**Core Changes**:
- Shift from heavy analysis in planning to targeted analysis in execution
- Reduce main process context by collecting only metadata (paths, structure)
- Simplify pre_analysis steps to be more declarative
**docs.md optimizations**:
- Phase 2: Lightweight metadata collection (get_modules_by_depth.sh + Code Index MCP)
- Planning phase now collects only paths and structure, not file content
- IMPL-001: Simplified to 2 pre_analysis steps (structure + tech stack from config files only)
- IMPL-002: Removed system_context loading, single focused analysis step
- IMPL-N (API): Replace rg with mcp__code-index__search_code_advanced for better structure
**doc-generator.md optimizations**:
- Add "Scope-Limited Analysis" philosophy
- Add "Optimized Execution Model" section
- Emphasize analysis is always limited to focus_paths to prevent context explosion
**Benefits**:
- Main process context stays minimal (no risk of overflow)
- Each task has controlled, scoped context
- Better separation: planning = metadata, execution = deep analysis
- More maintainable pre_analysis steps (less nested commands)
- Better use of MCP tools for structured data
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Implemented style-extract command to extract design styles from reference images using Gemini Vision and Codex.
- Added execution protocol detailing session validation, image analysis, and structured token generation.
- Introduced ui-generate command to create UI prototypes based on consolidated design tokens and synthesis specifications.
- Defined comprehensive execution steps for loading design systems, generating HTML/CSS prototypes, and ensuring accessibility compliance.
- Established a design-tokens-schema for consistent design token definitions across workflows, emphasizing OKLCH color format and semantic naming.
- Update doc-generator agent description to emphasize intelligent autonomous execution
- Add on_error field to all flow_control pre_analysis steps
- Clarify agent role as goal-oriented worker vs script runner
- Improve error handling documentation in workflow examples
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updated `/workflow:docs` command to enhance documentation planning and orchestration.
- Introduced new parameters for tool selection and scope filtering.
- Separated documentation planning from content generation, clarifying roles of docs.md and doc-generator.md.
- Added detailed planning process phases, including session initialization, project structure analysis, and task decomposition.
- Created new templates for API, module, and project-level documentation, ensuring comprehensive coverage and consistency.
- Enhanced output structure for better organization of generated documentation.
- Implemented error handling and completion checklist for improved user experience.
- Create test-context-gather.md for MCP-based coverage analysis
- Analyze test coverage gaps using MCP code-index tools
- Identify files requiring tests
- Detect test framework and conventions
- Generate test-context-package.json with coverage data
- Create test-concept-enhanced.md for Gemini-powered test analysis
- Analyze coverage gaps and implementation context
- Study existing test patterns and conventions
- Generate comprehensive test requirements
- Design test generation strategy
- Output TEST_ANALYSIS_RESULTS.md with test specifications
- Update test-task-generate.md to generate two tasks
- IMPL-001: Test generation task calling @code-developer
- IMPL-002: Test execution and fix cycle task calling @test-fix-agent
- Use TEST_ANALYSIS_RESULTS.md for test generation requirements
- Configure Gemini diagnosis and Codex resume mechanism for fixes
- Update test-gen.md to 5-phase workflow
- Phase 1: Create test session
- Phase 2: Gather test context (test-context-gather)
- Phase 3: Analyze test requirements (test-concept-enhanced)
- Phase 4: Generate test tasks (test-task-generate)
- Phase 5: Return summary
- Simplify documentation by removing redundant content
- Maintain all original intent while improving clarity
Complete workflow: Coverage analysis → Gemini analysis → Test generation → Test execution → Fix cycle
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Replace sequential execution model with explicit auto-continue pattern
- Document that each phase returns to user before auto-continuing
- Use TodoList state for progress tracking (no external state files)
- Store parsed data (sessionId, contextPath) in memory between phases
- Clarify error handling keeps phase in_progress without proceeding
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Major redesign of test-gen workflow based on simplified architecture:
Core Changes:
- Creates independent test session (WFS-test-[source]) instead of modifying existing session
- Implements 4-phase execution with concept-enhanced analysis integration
- Reuses IMPL-*.json format with meta.type="test-fix" (no new task type)
- Auto-detects cross-session context via workflow-session.json metadata
4-Phase Workflow:
1. Phase 1: Create test session with workflow_type and source_session_id metadata
2. Phase 2: Auto-detect and gather cross-session context (summaries, git changes, tests)
3. Phase 3: Analyze implementation with concept-enhanced (parallel Gemini + Codex)
4. Phase 4: Generate IMPL-001.json task with test strategy from ANALYSIS_RESULTS.md
Key Features:
- Session isolation: Test session separate from implementation session
- Parameter simplification: context-gather auto-detects via metadata (no --source-session param)
- Format reuse: IMPL-*.json with meta.type field (zero changes to execute/status tools)
- Analysis integration: concept-enhanced provides comprehensive test strategy and code targets
- Automatic cross-session: Collects source session artifacts automatically
Tool Modifications Required:
- session-start: Support test session metadata (workflow_type, source_session_id)
- context-gather: Auto-detect test session and gather from source session
