- Home
- Skills
- Product Management
- Rapidly: PRD to Clickable Prototype
More screenshots
Works with the AI Agents you already use
Rapidly: PRD to Clickable Prototype
Paste a PRD and get a clickable HTML prototype, a Figma Make prompt and a structured model of your product, in one conversation.
$9
Rapidly: PRD to Clickable Prototype
Example session with this skill installed
Prototype this PRD: customers submit refund requests, store managers approve or reject from a queue, refunds over €200 need a regional manager's second approval, finance downloads a daily CSV.
- Read your context and instructions
- Compiled the rapidly
- Generated the UI component
Summary
Rapidly shows the model it extracted (4 jobs, 3 flows, 4 actors, 1 rule, 3 open questions) and asks you to confirm it before generating anything.
Deliverable
prototype/index.html with 8 linked screens:
| Screen | What it shows |
|---|---|
| Overview | The PRD's three metrics |
| Refund queue | Filters for store managers |
| Request detail | Order and refund history |
| Decision | The €200 rule switches the button to "Send for second approval" |
| Regional manager approvals | List of refunds awaiting a second approval |
| Customer request form | Mobile form and confirmation |
| Finance export | Daily CSV download |
Plus figma_make_prompt.txt, cpm.json and prototype_plan.json.
Next steps
- Open
prototype/index.htmlin a browser and click through the flows - Say "review" to edit the model and regenerate
- Paste
figma_make_prompt.txtinto Figma Make for a high-fidelity version
index.html
TSX · React component
Example file from a real run - the skill writes it into your workspace.
Connects securely to your tools. The creator never sees your data.
About this skill
Most PRDs never get a prototype before engineering starts, because a designer is busy or the idea is still rough. Rapidly closes that gap. Give your agent a PRD, pasted or as a file, and Rapidly pulls out the problem, the outcome, the users, the jobs to be done and the step-by-step flows. It shows you that model first, so you fix a wrong flow before anything gets built. Then it writes a single HTML file you can open in any browser and click through: an overview with your real metrics, one screen per flow step, wizards, approvals and tables filled with sample data that uses your product's own terms. It also writes a Figma Make prompt for a high-fidelity version. Adding a feature to an existing app? Share a few screenshots and the prototype matches its colors, fonts and layout. Inferred details and open questions are flagged, never passed off as facts.
What it does
- Reads any PRD format: markdown, text, docs, pasted text, with optional problem, research or decision docs
- Builds a Canonical Prototype Model (problem, outcome, actors, jobs, flows, entities, rules, open questions) and asks you to confirm it before generating
- Picks the right output: clickable UI, data flow diagram, architecture diagram, or a mix
- Writes one self-contained
index.htmlthat opens from disk, with working navigation, wizards, approvals, toasts and no dead clicks - Matches an existing app from screenshots, or picks a layout that fits the product (dashboard, web app, phone, feed, marketplace)
- Writes a standalone Figma Make prompt with every screen and the design system
- Runs quality checks: every must-have job has a screen, every link has a target, no filler text, business rules are visible
- Flags sensitive content (customer names, credentials, confidential numbers) before it goes into a prototype
Why this beats prompting it yourself
A plain prompt jumps straight to HTML and invents whatever the PRD left out. Rapidly extracts a model first, shows it to you, marks every guess as inferred, and builds screens from that model, so the prototype traces back to the PRD and the files stay consistent when you iterate.
Who it's for
- Product managers who want stakeholders clicking through an idea before the first design review
- Founders and solo builders turning a spec into something to show users or investors
- Designers who want a fast first pass and a Figma Make prompt to start from
- Engineering leads who want to check a PRD's flows before estimating
Use cases
- Turn a one-page PRD into a demo for tomorrow's stakeholder review
- Show how a new feature fits into your existing app, using screenshots of the current UI
- Find gaps in a PRD: missing states, undefined owners, rules with no screen
- Generate a data flow or architecture diagram from a backend or data PRD
What you get
SKILL.mdwith the full 11-step workflow- 4 reference guides: extraction rules, model schema, Figma Make prompt structures, HTML prototype spec
- An HTML prototype shell with design tokens, components and navigation
- A sample PRD and its complete generated output
Frameworks and tools
Works in any agent that can read and write files: Claude Code, Cursor, Codex and similar. Screenshot matching needs an agent that can read images. No API keys, no installs, no scripts.
Known limitations
Prototypes are front ends with sample data. They do not call real APIs or save data. Screenshot matching gets close to an existing style, but exact brand fonts or custom icon sets may need a manual swap. Very long PRDs are condensed to the must-have jobs, and the rest is listed as not prototyped.
Find more skills like this on the Claude Skills Marketplace, the Codex Skills Marketplace and the Gemini Skills Marketplace, or browse Claude Skills for Product Management.
Prototypes are front ends with sample data. They do not call real APIs or save data. Screenshot matching gets close to an existing style, but exact brand fonts or custom icon sets may need a manual swap. Very long PRDs are condensed to the must-have jobs, and the rest is listed as not prototyped.
How to use
- 1
Connect your agent to the Agensi MCP
Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup
- 2
Sign in to Agensi
Your agent shows a sign-in link. Sign in and click Allow.
- 3
Ask for what you need
Your agent loads this skill and uses it on your task. It always asks before buying a skill.
Reviews
No reviews yet
Be one of the first to try it. Every listed skill goes through the trust checks below.
Security scanned
Checked by our 8-point scan before listing
Fresh listing
Recently published to Agensi
Money-back guarantee
Doesn't do what its page promises? You get your money back.
Trust & safety
Security scan
Security scan passed 21 days ago
Permissions
File Scopes
Read: the PRD and any docs or screenshots you point it to. Write: a prototype/ output folder only; it never edits your PRD. Network: none by the skill. The generated index.html loads Google Fonts and the Lucide icon library (pinned version) when opened. Use --offline for a prototype with no external requests.