Brain dump (raw)
Original input from /new-project. Stored as raw evidence per FORCED_RULE_PO_02; never edited in place.
# Live Poker Streaming Trading Platform — Handoff Document
## Product Overview
A trading platform where users buy and sell shares of poker players during their live-streamed sessions. Share values track the player's chip stack over the course of the session, with settlement happening when the player cashes out. v1 launches as a free-to-play entertainment product using Gold Coins (GC), partnered with a local poker room.
## Core Mechanics
### Shares and Pools
- When a player buys in for $X, exactly X shares are minted (1 share = 1 GC face value)
- Example: Phil buys in for $1000 → 1000 shares minted, pool sized at 1000 GC face value
- Total share supply is fixed at the player's buy-in amount
- Pool reserves mirror the chip stack throughout the session
### The IPO
- Before each session, shares are offered to the market
- Pricing is auction-style — users bid for shares
- You (the platform) act as initial market maker, holding all 1000 shares and selling into demand
- Final clearing price is whatever users are willing to pay
- Pool gets funded at face value (1000 GC); any premium above face value goes to you as platform revenue
- Unsold shares stay in your inventory — you hold them through the session
- This guarantees sessions always run (no failed IPOs)
### Settlement
- Session ends when player cashes out
- Operator manually enters the final chip stack amount
- Each share redeems at `final_stack / total_shares` GC
- All shareholders get paid out of the pool proportionally
- The platform's market-making position settles like any other holder
### Trading During the Session
- Peer-to-peer order book (bids and asks)
- No bonding curve, no AMM — pure order matching
- Traders post limit orders or take market orders
- Order book runs continuously through the session, including during hands
- No trading pauses, no circuit breakers — chaos is a feature, not a bug
- Real-time order book updates broadcast to all viewers
- Book value is NOT displayed live; traders watch the stream and price shares based on what they see
## Session Format
- Single buy-in declared upfront (e.g., $1000)
- Fixed blinds, no level escalation
- Player can sit at the table with less than their full bankroll, reloading from bankroll as they choose
- Session ends when bankroll is exhausted or player voluntarily cashes out
- No rebuys beyond the declared bankroll cap
- This is essentially a "capped session" cash game format, purpose-built for the trading product
## The Currency System
### Gold Coins (GC) — v1
- $1 = 10 GC (purchase rate)
- 100 GC = $1 of room time/services (redemption rate)
- Redemption ratio creates 10x leverage in user's favor
- Used for all trading and IPO subscription
- Not redeemable for cash — only for poker room services and merchandise
### Revenue Mechanics
- IPO premium: revenue from market-making spread (users overpay during IPO; you keep the surplus)
- Trading fees: small percentage on matched trades (optional for v1)
- Settlement fee: small skim on pool at settlement (optional for v1)
- Most margin comes from unredeemed GC and the favorable purchase-to-redemption ratio
### Example Economics — Popular Session
- Phil buys in $1000 → 1000 shares
- IPO clears at 3 GC per share (3x face value)
- Users pay 3000 GC total = $300 real money
- Pool gets 1000 GC face value
- You keep 2000 GC premium
- If user redeems 1000 GC for room time: you owe $10 of services → keep $290 margin
- If user redeems nothing: you keep $300
## Partner Poker Room
- Single local poker room for v1 (relationship already exists)
- Room provides: live stream, players, venue, redemption fulfillment
- You provide: trading platform, audience engagement layer
- Redemption catalog: cash game session time, tournament entries, food and drink, merchandise, lessons, VIP perks
- Suggested redemption ratios protect room margins (e.g., 100 GC = $1 of room time)
- Settlement frequency with room: weekly or monthly payments for actual redemptions
## Legal Structure — v1
- Pure entertainment product, no cash redemption
- GC is non-redeemable for cash; only for partner room services
- Same legal category as mobile games with in-app purchases
- No KYC required
- No sweepstakes infrastructure
- No state-by-state geofencing complexity
- Standard mobile app terms of service
- Minimal legal cost (a few thousand for basic terms)
### Future Phases (Not v1)
- Phase 2: Add Sweeps Coin layer for cash redemption (sweepstakes model, à la Global Poker)
- Phase 3: Skill-gaming product in permissive states
- Phase 4: Regulated product nationally (CFTC/DCM path)
- Phase 5: International expansion
These are deferred until v1 has traction and investors.
