Enable test generation for existing codebases without requiring prior
workflow session. Supports both session-based (WFS-xxx) and prompt-based
(text description or file path) inputs with automatic mode detection.
Key improvements:
- Prompt mode: Generate tests from feature description or requirements file
- Backward compatible: Existing session-based usage unchanged
- Flexible input: Accepts session ID, text description, or file path
- Use case: Testing existing APIs without workflow history
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Renamed /workflow:test-execute to /workflow:test-cycle-execute to better
communicate the iterative cycle nature of the command (test → diagnose → fix → retest).
Changes:
- Renamed test-execute.md to test-cycle-execute.md
- Updated command metadata and document title
- Updated 9 internal references within the command file
- Updated reference in test-fix-gen.md related commands section
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Add comprehensive skill system and batch replan functionality:
- New skill modules for Gemini, Qwen, Codex, and context search
- Enhanced task replan command with batch mode
- Improved action plan verification
- Complete skill documentation with examples
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Enhanced layout-extract and style-extract workflows with two-phase generation
process, allowing users to preview and select preferred options before full
generation.
Key changes:
- Split generation into two agent tasks:
* Task 1: Generate concept options with previews
* Task 2: Develop selected option into complete output
- Add Phase 1.5 interactive user selection:
* Present concept options with visual previews
* Capture user selection via AskUserQuestion tool
* Save selection to user-selection.json
- Layout workflow improvements:
* Generate layout concept options with ASCII wireframes
* User selects preferred structural approach
* Generate detailed templates only for selected concepts
- Style workflow improvements:
* Generate design direction options with color/typography previews
* User selects preferred design philosophy
* Generate complete design system only for selected direction
- Better user experience:
* Avoid generating unwanted variants
* Allow informed decisions before full generation
* Reduce wasted computation on unused variants
This change improves workflow efficiency and user control over design output.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Renamed test-gen to test-fix-gen and created new test-execute command for
dynamic iterative test-fix cycles. Improved task generation flexibility and
clarified data flow between planning and execution phases.
Key changes:
- Rename test-gen.md → test-fix-gen.md to reflect task generation purpose
- Add test-execute.md for dynamic test-fix workflow execution with CLI analysis
- Enhance test-fix-gen.md:
* Emphasize code-developer's understanding and analysis responsibilities
* Support flexible task count (minimum 2, expandable based on complexity)
* Add examples for multi-module test generation scenarios
- Clarify test-execute.md data flow:
* failure_context contains raw test output and error messages
* fix_strategy generated by CLI tool (Gemini/Qwen) analysis
* IMPL-001 (test generation) executes first, then IMPL-002 (test-fix cycle)
This improves workflow clarity and enables adaptive test-fix strategies for
complex projects requiring multiple test generation tasks.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Fixes critical P0 issue where animation-tokens.json wasn't consumed by the
generate command, breaking the value chain. The animation extraction system
now properly flows through: animation-extract → tokens → generate → prototypes.
Changes:
- Added animation-extract command with hybrid CSS extraction + interactive
fallback strategy
- Updated generate.md to load and inject animation tokens into prototypes
- Added CSS animation support (custom properties, keyframes, interactions,
accessibility)
- Integrated animation extraction into explore-auto and imitate-auto workflows
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
Automatically updated by linter to reflect new --count parameter
in argument hint for command usage clarity.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
**Problem**: synthesis.md incorrectly used action-planning-agent
- synthesis-specification.md is conceptual analysis document (WHAT to build)
- action-planning-agent is designed for execution plans (HOW to build)
- Mismatch between agent capabilities and task requirements
**Solution**: Change to conceptual-planning-agent
- Aligns with task nature (conceptual synthesis and multi-perspective integration)
- Consistent with other role analysis commands (data-architect, system-architect, etc.)
