AI Product Design Toolkit - 126 Claude Prompts + Method by M M AzazAI Product Design Toolkit - 126 Claude Prompts + Method | Contra
Use AI to move faster, without inventing the evidence.
Claude can draft a persona, a PRD or a journey map in a minute, and all of them will read as plausible. The AI Product Design Toolkit is a working method that gets you that speed while keeping AI guesses clearly separate from what your users actually showed you.
What’s inside
Playbook: 8 chapters covering six modules, from Product Discovery through UX Research, Strategy, UX Architecture and UI Design to Validation & Handoff. Each module covers when to use AI, when not to, the workflow, quality checklists and common mistakes.
126 prompts, each built in seven parts (role, context, objective, inputs, constraints, output format, quality bar) and linked to the prompts before and after it. You can find them by module, by deliverable, or by what you already have.
11 editable templates that match the prompts’ outputs: PRD, research plan, interview guide, synthesis, user journey, component spec, handoff spec, UX audit and more.
Claude workflows: Project instructions to paste in, a guide to which files to upload, 7 step-by-step runbooks, and 3 optional Claude Code skills.
A complete case study: one fictional SaaS product taken from a founder’s notes to a tested, specified UI, including the full PRD, a token set and 8 screens.
Quick reference: a one-hour quick start, a one-page cheat sheet and an Evidence Ladder card.
What makes it different
The Evidence Ladder: every claim is tagged from H0 (AI hypothesis) to V4 (validated), so you can see at a glance which parts are evidence and which are guesses.
Human decision points: the toolkit marks the decisions that stay with you. Claude prepares the options, and you make the call.
No invented facts: prompts that touch competitors or markets work only from sources you paste in.
Built for product designers, freelance UX/UI designers, early-stage founders and small product teams.
Delivered as plain Markdown, HTML, CSV and JSON files that work in any editor, Notion, Google Docs or a Git repository. The case-study data is fictional.

Get it for$0.00
Tags
Claude
Digital Marketer
Graphic Designer
Interaction Designer
Product Designer
Product Manager
Product Marketer
Product Strategist
UI Designer
UX Designer
UX Researcher
AI prompts
UI kits
Product created by
M M Azaz Dhaka, Bangladesh
1
Paid projects
110
Followers
Get it for$0.00
Tags
Claude
Digital Marketer
Graphic Designer
Interaction Designer
Product Designer
Product Manager
Product Marketer
Product Strategist
UI Designer
UX Designer
UX Researcher
AI prompts
UI kits
Use AI to move faster, without inventing the evidence.
Claude can draft a persona, a PRD or a journey map in a minute, and all of them will read as plausible. The AI Product Design Toolkit is a working method that gets you that speed while keeping AI guesses clearly separate from what your users actually showed you.
What’s inside
Playbook: 8 chapters covering six modules, from Product Discovery through UX Research, Strategy, UX Architecture and UI Design to Validation & Handoff. Each module covers when to use AI, when not to, the workflow, quality checklists and common mistakes.
126 prompts, each built in seven parts (role, context, objective, inputs, constraints, output format, quality bar) and linked to the prompts before and after it. You can find them by module, by deliverable, or by what you already have.
11 editable templates that match the prompts’ outputs: PRD, research plan, interview guide, synthesis, user journey, component spec, handoff spec, UX audit and more.
Claude workflows: Project instructions to paste in, a guide to which files to upload, 7 step-by-step runbooks, and 3 optional Claude Code skills.
A complete case study: one fictional SaaS product taken from a founder’s notes to a tested, specified UI, including the full PRD, a token set and 8 screens.
Quick reference: a one-hour quick start, a one-page cheat sheet and an Evidence Ladder card.
What makes it different
The Evidence Ladder: every claim is tagged from H0 (AI hypothesis) to V4 (validated), so you can see at a glance which parts are evidence and which are guesses.
Human decision points: the toolkit marks the decisions that stay with you. Claude prepares the options, and you make the call.
No invented facts: prompts that touch competitors or markets work only from sources you paste in.
Built for product designers, freelance UX/UI designers, early-stage founders and small product teams.
Delivered as plain Markdown, HTML, CSV and JSON files that work in any editor, Notion, Google Docs or a Git repository. The case-study data is fictional.