## Product Architecture
### Standalone vs. ValueBet Integration
- Built as a standalone product separate from ValueBet
- No shared infrastructure for v1 — avoid coupling complexity
- Integration possible later if both products warrant it
- Each product has its own brand, codebase, and operations
### Tech Stack (Implied)
- Webapp, mobile-responsive (not native)
- Auth: Supabase or similar
- Payments: Stripe
- Database: Postgres-based, with strong transactional integrity
- Real-time: WebSockets or Supabase realtime for order book broadcasts
- Hosting: Vercel or similar
### Stream Integration
- No video infrastructure built by you
- Stream lives on the room's existing platform (YouTube/Twitch)
- Your app is purely trading; users watch the stream separately
- Trading data drawn from operator manual entry (chip stack updates at session end), not live API integration
- No PiP, no embedded video in v1 — keeps mobile UX uncluttered
## Build Scope — v1
### Required Components
1. User auth and signup
2. GC wallet with deposit/balance tracking
3. Stripe integration for GC purchases
4. Admin panel: create sessions, manage player roster, enter final chip stack, trigger settlement
5. IPO subscription interface
6. Order book UI and matching engine
7. Trade execution with ledger-style accounting
8. Settlement logic (compute final share value, distribute GC to shareholders)
9. Portfolio view (user's holdings, P&L, history)
10. Session state machine (created → IPO open → IPO filled → session active → settled)
### Not in v1
- Live chip stack data feed
- Real-time book value display
- Stream embedding
- Multiple simultaneous sessions/tables
- Sweepstakes / cash redemption
- Multi-room partnerships
- Native mobile app
- Bot trading APIs
## UX Standards
The bar is PrizePicks-level polish — mobile-first, fast, low-friction, addictive.
### Design Principles
- One or two taps per action; minimal confirmation dialogs
- Visual hierarchy with big player photos and clear numbers
- Mobile-native gestures (swipe, tap, drag)
- No clutter — hide complexity until needed
- Smart defaults and presets
- 60fps animations throughout
- Haptic feedback on key actions
- Real-time order book and trade updates with smooth animations
### Key Screens
1. **Lobby:** Vertical feed of player cards with live indicators, current price, mini chart, your position, trade button
2. **Player trading view:** Player info, order book depth chart, your position, large buy/sell buttons, recent trades feed
3. **Buy/sell sheet:** Slide-up bottom sheet with amount slider, price input, single-tap confirmation
4. **Portfolio view:** All holdings across players, sortable, drillable
5. **Settlement notification:** Push notification + in-app animation when sessions end
### Three UX Modes to Design For
1. **Watching mode:** Glanceable, minimal info, fast action while user is focused on stream
2. **Researching mode:** Lobby with full information for browsing
3. **Reacting mode:** One-tap simplicity for big moments (all-ins, swings)
## Timeline
- Functional v1 (rough): 1-2 weeks with AI assistance
- PrizePicks-polished v1: 3-4 weeks realistic
- Suggested approach: ship functional v0.5 internally for personal use during real streams, then polish based on observed pain points
## Open Questions / Deferred Decisions
1. **Player consent and revenue share** — what players receive for opting in (TBD, figure out later)
2. **Trading fees structure** — flat fee, percentage, none for v1?
3. **Settlement fee** — yes/no for v1
4. **Operator UX for entering final chip stack** — needs to be fast and error-resistant
5. **Onboarding flow** — get user from signup to first trade in <60 seconds
6. **Pre-launch GC sale** — how to structure, pricing tiers, bonus allocations for early buyers
7. **First session lineup** — which players, which day, which streams
8. **Anti-bot measures** — likely web/mobile UI only, no API access in v1
9. **Branding** — product name, positioning, visual identity
## Strategic Context
This follows the same pattern as TowMarX/SiXiS:
- ValueBet could be framed as the platform engine for live trading markets
- Sports prediction (current ValueBet) is the first vertical
- Live poker trading is the second vertical
- Future verticals possible (esports, live events, etc.)
- Each vertical can be standalone now and consolidated later if valuable
The v1 launch is designed to:
1. Validate the trading mechanic with real users
2. Generate data on engagement, GC purchase behavior, retention
3. Build a story for investors (traction, not just an idea)
4. Establish the partner room relationship as a template for future rooms
5. Stay legally simple while proving product-market fit
---
We only want to start with a landing page:
What the landing page needs to do
Two audiences, two jobs:
For the poker room (your pitch target): Show them this is a real, polished concept worth partnering on. Not vaporware. They should see it and think "this is happening with or without me, I want in."
For potential users (email signups): Build a waitlist that proves audience demand. Your pitch to the room gets stronger with every signup.
