CAPACITR — Technical Specification

00

Thesis: Signal-to-Position

The Problem

Markets exist everywhere — prediction markets, perpetual futures, options — but finding the right market at the right time is still manual, fragmented work. A trader who follows AI policy news has to manually search Polymarket for relevant prediction markets, check HyperLiquid for NVIDIA perps, and scan Deribit for options plays.

The information that drives trading decisions — trending topics, breaking news, social sentiment, expert chatter — lives in completely different places than the markets themselves. Nobody connects them.

The Insight

Every tradeable market is downstream of a real-world narrative. "Will the Fed cut rates?" is a Polymarket prediction, a BTC perpetual trade, and a set of Treasury options all at once. The narrative is the root. The markets are leaves on the same tree.

If you know what someone cares about and you can see what the world is talking about right now, you can surface exactly the markets that matter to them, ranked by:

Relevance

Does this market match their interests and expertise?

Volume & Liquidity

Is there enough depth to take a meaningful position?

Social Momentum

Are people talking about this with increasing intensity?

Sentiment Signal

Is the conversation bullish, bearish, or uncertain?

News Attention

How much mainstream press is covering this narrative?

Timeliness

Is this breaking right now or fading?

The Product

Capacitr is the bridge between signals and positions.

01. Absorb Signals

Poll X, Reddit, Farcaster, news APIs, Google Trends, and crypto feeds continuously. Extract trending narratives.

02. Charge with Intelligence

LLM analysis categorizes, summarizes, and extracts tradeable entities. "NVIDIA earnings beat" connects to an NVIDIA perp, a semiconductor option, and a prediction market about AI chip exports.

03. Discharge Matched Markets

For every narrative, surface positions across three market types:

Predictions (Polymarket) — Binary Outcomes Perps (HyperLiquid) — Directional, Leveraged Options (Deribit/Aevo/Derive) — Asymmetric Risk/Reward

04. Personalize the Feed

Interests, social graph, and engagement patterns continuously refine what surfaces. The feed gets sharper over time.

05. Enable Instant Action

Every feed card shows the narrative AND matching markets with live prices, volume, and entry points. Read a story, evaluate the trade — same view.

Who It's For

Crypto Traders

Want to trade narratives, not just charts. See a tweet, take a position.

Prediction Market Enthusiasts

Want relevant markets surfaced to them instead of browsing thousands.

Subject Matter Experts

Policy wonks, sports analysts, tech insiders — deep domain knowledge, no time to hunt for matching markets.

Degens

Fastest path from "I just saw this tweet" to "I have a position on this."

The Moat

The value compounds. As users engage — clicking, expanding, trading — Capacitr learns which signal-to-market connections actually convert. The matching algorithm improves. The feed sharpens. The proprietary dataset of "which narratives drive which markets" becomes a unique asset that no individual exchange or news platform has.

01

Product Scope

⚡

Phase 01 — Link Drop MVP

PROTOTYPE // WEEK 01-02

Enter the terminal. Paste a link to an X post or news article. System scrapes content, analyzes with LLM, returns matched Polymarket predictions + HyperLiquid perps. No auth required.

Web Only No Auth Single API Call

🔋

Phase 02-04 — Full Discharge

PRODUCTION // WEEK 03-10

React Native app. Privy auth with social login. Calibrate the terminal to show markets that matter to you. Auto-generated trending feed with market overlays. Social data enrichment from X and Farcaster.

React Native Privy Auth Personalized Feed

02

System Architecture

📱 Client LayerReact Native (Expo) + Privy SDK

▼ REST / WS

⚡ API GatewayHono // /auth /feed /scrape /markets

▼

🔐 AuthPrivy

📡 PollerBullMQ

🔍 ScraperJina / X API

🎯 MatcherLLM + Search

▼

🗄️ PostgreSQLSupabase

⚡ RedisUpstash + BullMQ

🧠 LLMOpenAI / Claude

▼

𝕏 Twitter

📰 NewsAPI

📊 Polymarket

📈 HyperLiquid

🔮 Farcaster

Tech Stack

Layer	Technology	Rationale
Mobile Client	`React Native (Expo)`	Cross-platform, Privy RN SDK support
Web Client	`Next.js`	Prototype-first, SSR for SEO
API Server	`Node.js + Hono`	Fast, typed, edge-compatible
Database	`PostgreSQL (Supabase)`	Managed, real-time subscriptions
Cache / Queue	`Redis (Upstash) + BullMQ`	Serverless Redis, reliable job queues
LLM	`GPT-4o / 4o-mini`	Best cost/quality for classification
Image Gen	`DALL-E 3 / Flux`	Editorial illustrations for feed cards
Hosting	`Vercel + Fly.io`	Vercel for web, Fly for API workers

03

Polymarket Trading Integration Plan (App-Native)

CRITICAL IMPLEMENTATION NOTE

Polymarket trading requires user-signed orders (non-custodial). Builder attribution is attached per order via builderCode. Authentication for trading is two-layer: L1 wallet signature for API credential derivation, then L2 HMAC headers for CLOB requests.