- Agent specializes in conceptual analysis rather than task decomposition
**Changes**:
- Update allowed-tools: action-planning-agent → conceptual-planning-agent
- Update description to note agent usage
- Add "Conceptual Focus" rationale explaining agent specialization
- Update all agent references throughout document (8 locations)
- Maintain consistency with role analysis command pattern
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add --count N parameter to specify number of roles (default: 3, max: 9)
- Add Parameter Parsing section with count handling logic
- Update role selection logic to use dynamic count value
- Update Usage section with parameter documentation
- Update TodoWrite examples to reflect dynamic role count
- Update parallel agent invocation examples with N roles
- Add role ranking and diversity selection guidelines
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Add --urls parameter support to style-extract and layout-extract
- Auto-trigger computed styles extraction when URLs provided (no manual flag)
- Auto-trigger DOM structure extraction for layout analysis
- Streamline imitate-auto workflow by removing redundant REPORT statements
- Maintain clear phase headers while reducing verbose output
- Enable graceful fallback to visual analysis if Chrome DevTools unavailable
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Reduce intelligent-tools-strategy.md from 504 to 325 lines (35.5% reduction)
- Merge Gemini/Qwen tool specifications into unified Analysis Tools section
- Consolidate command templates and remove duplicate examples
- Compress usage scenarios while preserving all critical information
- Simplify guidelines to concise bullet points
- Fix incorrect codex chat calls in discuss-plan.md to use proper exec format
All core decision rules, permission framework, and command templates preserved.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Update enhance-prompt.md to use correct file reference format
- Change from @~/.claude/workflows/intelligent-tools-strategy.md to ~/.claude/workflows/intelligent-tools-strategy.md
- Align with other command files' reference format
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Combine IMPL-005 (ARCHITECTURE.md) and IMPL-006 (EXAMPLES.md) into one task
- New IMPL-005: Generate both ARCHITECTURE.md and EXAMPLES.md together
- Reduces task count and simplifies workflow execution
- Both documents share similar analysis context from project README and modules
- Update HTTP API task from IMPL-007 to IMPL-006
- Maintain sequential task numbering after merge
- Update all references:
- Task hierarchy documentation
- Task generation logic
- Session structure examples
- Execution command examples
Benefits:
- Streamlined workflow with fewer tasks to execute
- Architecture and examples are logically related (structure + usage)
- Reduced overhead in task management and execution
- Fix get_modules_by_depth.sh to avoid filtering 'core' directories
- Remove overly broad "core" exclusion pattern
- Keep "*.core" for actual core dump files
- Resolves issue where directories named 'core' were incorrectly filtered
- Refactor docs workflow from path-based detection to mode selection
- Replace is_root logic with explicit --mode parameter
- Add --mode full: generate complete documentation (modules + project-level)
- Add --mode partial: generate module documentation only
- Default to full mode for better user experience
- Simplify configuration by removing complex path comparison logic
These changes provide better control over documentation generation
and fix directory discovery issues in Python projects with core/ folders.
- Updated implementation_approach in task-generate-tdd.md to use an array format with detailed steps for minimal code implementation and iterative testing.
- Enhanced task-generate.md to support CLI execution mode with Codex, including examples for task execution and session management.
- Modified test-task-generate.md to incorporate CLI execution mode, allowing Codex to autonomously generate tests and execute iterative test-fix cycles.
- Revised task-core.md and workflow-architecture.md to adopt a step-based implementation approach, improving clarity and dependency management in task execution.
- Improved documentation in workflow-architecture.md to clarify flow control structure and variable referencing.
- Introduced the discuss-plan command to facilitate iterative discussions among Gemini, Codex, and Claude.
- Defined roles and priorities for each model in the discussion loop.
- Established a structured workflow for input processing, discussion rounds, and user review.
- Implemented TodoWrite tracking for progress across rounds.
- Specified output routing for logs and final plans.
- Included examples and best practices for effective usage.
- Introduced a unified API documentation template covering both Code API and HTTP API sections.
- Created a folder navigation documentation template for directories containing subdirectories.
- Developed a module README documentation template focusing on module purpose, usage, and dependencies.
- Added a project architecture documentation template synthesizing module information and system design.
- Implemented a project examples documentation template showcasing practical usage scenarios.
- Established a project-level documentation template outlining project overview, architecture, and navigation.
**Summary:**
Updated all 62 command files in `.claude/commands` directory to improve parameter documentation clarity by replacing `examples` field with descriptive `argument-hint` field.