Landing page structure
Above the fold:
Logo + name (Sweats)
One-line value prop: "Trade shares of poker players live"
Hero visual (mock screen of the trading UI on mobile)
Email capture CTA: "Get early access"
How it works section:
3 simple steps with icons
"Player buys in → You buy shares → Trade as they play → Cash out when they cash out"
The product section:
Mock screenshots of lobby, trading view, portfolio
Bullet on key features: live order book, real-time trading, leaderboards
Why now section:
Poker streaming is bigger than ever
Audiences want skin in the game
Built for the streaming era
Partner section (subtle):
"Powered by [Poker Room Name]" placeholder
Or "Now partnering with select poker rooms"
Footer:
Email signup again
Social links (even if empty for now)
Contact email for partnerships
The pitch deck for the room
Separate from the landing page, but referenced. A 5-7 slide deck:
The opportunity: Poker streaming is a $X market, audiences want deeper engagement
The product: Sweats lets viewers trade shares of players in real time
The mechanic: Quick visual of how IPO + trading works (with simple numbers)
What we need from you: Stream access, players opt in, redemption catalog
What you get: New revenue stream, deeper audience engagement, novel content angle, foot traffic from GC redemptions
The ask: Pilot partnership for 90 days, exclusive in your market
Why us: Already built X, can launch in Y weeks, here's the waitlist signal
What to build first
Order of operations:
Domain + basics (1 day): Register sweats.poker, set up email, social handles
Landing page (2-3 days): Use Framer, Webflow, or just a Next.js site on Vercel. Polished but not overengineered
Mock UI screenshots (1-2 days): Design 3-4 mobile screens in Figma — lobby, trading, portfolio. These sell the concept
Pitch deck (1 day): 5-7 slides, clean design, real numbers
Email setup (half day): ConvertKit or Beehiiv for waitlist, simple welcome email
Pre-launch announcement post (1 day): X/Twitter, poker subreddits, maybe a short demo video
Total: about a week to be ready to pitch.
The pitch meeting
Walk in with:
Live landing page on your laptop
Signup count (even if it's 50, it's signal)
Printed pitch deck
A clear ask: "90-day pilot, you provide stream + one player consent, we provide platform + audience"
Don't overpromise. Frame it as a test, not a commitment. They have low downside, you have a real shot at validation.
One specific thing to test in the pitch
Find out what their pain points actually are before you pitch the solution. Open with questions:
How are you thinking about growing your audience?
What's working in your streaming strategy?
Have you tried interactive content?
If their pain matches your solution, the pitch lands. If it doesn't, you've learned to adjust the pitch — or move to a different room.
Mock UI matters more than copy
Spend extra time on the mock screens. Investors and partners visualize products through screens, not paragraphs. A polished mobile mockup of the trading UI sells the concept in 5 seconds. Bad mockups undermine even good ideas.
Use Figma. Pull design inspiration from PrizePicks, Robinhood, Polymarket. Doesn't need to be functional — just needs to look real and beautiful.
Event timeline (19)
Every event tagged with this draft_id, in chronological order.
discovery_prompt_generated
system · 2026-05-14 15:06:41+00:00
Generated the discovery drill prompt for draft d662c946-5083-455f-837e-7399e79ad70b using prompt_templates.discovery_drill v3.
project_draft_started
system · 2026-05-14 15:06:41+00:00
You started a new project draft via /new-project (draft_id=d662c946-5083-455f-837e-7399e79ad70b).
draft_saved
system · 2026-05-14 15:06:41+00:00
You saved the brain-dump for draft d662c946-5083-455f-837e-7399e79ad70b: "# Live Poker Streaming Trading Platform — Handoff Document
## Product Overview
A trading platform where users buy and sell shares of poker players during their live-streamed sessions. Share valu…"
ui_launch_event
system · 2026-05-14 15:07:05+00:00
You clicked 'Copy prompt' on /new-project for draft d662c946-5083-455f-837e-7399e79ad70b — prompt taken to Claude Code (prompt_hash=861721a41f6e…).
discovery_answer_logged
tommy · 2026-05-14 15:08:52+00:00
Q: "Is the scope just the landing page itself, or does it include the mock UI mockup…" — A: "Mockup UI as well — both the landing page and the mock UI screens it displays are in scope for this project."
discovery_answer_logged
tommy · 2026-05-14 15:10:55+00:00
Q: "Who/what produces the mockups, and how do they live in the project?" — A: "Path 1: mockups built as real HTML/CSS components inside the landing page (interactive-looking, rendered by the site its…"
discovery_answer_logged
tommy · 2026-05-14 15:11:56+00:00
Q: "Is 'Sweats' the locked product name, and is sweats.poker already registered or d…" — A: "Name 'Sweats' is locked. Domain sweats.poker is already purchased."