Source: Polymarket docs — builder overview + CLOB auth/signature model.

Execution Architecture

01. Discovery → App Deep Link

Route markets/discovery clicks into app.capacitr.xyz/trade?venue=polymarket&market=... so trade intent is preserved.

02. Wallet + Signature Type Selection

Privy wallet is user identity. Use deposit-wallet flow for new API users (recommended: POLY_1271, signature type 3), with funderAddress set to deposit wallet.

03. Provision Trading Credentials

Use user wallet signature (L1) to create/derive CLOB API creds, then persist encrypted L2 creds (api key/secret/passphrase) server-side.

04. Funding + Allowance Readiness

Guide user through gasless relayer/deposit wallet operations, check balances + allowances before enabling submit.

05. Place Attributed Orders

Create EIP-712 signed order payload and submit to CLOB with builderCode attached on every order.

06. Reconcile Positions + Attribution

Track fills/positions in Capacitr ledger and reconcile builder-attributed volume for leaderboard + internal analytics.

Builder Program Requirements (Engineering Checklist)

Item	Requirement	Why It Matters
Builder profile	Create builder profile and register builder code in Polymarket settings.	Enables volume attribution and program benefits.
Order attribution	Attach `builderCode` to every submitted order.	Only attributed orders count toward builder volume.
Wallet mode	New API users should use deposit-wallet flow + signature type 3 (`POLY_1271`).	Best path for gasless ops and current recommended integration.
Auth model	L1 (wallet signature) + L2 (HMAC headers) both required for trading.	Prevents failed order submissions and auth drift.
Signing ownership	User still signs order payloads even with L2 headers.	Non-custodial guarantee and compliance with protocol model.

Readiness State Machine (User UX)

wallet_missing            -> prompt Privy wallet create/connect
deposit_wallet_missing    -> deploy via relayer flow
api_creds_missing         -> create/derive L2 credentials
balance_or_allowance_low  -> fund wallet / approve spending
ready_to_trade            -> show confirm + submit
order_submitted           -> open/filled/canceled/error states
    

Capacitr Services to Implement

pm-intent service

Builds and stores intent from discovery deep links (market, side, size, source surface).

pm-wallet service

Manages Privy wallet binding, deposit-wallet provisioning state, and signature type metadata.

pm-auth service

Handles L1 API key derivation and encrypted L2 credential lifecycle (rotate/recover/revoke).

pm-execution service

Creates/sends signed CLOB orders, injects builderCode, tracks order lifecycle, and handles retries/idempotency.

pm-risk service

Pre-trade checks for market status, liquidity, slippage, max size, and user exposure limits.

pm-ledger + attribution

Canonical records for orders/fills/positions + builder volume telemetry for leaderboard reconciliation.

API Surfaces Used

Surface	Use Case
`gamma-api.polymarket.com`	Public discovery/search/events/markets metadata.
`clob.polymarket.com`	Authenticated order placement/cancel/query with L2 headers.
`data-api.polymarket.com`	Trades, open interest, and market analytics enrichment.
`builder-relayer client`	Gasless deposit-wallet deployment + wallet batch operations.

Rollout Plan

STAGE 01

Internal Dogfood

Single market category, market orders only, strict notional caps, attribution verification on every fill.

STAGE 02

Allowlist Beta

Expand market set, add order cancellation/edit flows, monitor failure classes and funding-friction conversion.

STAGE 03

General Availability

Broader market coverage, higher limits, stronger risk controls, and paper/live routing controls in one execution UX.

Q

Quotient Intelligence

AI-computed fair odds for Polymarket markets — q-api.quotient.social · Auth: x-quotient-api-key header

What It Does

Quotient provides AI-computed fair odds for Polymarket markets, identifying mispriced opportunities where the market price diverges from the model's estimated probability. Markets with >5% edge get orange badges showing the spread, direction, and a BLUF (Bottom Line Up Front) analysis.

