2
   - Added CLI tool usage guidelines

3. GETTING_STARTED Enhancements (2 files):
   - Added Design Philosophy section explaining multi-model CLI integration
   - Added comprehensive CLI tool usage guide with common workflows
   - Reorganized quick start to emphasize automatic session creation
   - Added examples for bug fixes and feature development

Files modified:
- Agent config: .claude/agents/general-purpose.md
- Commands: 7 files in .claude/commands/
- Workflows: 5 files in .claude/workflows/
- Documentation: README.md, README_CN.md, GETTING_STARTED.md, GETTING_STARTED_CN.md

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-23 21:18:36 +08:00
catlog22
69a654170a docs: 更新 test-cycle-execute 和 test-fix-gen 文档，增强命令范围和责任边界的描述，明确测试失败处理流程 2025-10-23 17:50:47 +08:00
catlog22
12d010c1d8 docs: 更新文档，支持多个分析文件的动态发现和加载，增强任务生成和合成文档的灵活性 2025-10-23 09:48:32 +08:00
catlog22
d9cee7f17a docs: 更新 auto-parallel 文档，优化任务代理调用模板，增强用户意图对齐和执行指令的描述 2025-10-22 23:37:15 +08:00
catlog22
598efea8f6 docs: 更新 task-generate 文档，增强上下文智能和技术指导的描述，优化任务执行优先级和依赖关系 2025-10-22 23:03:52 +08:00
catlog22
8b8c2e1208 docs: 更新 action-plan-verify 文档，优化 TodoWrite 任务跟踪和修复工作流程 2025-10-22 22:31:06 +08:00
catlog22
d3f8d012a1 docs: 更新文档，增强CLI工具使用说明，明确命令行参数和上下文组装指南 2025-10-22 22:17:00 +08:00
catlog22
83664cb777 feat: migrate to Gemini CLI v0.11.0-nightly with native prompt support 
## Major Changes

### Gemini CLI Integration (google-gemini/gemini-cli#11228)
- Migrate from wrapper scripts to native Gemini CLI v0.11.0-nightly
- Remove `-p` flag requirement for prompt strings
- Deprecate `gemini-wrapper` and `qwen-wrapper` scripts
- Update all commands and workflows to use direct CLI syntax

### Command Syntax Updates
- **Before**: `gemini -p "CONTEXT: @**/* prompt"`
- **After**: `gemini "CONTEXT: @**/* prompt"`
- Apply to all 70+ command files and workflow templates
- Maintain backward compatibility for Qwen fallback

### File Pattern Migration
- Replace `@{CLAUDE.md}` with `@CLAUDE.md`
- Replace `@{**/*}` with `@**/*`
- Update all file references to use direct @ notation
- Remove legacy brace syntax across all documentation

### Documentation Improvements
- Reorganize `intelligent-tools-strategy.md` structure
- Add Quick Start section with decision matrix
- Enhance `--include-directories` best practices
- Add comprehensive command templates and examples
- Improve navigation with clearer section hierarchy

### Files Modified (75+ files)
- Commands: All CLI commands updated (cli/, workflow/, task/, memory/)
- Workflows: Core strategy files and templates
- Agents: CLI execution agent and doc generator
- Templates: Planning roles and tech stack guides

## Breaking Changes
- Gemini CLI v0.11.0-nightly or later required
- Old wrapper scripts no longer supported
- Legacy `@{pattern}` syntax deprecated

## Migration Guide
Users should:
1. Update Gemini CLI to v0.11.0-nightly or later
2. Remove `-p` flag from existing commands
3. Update file patterns from `@{pattern}` to `@pattern`

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-21 14:46:16 +08:00
catlog22
f8d5d908ea feat: 修复workflow用户意图传递链，确保原始提示词贯穿全流程 
## 核心问题
根据Gemini深度分析，用户原始提示词在brainstorm→plan流程中逐步被稀释，
synthesis环节完全不接收用户提示词，导致最终产出偏离用户意图。

## 关键修改

### 1. synthesis.md - 修复最严重偏差环节
- 新增从workflow-session.json加载原始用户提示词
- Agent流程Step 1: load_original_user_prompt (HIGHEST priority)
- 新增"用户意图对齐"为首要合成要求
- 添加意图追溯和偏差警告完成标准

### 2. concept-clarify.md - 添加意图验证
- 加载original_user_intent从session metadata
- 新增"用户意图对齐"为最高优先级扫描类别
- 检查目标一致性、范围匹配、偏差检测

### 3. action-plan-verify.md - 强化验证门禁
- workflow-session.json作为主要参考源
- 新增"用户意图对齐"为CRITICAL级别检测
- 违反用户意图标记为CRITICAL严重性

### 4. plan.md - 确立意图优先级
- 明确用户原始意图始终为主要参考
- 优先级规则: 当前用户提示词 > synthesis > topic-framework

### 5. artifacts.md - 推广结构化格式
- 添加推荐的GOAL/SCOPE/CONTEXT结构化格式
- 强调用户意图保存的重要性

### 6. auto-parallel.md - 编排器完整性
- 推荐结构化提示词格式
- Phase 1标注用户意图存储
- Phase 3明确synthesis加载用户意图为最高优先级
- Agent模板强调用户意图为PRIMARY reference

## 影响力提升
用户提示词影响力: 30% → 95%

## 参考
分析报告由Gemini生成，识别了5个关键问题点并提供4条改进建议

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>
 2025-10-19 20:25:45 +08:00
catlog22
b8935777e7 refactor: enhance prompt generation process with direct output and improved internal analysis 2025-10-18 22:33:29 +08:00