discovery_answer_logged
tommy · 2026-05-14 15:14:14+00:00
Q: "Is the pitch deck part of this project, or out of scope?" — A: "Pitch deck is OUT of scope. This project is just the landing page (plus the in-page mockup UI components). User shared a…"
discovery_answer_logged
tommy · 2026-05-14 15:17:38+00:00
Q: "Which visual palette direction, and any existing Sweats brand assets to design a…" — A: "Borrow PrizePicks styling overall (dark gradient, two-phone hero, big CTA), but swap purple for red. No existing Sweats …"
discovery_answer_logged
tommy · 2026-05-14 15:18:36+00:00
Q: "Where do email signups go, and do you want welcome-email/broadcast capability?" — A: "Path 1: new dedicated Supabase project for Sweats with a waitlist table. Just capture addresses for v1 — no welcome-emai…"
discovery_synthesis_logged
claude · 2026-05-14 15:40:26+00:00
Claude synthesized the discovery drill intent for draft d662c946-5083-455f-837e-7399e79ad70b: "Z1: Build a polished single-page marketing site for Sweats at sweats.poker (Next.js on Vercel, mobile-first, dark-PrizePicks aesthetic with red swapped for purple, green CTA) whose job is dual-audienc…"
wizard_step_completed
claude · 2026-05-14 15:41:03+00:00
Claude marked the discovery drill complete for draft d662c946-5083-455f-837e-7399e79ad70b. Wizard advances from discovery to intent-freeze step.
brain_response_logged
deepseek · 2026-05-14 16:09:34+00:00
DEEPSEEK Round 1 response for draft d662c946-5083-455f-837e-7399e79ad70b: "DeepSeek (Expert mode, Thought for 70s) — Round 1 position
Agrees with orchestrator instinct on HERO. Disagrees on BELOW-FOLD vs SKIP.
Recommendation:
- HERO (two tilted phones): LOBBY + PLAYER TRAD…"
brain_response_logged
claude · 2026-05-14 16:10:24+00:00
CLAUDE Round 1 response for draft d662c946-5083-455f-837e-7399e79ad70b: "Claude.ai (Opus 4.7, inside Sweats.Poker project) — Round 1 position
Disagrees with orchestrator instinct on HERO. Flips TRADING VIEW out of hero.
Recommendation:
- HERO (two tilted phones): LOBBY +…"
brain_response_logged
gpt · 2026-05-14 16:10:44+00:00
GPT Round 1 response for draft d662c946-5083-455f-837e-7399e79ad70b: "GPT (Instant mode, inside sweats.poker project) — Round 1 position
Reframes the question before answering. Calls the orchestrator instinct "mostly right" but pushes back on the framing.
Key reframe:…"
brain_response_logged
deepseek · 2026-05-14 16:37:12+00:00
DEEPSEEK Round 2 response for draft d662c946-5083-455f-837e-7399e79ad70b: "DeepSeek Round 2 — SHIFTED.
Final picks:
- Hero (two-phone): LOBBY + BUY/SELL SHEET
- Below-fold: TRADING VIEW + PORTFOLIO
- Skip for v1: SETTLEMENT NOTIFICATION
What moved it: Claude.ai's 5-second-…"
brain_response_logged
claude · 2026-05-14 16:37:33+00:00
CLAUDE Round 2 response for draft d662c946-5083-455f-837e-7399e79ad70b: "Claude.ai Round 2 — HELD on hero, SHIFTED on below-fold (pulled settlement back IN as cashout-receipt, dropped PORTFOLIO).
Final picks:
- Hero (two-phone): LOBBY + BUY/SELL SHEET (held from R1)
- Bel…"
brain_response_logged
gpt · 2026-05-14 16:37:54+00:00
GPT Round 2 response for draft d662c946-5083-455f-837e-7399e79ad70b: "GPT Round 2 — UPDATED partially. Synthesizes Claude's comprehension-tax critique but rejects the buy/sell-sheet-in-hero solution.
Final picks:
- Hero (two-phone):
- FRONT/DOMINANT phone: LOBBY
- …"
draft_promoted_to_project
claude · 2026-05-15 02:15:35.693834+00:00
Draft d662c946 promoted to active project — Sweats Trading Platform v1. Card 1 (Foundation & Auth) shipped end-to-end; spec sheet locked at cycle 879ca7b7.