By cyotee
Interactive design sessions for defining tasks in tasks/ directory with repo-prefixed IDs and dependency validation
npx claudepluginhub cyotee/cyotee-claude-plugins --plugin designInteractive design session to create a new task
Digest an existing design document into individual tasks
Create new tasks from code review suggestions
Initialize task management directory structure in a repository
Interactively create the global project PRD.md
Audit task definitions for quality and completeness (not code review - see /backlog:review for that)
Interactive design sessions for creating and refining tasks with user stories in the tasks/ directory structure.
/plugin marketplace add cyotee/cyotee-claude-plugins
/plugin install design@cyotee
/design <feature-description>Start an interactive design session to create a new task.
Aliases: /design:task <feature-description>
What it does:
Example usage:
/design Add a dark mode toggle to the settings page
Interactive process:
Round 1 - Scope & Purpose:
- What problem does this solve?
- Which layer does this belong to?
- What are the key user actions?
Round 2 - Technical Approach:
- What existing patterns should this follow?
- What dependencies does this have?
- What external systems does this interact with?
Round 3 - Implementation Details:
- What are the key design decisions?
- What are the acceptance criteria?
- What tests are needed?
Output: A task directory with:
tasks/[PREFIX]-[N]/PRD.md - Requirements and acceptance criteriatasks/[PREFIX]-[N]/PROGRESS.md - Agent progress logtasks/[PREFIX]-[N]/REVIEW.md - Review findingstasks/INDEX.md/design:initInitialize the tasks/ directory structure in a repository.
What it does:
Created structure:
tasks/
├── 0/ # Template directory
│ ├── PRD.md # Task requirements template
│ ├── PROGRESS.md # Progress log template
│ └── REVIEW.md # Review findings template
├── archive/ # Completed tasks moved here
└── INDEX.md # Task index/status overview
/design:prdInteractive PRD creation for project-level requirements.
What it does:
/design:reviewReview all tasks for quality and completeness.
What it does:
/design:review <task-id>Review and refine a specific task.
Example usage:
/design:review P-7
Review checklist:
Layers are detected dynamically:
tasks/INDEX.md for layer name and prefixTasks created by /design follow this structure:
---
task: N
title: [Title]
status: pending
layer: [Layer Name]
worktree: feature/[kebab-case-name]
created: YYYY-MM-DD
dependencies: []
---
# Task [PREFIX]-N: [Title]
## Description
[2-3 sentences explaining the feature]
## Dependencies
- [PREFIX]-X must complete first
- External system Y must be available
## User Stories
### US-N.1: [Story Title]
As a [role], I want to [action] so that [benefit].
**Acceptance Criteria:**
- [ ] Specific, testable criterion 1
- [ ] Specific, testable criterion 2
## Files to Create/Modify
**New Files:**
- `path/to/NewFile.sol` - Description
**Tests:**
- `test/path/Test.t.sol` - Description
## Inventory Check
- [ ] Prerequisite 1 exists
- [ ] Prerequisite 2 works
## Completion Criteria
- [ ] All user stories implemented
- [ ] Tests pass
- [ ] Documentation updated
| File | Purpose |
|---|---|
tasks/INDEX.md | Task index with status table |
tasks/[ID]/PRD.md | Task requirements |
PRD.md | Global project requirements |
CLAUDE.md | Project context |
1. /up:plan # See existing tasks
2. /design:init # Initialize tasks/ (first time)
3. /design <feature> # Create new task
4. /design:review # Refine tasks
5. /backlog:launch # Start working on a task
MIT
Comprehensive UI/UX design plugin for mobile (iOS, Android, React Native) and web applications with design systems, accessibility, and modern patterns
Share bugs, ideas, or general feedback.
Ultra-compressed communication mode. Cuts ~75% of tokens while keeping full technical accuracy by speaking like a caveman.
Standalone image generation plugin using Nano Banana MCP server. Generates and edits images, icons, diagrams, patterns, and visual assets via Gemini image models. No Gemini CLI dependency required.
AI image generation Creative Director powered by Google Gemini Nano Banana models. Claude interprets intent, selects domain expertise, constructs optimized prompts, and orchestrates Gemini for best results.