- task-generate: Extract test strategy from ANALYSIS_RESULTS.md
- concept-enhanced: No changes (already supports analysis)
- execute: No changes (already dispatches by meta.agent)
Data Flow:
User → session-start (creates test session) → context-gather (auto cross-session)
→ concept-enhanced (analyzes + generates strategy) → task-generate (creates task)
→ Returns summary → User triggers /workflow:execute separately
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Major role restructuring to improve workflow efficiency and clarity:
## New Roles Added (4)
- product-owner: Backlog management and user story definition
- scrum-master: Sprint planning and agile process facilitation
- ux-expert: User experience optimization and usability testing
- subject-matter-expert: Domain expertise and industry standards
## Roles Removed (5)
- business-analyst → functionality split to product-owner and scrum-master
- feature-planner → merged into product-owner responsibilities
- innovation-lead → integrated into subject-matter-expert
- security-expert → integrated into subject-matter-expert
- user-researcher → merged into ux-expert
## Files Updated
### Command Files (.claude/commands/workflow/brainstorm/)
- Created: product-owner.md, scrum-master.md, ux-expert.md, subject-matter-expert.md
- Deleted: business-analyst.md, feature-planner.md, innovation-lead.md, security-expert.md, user-researcher.md
- Updated: artifacts.md, auto-parallel.md, auto-squeeze.md, synthesis.md
### Planning Templates (.claude/workflows/cli-templates/planning-roles/)
- Created: product-owner.md, scrum-master.md, ux-expert.md, subject-matter-expert.md
- Archived: Moved 5 deprecated roles to _deprecated/ with migration guide
- Added: _deprecated/README.md with deprecation details and migration paths
### Agent Configurations
- Updated conceptual-planning-agent.md with new role mappings
- Updated action-planning-agent.md with new role references
### Documentation
- Updated README.md brainstorming role tables and descriptions
- Updated README_CN.md with Chinese translations for new roles
- Updated workflow documentation files with new role references
## Breaking Changes
Commands for removed roles are no longer available. Use replacement roles:
- /workflow:brainstorm:business-analyst → use product-owner or scrum-master
- /workflow:brainstorm:feature-planner → use product-owner
- /workflow:brainstorm:innovation-lead → use subject-matter-expert
- /workflow:brainstorm:security-expert → use subject-matter-expert
- /workflow:brainstorm:user-researcher → use ux-expert
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Updated the `/workflow:plan` command description to clarify its orchestration of a 4-phase planning workflow.
- Revised the execution flow and core planning principles for improved clarity and structure.
- Removed the `ANALYSIS_RESULTS.md` file as it is no longer needed in the workflow.
- Enhanced the `concept-enhanced` tool documentation to specify mandatory first steps and output requirements.
- Expanded the `ui-designer` role template to include detailed design workflows, output requirements, and collaboration strategies.
- Introduced new design phases with clear outputs and user approval checkpoints in the UI designer template.
- Implemented task-generate-agent for autonomous task generation using action-planning-agent with discovery and output phases.
- Introduced task-generate command to generate task JSON files and IMPL_PLAN.md from analysis results with automatic artifact detection and integration.
- Enhanced documentation for both commands, detailing execution lifecycle, phases, and output structures.
- Established clear integration points and error handling for improved user experience.
- Updated intelligent-tools-strategy.md to include `--skip-git-repo-check` for Codex write access and development commands.
- Improved context gathering and analysis processes in mcp-tool-strategy.md with additional examples and guidelines for file searching.
- Introduced new command concept-enhanced.md for enhanced intelligent analysis with parallel CLI execution and design blueprint generation.
- Added context-gather.md command for intelligent collection of project context based on task descriptions, generating standardized JSON context packages.
- Implemented `/workflow:plan` for creating detailed implementation plans with task decomposition and context gathering.
- Added `/workflow:resume` for intelligent session resumption with automatic progress analysis.
- Introduced `/workflow:review` for executing the final phase of quality validation and generating review reports.
- Developed `/workflow:status` to provide on-demand views of workflow status and task progress.
- Created `/workflow:test-gen` to generate comprehensive test workflows based on completed implementation tasks, ensuring full test coverage.
- Introduced `/workflow:test-gen` command to automate test workflow generation based on completed implementation tasks, including detailed lifecycle phases, task decomposition, and agent assignment.
- Implemented `/workflow:concept-eval` command for pre-planning evaluation of concepts, assessing feasibility, risks, and optimization recommendations using strategic and technical analysis tools.
- Added `/workflow:docs` command for generating hierarchical architecture and API documentation, with structured task creation and session management.
- Developed `/workflow:status` command to provide on-demand views of workflow state, supporting multiple formats and validation checks for task integrity and relationships.
- Removed deprecated `gemini-init.md` command and migrated its functionality to a new structure under `.claude/commands/gemini/`.
- Introduced `auto-parallel.md` for parallel brainstorming automation with dynamic role selection and concurrent execution.
- Added `auto-squeeze.md` for sequential command coordination in brainstorming workflows, ensuring structured execution from framework generation to synthesis.
