Claude/ChatGPT Prompt to Build a P2P Marketplace Trust Interface

What this prompt does

This prompt engineers the complete trust layer of a peer-to-peer marketplace across all eight outputs: listing cards, listing detail pages, seller profiles, in-app messaging, escrow checkout, mutual review systems, dispute resolution flows, and working framework code. The template explicitly bans anti-patterns like off-platform communication and hidden seller identity — the same problems that forced Airbnb, Vinted, and eBay to retrofit trust systems after fraud patterns emerged at scale.

What makes the template effective is its sequenced architecture. It follows the actual buyer journey — browse, inspect, contact, pay, review, dispute — rather than jumping straight to UI components. The industry rules enforce escrow payment prominence and seller verification badges at every touchpoint, which are the two signals buyers actually check before committing money to a stranger. The framework variable ensures the final output is a usable component, not a wireframe description.

The result is a complete design brief plus working code for a listing detail page with seller profile and offer CTA, ready to hand off or iterate on.

When to use it

• Building a secondhand fashion app where buyer protection and return disputes need to be airtight
• Designing a freelance services marketplace where milestone escrow and partial release matter
• Launching a local goods exchange (electronics, furniture) where meetup vs. shipping options affect trust differently
• Prototyping a rental platform (tools, cameras, gear) where deposit handling and damage disputes are the core UX problem
• Adding a resale channel to an existing e-commerce brand and needing the full trust scaffolding fast
• Pitching a marketplace concept to stakeholders who need to see the complete trust flow, not just the listing card

Example output

Marketplace: Local used electronics
Transaction type: Buy/sell with escrow, meetup or shipping
Trust challenge: Buyers fear counterfeit items; sellers fear chargebacks

Generated:
- Listing card: thumbnail, £85 price, 4.8 (63 reviews), "Verified Seller",
  London · 2 days ago, save/share icons
- Listing detail: 6-image gallery with zoom, condition notes, "Make Offer"
  and "Buy Now" CTAs, seller profile preview pinned to sidebar
- Seller profile: ID-verified badge, 98% response rate under 1 hr, 140 sales,
  PayPal/bank transfer escrow supported
- Messaging: in-app chat with image attach, "Counter Offer" button inline,
  phone number sharing blocked with inline warning
- Checkout: escrow hold explanation ("funds released after you confirm receipt"),
  buyer protection badge, meetup pin or tracked shipping toggle
- Review system: mutual review prompt for both parties, star + text, verified
  purchase badge, 14-day review window enforced
- Dispute flow: 3 steps — raise issue -> upload photos/chat log -> moderator
  assigned within 24 hr; auto-refund if seller non-responsive over 72 hr
- React component: ListingDetail with SellerCard, MakeOfferModal, EscrowBadge

Pro tips

• Set trust_challenge to the specific fear your users have (counterfeit goods, no-shows, chargebacks) — the template shapes dispute flow and verification emphasis around it, so vague inputs like "building trust between strangers" produce generic outputs that miss the actual risk.
• For transaction_type, specify whether escrow is instant release or milestone-based. The checkout and dispute sections behave differently and the model needs that distinction to generate accurate copy and flow logic — not just visual placeholders.
• Run the prompt twice with different framework values (React vs. Alpine.js) and diff the component structures. The offer CTA wiring differs meaningfully: React output tends toward controlled modal state, Alpine.js toward inline x-data toggling. Seeing both before you commit to a stack is worth the extra run.
• If your marketplace is voice or video based, the messaging section's no-phone-number rule will produce a chat-only interface. Override it by adding an explicit instruction at the end of the prompt body — something like "Allow voice/video call initiation from within the messaging screen" — rather than trying to route this through trust_challenge, which is a context field, not an override mechanism.
• The escrow explanation step is where checkout flows most often break down in user testing on P2P products. Treat the AI-generated copy there as a hypothesis rather than final copy — run it past a handful of target users before assuming the framing lands, regardless of how polished the output looks.