| name | dev-starter-prompts |
| description | Generates kick-start prompts for beginning development in any IDE. Use when transitioning from planning to coding or when starting a new project and needing structured initial prompts. |
| disable-model-invocation | true |
| model | sonnet |
| effort | medium |
| license | MIT |
Kick-Start Prompt Generation
Why This Matters
When you open a new project in an IDE, you are looking at a blank folder. If you are new to programming, it is hard to understand how to begin. Kick-start prompts are starting messages you send to your AI coding tool to set everything up.
The rule of thumb: do not overcomplicate them. Avoid deep technical details or code snippets. These prompts should be high-level references that point to your context documents – the files from your planning activities placed into the project folder.
Interaction Method
If AskUserQuestion is available, use it for all prompts below.
If not, present each question as a numbered list and wait for a reply before proceeding. For multiSelect questions, accept comma-separated numbers (e.g. 1, 3). Never skip or auto-answer without explicit user consent.
Step 0: Before starting
-
Announce your execution plan: Before doing anything, state what you will do in this activity: "Here's what I'm going to do: 1) check which planning documents you have ready, 2) determine which AI coding tool you'll use, 3) generate 3–5 focused kick-start prompts customized to your project, 4) review and customize them with you." This is a commitment device – harder to skip steps you just announced.
-
Conditional teaching: Ask the user if they are familiar with kick-start prompts and how they work. If yes, give a one-sentence refresher. If no, explain it in simple terms: these are the first messages you send to your AI coding tool to set up the project – high-level instructions that point to your context documents, not deep technical commands.
Required: ALWAYS ask the question, ALWAYS give the refresher when the user says yes. Never skip this step because the user "is a designer" or "already demonstrated familiarity earlier." Users want a memory refresh on every activity, including ones they know. Phrases like "I'll skip the explainer (you're a designer)" are forbidden — they signal the model has decided ON BEHALF OF the user that a refresher isn't needed. The user, not the model, decides what's redundant. The refresher takes one sentence; the cost is trivial; the value to a tired user mid-session is high.
-
Output presentation rule: Present output incrementally – one section at a time. After each section, discuss with the user, get their input, then move to the next. Never dump an entire deliverable at once.
-
Challenge ideas: After the user shares an idea or decision, challenge it – surface blind spots, edge cases, future implications. Then let the user decide with full perspective. This is not about being negative – it's about pressure-testing ideas so the user makes better decisions.
BLOCKING REQUIREMENT: Wait for the user to acknowledge the plan before proceeding to Step 1.
Step 1: Gather Available Context
question: "Which planning documents do you have ready?"
header: "Available Documents"
options:
- label: "MVP Requirements"
description: "Prioritized feature list with acceptance criteria"
- label: "Information Architecture"
description: "Screen inventory, navigation, user flows"
- label: "Design references or Figma designs"
description: "Visual direction for the product"
- label: "Business Plan / StoryBrand"
description: "Messaging, positioning, or revenue model"
- label: "Target Audience / Personas"
description: "User profiles and pain points"
- label: "CLAUDE.md already created"
description: "Global rules file is ready"
- label: "None yet"
description: "Starting without planning documents"
allowMultiSelect: true
multiSelect: true # User can have multiple document types ready
BLOCKING REQUIREMENT: Wait for the user's answer before proceeding to Step 2.
Step 2: Determine the AI Tool
question: "Which AI coding tool will you use?"
header: "Development Tool"
options:
- label: "Claude Code (terminal)"
description: "Anthropic's CLI tool – supports agents, skills, MCPs"
- label: "Cursor"
description: "AI-powered IDE with built-in code generation"
- label: "Claude Code inside Cursor's terminal"
description: "Best of both worlds – Cursor's IDE with Claude Code's capabilities"
- label: "Other IDE"
description: "Windsurf, Kiro, or another AI-assisted editor"
multiSelect: false # User must choose one development tool
BLOCKING REQUIREMENT: Wait for the user's answer before proceeding to Step 3.
Step 3: Generate the Prompts
If you save the generated prompts to a file under .design-engineer-plugin/design/dev/ (e.g., .design-engineer-plugin/design/dev/kickstart-prompts.md), ensure the parent directory exists first: run mkdir -p .design-engineer-plugin/design/dev (Bash). The plugin uses lazy folder scaffolding – folders are created by the skill that needs them, not upfront.
Using the template from kickstart-template.md and the methodology from dev-prep-assistant.md, generate a set of 3-5 focused prompts:
Prompt 1: Project Initialization
- Create folder structure
- Install dependencies using official documentation (never hardcoded commands)
- Set up configuration files
- Reference: point to relevant planning documents
Prompt 2: Foundation Setup
- Establish the design token system or styling foundation
- Set up navigation structure based on Information Architecture
- Create the app entry point
- Reference: point to Information Architecture document
Prompt 3: First Feature Implementation
- Implement the most critical feature from MVP Requirements
- Reference the design files for visual direction
- Specify that AI should ask clarifying questions before implementing
Prompt 4: Design System Extraction (after first feature)
- Refactor the first implementation to extract reusable tokens and components
- Create the design system structure from implemented code
- Establish naming conventions for future development
Prompt 5: Second Feature (with patterns)
- Implement the next priority feature reusing patterns from the first
- Validate that the design system is being followed
Step 4: Review and Customize
Present the generated prompts to the user for review. Each prompt should:
- Be one clear objective (not multiple goals)
- Reference existing documents by file path
- Specify behavioral expectations (ask questions first, wait for approval)
- Avoid code snippets or deep technical details
- Focus on outcomes, not implementation methods
Content Integrity
- No fabrication: Only include content the user explicitly provided or that was read from an existing deliverable file. If you see a gap – a missing document, an unaddressed workflow step – ask via AskUserQuestion. Never fill gaps silently. Never invent project details, file paths, or tech stack choices. Never attribute content to a deliverable you haven't Read.
- Read before reference: When referencing any previous deliverable in your output, you MUST Read the file first. Do not quote from memory – read the actual file and use its actual content.
Anti-slop Writing
Before generating any text for the deliverable, read anti-slop-writing.md and apply its rules. Scan your output before presenting it to the user.
Decision Hierarchy
- User's direct input – their tool preference, their project priorities
- Existing documentation – prompts reference real documents, not assumptions
- AI suggestions – propose prompt structure and sequence
What Comes Next
After kick-start prompts are generated, the user begins development. Suggest dev-status-tracking to set up status tracking, or dev-agent-setup if they want automated agent workflows.
Resource Files