- Updated `plan.md` to improve clarity on project structure analysis and command execution strategies.
- Enhanced error handling and session management across all commands to ensure robustness and user guidance.
- Improved documentation for generated files and command options to facilitate user understanding and usage.
- Add "Task Merging Over Decomposition" as core principle
- Define concrete decomposition criteria (2500 lines or 6+ files)
- Standardize error messages to English for consistency
- Improve workflow planning efficiency by reducing unnecessary task splits
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Move MCP Tools Integration from prominent position to Task JSON Schema section
- Simplify complex command examples by breaking into multiple separate commands
- Remove repository URLs and make MCP integration information more concise
- Improve readability by avoiding complex bash conditionals in examples
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
## New Features
- **MCP Tools Integration**: Added support for Model Context Protocol tools
- Exa MCP Server: External API patterns and best practices
- Code Index MCP: Advanced internal codebase search and indexing
- **Enhanced Workflow Planning**: Updated pre_analysis to include MCP tool steps
- **Documentation Updates**: Added MCP tool setup guides and usage examples
## Changes
### Core Components
- Updated `plan.md` with MCP integration principles and implementation approach guidelines
- Added MCP tool steps in pre_analysis workflow: `mcp_codebase_exploration`, `mcp_external_context`
- Enhanced context accumulation with external best practices lookup
### Documentation
- Added comprehensive MCP tools section in both English and Chinese README
- Updated installation requirements and integration guidelines
- Added GitHub repository links for required MCP servers
### Agent Enhancements
- Updated multiple agents to support MCP tool integration
- Enhanced context gathering capabilities with external pattern analysis
## Technical Details
- MCP tools provide faster analysis through direct codebase indexing
- Automatic fallback to traditional bash/CLI tools when MCP unavailable
- Enhanced pattern recognition and similarity detection capabilities
🧪 **Experimental**: MCP integration is currently experimental and optional
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Update agent definitions with enhanced context and flow control
- Standardize command templates for consistent CLI tool integration
- Improve gemini-wrapper with better token management and path handling
- Simplify qwen-wrapper for streamlined execution
- Enhance intelligent-tools-strategy with directory navigation patterns
- Add proper quoting for Windows path compatibility in all commands
- Update workflow planning documentation with current architectural insights
These changes align the command system with the four-layer architecture
analysis and improve cross-platform compatibility.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add concept-eval.md command for concept evaluation workflow
- Add concept-eval.txt template for structured concept analysis
- Enhance plan.md with required IMPL_PLAN.md structure format
- Define standardized format for project planning documentation
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add optional analyze_implementation step using Codex CLI tool
- Include -C parameter for directory-focused development analysis
- Enhance pre_analysis flow with both Gemini and Codex options
- Improve agent context loading with implementation-focused patterns
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Create new workflow command to initialize Gemini CLI configuration
- Generates .gemini file with contextfilename set to "CLAUDE.md"
- Creates .geminiignore with technology-specific filtering rules
- Uses get_modules_by_depth.sh for workspace analysis and tech stack detection
- Supports preview mode and custom output paths
- Includes backup functionality for existing files
- Supports major tech stacks: Node.js, Python, Java, Docker, etc.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Fix execute.md: Clarify that pre_analysis steps are passed to agents for execution, not executed by workflow:execute command
- Fix plan.md: Add explicit note that pre_analysis steps are generated for agent execution
- Streamline verbose sections while preserving critical operational details
- Maintain consistency between plan and execute commands regarding pre_analysis handling
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Clarify that /workflow:plan creates plans but doesn't execute implementation.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Rename /workflow:verify to /workflow:plan-verify to better reflect its role in the workflow lifecycle between planning and execution phases.
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
Implement /workflow:test-gen command for generating comprehensive test workflows based on completed implementation tasks. Creates dedicated test sessions with full coverage planning including unit, integration, e2e, performance, security, and documentation tests.
Key features:
- Analyzes completed IMPL-* tasks to generate corresponding TEST-* tasks
- Follows established workflow architecture with 5-field JSON schema
- Integrates with existing TodoWrite progress tracking and agent orchestration
- Supports automatic session detection and manual session specification
- Maps implementation patterns to appropriate test strategies
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Create dedicated /workflow:brainstorm:artifacts command for topic discussion and decomposition
- Unify document structure between auto.md and artifacts.md commands
- Fix .brainstorming path references to use proper .workflow/WFS-[topic]/ structure
- Add discussion-summary.md and component-analysis.md as shared foundation documents
- Separate synthesis/ folder generation to dedicated synthesis command
- Update auto.md workflow to include Interactive Discussion → Topic Decomposition phases
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Create /workflow:verify command for plan validation before execution
- Implement gemini strategic analysis and codex technical analysis
- Add cross-validation framework with user approval workflow
- Enhance brainstorm auto.md with comprehensive pre_analysis configuration
- Include specialized verification prompt templates for structured analysis
🤖 Generated with [Claude Code](https://claude.ai/code)
Co-Authored-By: Claude <noreply@anthropic.com>