**Changes:**
- Added/improved `argument-hint` for all commands based on usage patterns
- Removed `examples` field and all example items from YAML headers
- Maintained all other YAML fields (name, description, usage, allowed-tools)
- Deleted obsolete commands: workflow/issue/*, workflow/session/pause.md, workflow/session/switch.md
- Cleaned up temporary analysis files
**Rationale:**
The `argument-hint` field provides clearer, more concise parameter documentation than example lists, improving command discoverability and usability in the Claude Code interface.
**Files Modified:** 62 command files
**Lines Changed:** -1570 insertions, +192 deletions
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Implemented the `/workflow:ui-design:explore-layers` command for depth-controlled UI capture.
- Defined depth levels from full-page screenshots to Shadow DOM exploration.
- Included setup, validation, and navigation phases with error handling.
- Captured screenshots at various depths and generated a comprehensive layer map.
- Added examples and usage instructions for ease of use.
- Introduced a new workflow named `imitate-auto-v3` that allows for batch screenshot capture and optional token refinement.
- Implemented a comprehensive 5-phase execution protocol to handle initialization, screenshot capture, style extraction, token processing, and UI generation.
- Added support for multiple target URLs through a `--url-map` parameter, enhancing flexibility in design replication.
- Included optional parameters for session management and token refinement, allowing users to customize their workflow experience.
- Enhanced error handling and reporting throughout the workflow to ensure clarity and ease of debugging.
- Deprecated the previous `imitate-auto.md` workflow in favor of this more robust version.
- Remove npx playwright to prevent auto-installation triggers
- Add explicit tool availability checks using 'which' command
- Check for playwright-cli, google-chrome, chrome, and chromium executables
- Only attempt screenshot capture if tools are already installed
- Provide clear tool availability reporting to users
- Gracefully fallback to manual mode when no tools available
- Support chromium browser on Linux systems
This ensures the workflow never triggers package installations and works
safely in environments without screenshot tools installed.
Add Phase 0.1/1.1 memory check to all UI design workflow commands to skip
execution when outputs already exist. This optimization dramatically reduces
redundant execution time from minutes to <1 second.
Changes:
- extract.md: Check for style-cards.json, skip Phase 0.5-3 if exists
- consolidate.md: Check for style-1/design-tokens.json, skip Phase 2-6
- generate.md: Check for compare.html, skip Phase 1.5-4
- update.md: Check for current design run in synthesis-spec, skip Phase 2-5
Benefits:
- >99% time reduction for repeated workflow execution
- Automatic cache detection without manual flags
- Clear reporting of skipped phases
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Agent now directly generates HTML with final CSS references (no placeholders)
- Remove tokens.css dependency - agents create self-contained CSS from design-tokens.json
- Simplify ui-generate-preview-v2.sh (no placeholder replacement logic)
- Update Phase 2.5 validation to check actual href attributes
- Remove Phase 1.6 (token conversion step)
- Improve agent instructions with direct CSS value usage from design tokens
Benefits:
- Simpler workflow with fewer intermediate steps
- More flexible CSS generation - agents can adapt token values based on design_attributes
- Better style differentiation across variants
- Reduced dependencies and potential error points
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Replace exa-code with web-search for diverse layout inspiration
- Simplify from keyword pools to direct "common variations" search
- Add selection strategy for agents to pick different structural patterns
- Support unlimited layout variants with natural diversity
Impact: Prevents layout architecture convergence by leveraging web search
for real-world layout variations instead of code examples.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
- Implemented a new script to generate compare.html and index.html for UI prototypes.
- Auto-detects styles, layouts, and targets from HTML file patterns.
- Generates an interactive comparison matrix and a simple index of prototypes.
- Includes a detailed PREVIEW.md guide with usage instructions and file structure.
Clarified file writing patterns across UI design workflows to prevent confusion
between "generate content as output then save" vs "direct write to file" approaches.
Changes:
- extract.md: Merged Phase 2 & 3 to unify style synthesis and file write operations
* Updated phase title to emphasize direct file writing
* Added explicit note about internal processing without context output
* Consolidated parse-and-write step into single atomic operation
- imitate-auto.md: Added clarifying comments for orchestrator file transformation
* Documented that Phase 2 uses orchestrator-level read-process-write pattern
* Explained this as performance optimization to bypass consolidate phase
The entire UI design workflow now consistently follows clear file operation patterns
with no ambiguous intermediate output steps.
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude <noreply@anthropic.com>