32% EDGE — BUY NO

API Endpoints

Endpoint	Purpose	Key Params
`GET /api/v1/markets/mispriced`	Batch mispriced markets with spread, direction, BLUF analysis	—
`GET /api/v1/markets/lookup`	Lookup specific market by slug or condition_id	slug, condition_id

Response Fields

Field	Type	Description
`question`	string	Market question text
`slug`	string	Polymarket market slug
`quotient_odds`	number	AI-computed fair probability
`market_odds`	number	Current Polymarket price
`spread`	number	Absolute difference between Q odds and market odds
`spread_direction`	string	`q_higher` or `q_lower`
`bluf`	string	Bottom Line Up Front — reasoning summary
`signal_count`	number	Number of signals informing the estimate
`volume_24h`	number	24-hour trading volume

Integration Flow

01. Fetch Mispriced

Per-query fetch from /api/v1/markets/mispriced with API key header

02. Keyword Match

Match mispriced markets to user search/feed topics via keyword overlap

03. Merge & Sort

Merge into predictions array, sorted by spread descending. Markets with >5% edge get orange badges

04. Display Edge

Badge shows "32% EDGE — BUY NO" with BLUF analysis text

Used In

Market Discovery /discover In-App Markets /markets Feed Enrichment Standalone Lookup by Slug Agent Tool: getMispriced

04

Hyperliquid Trading Integration Plan (App-Native)

CRITICAL IMPLEMENTATION NOTE

Builder codes are fee attribution, not custody delegation. Capacitr should not place all user trades from a single house wallet. User wallet remains the trading identity. Capacitr orchestrates flow, UX, and attribution.

Source: Hyperliquid docs — /trading/builder-codes, /for-developers/api/exchange-endpoint.

Execution Architecture

01. Discovery → App Deep Link

Replace external Hyperliquid link-outs from markets with app deep links: app.capacitr.xyz/trade?venue=hyperliquid&coin=BTC&side=buy&size=....

02. Wallet Readiness

Privy wallet create/connect. Treat that address as user Hyperliquid identity. No separate custodial account provisioning.

03. Funding Path

Fast path: already funded perps collateral. Fallback path: fund Capacitr wallet → bridge/swap to USDC on Arbitrum → deposit to Hyperliquid perps account.

04. Builder Approval

One-time ApproveBuilderFee per user + builder address, signed by user main wallet (not agent/API wallet).

05. Place Order

Submit signed order via Hyperliquid exchange endpoint with optional builder field: {"b":"builder_address","f":10} (10 = 1 bp).

06. Reconcile + Attribute

Persist order/fill state internally, query referral state, and backfill builder fills for analytics + revenue reconciliation.

Builder Code Semantics (What Engineering Must Implement)

Item	Requirement	Why It Matters
Builder approval	User must sign `ApproveBuilderFee` from main wallet.	Without this, builder fees cannot be charged or attributed.
Order payload	Include optional `builder: {"b": address, "f": number}`.	Per-order flexibility and explicit fee selection.
Fee bounds	Perps max builder fee is 0.1% (docs).	Prevent invalid orders and trust issues.
Builder account constraints	Builder address must have at least 100 USDC perps account value and standard account abstraction mode.	Required for production reliability.
Approval limits	User can have max 10 active builder approvals.	UI should manage approval slots and revocation.

Readiness State Machine (User UX)

wallet_missing        -> prompt Privy wallet create/connect
account_unfunded      -> funding stepper (fund, bridge/swap, deposit)
builder_not_approved  -> sign ApproveBuilderFee
ready_to_trade        -> show confirm + submit
order_submitted       -> live status (resting/filled/error)
    

Backend Services (Capacitr App)

trade-intent service

Parses deep-link params from discovery surfaces and creates resumable trade intent records.

hl-account service

Queries user account readiness, balances, open orders, and margin status by wallet address.

hl-builder service

Checks maxBuilderFee, manages fee presets, and tracks approval state per user.

hl-execution service

Builds order actions, enforces guardrails, submits orders, and maps Hyperliquid statuses into app status enums.

funding-orchestrator

Coordinates wallet funding, bridge/swap into USDC on Arbitrum, and perps deposit status transitions.

trade-ledger + analytics

Canonical internal record for every order/fill, builder fee attribution, and execution funnel conversion.

API Surfaces Used

POST https://api.hyperliquid.xyz/info
  - type: meta, metaAndAssetCtxs, allMids, openOrders, maxBuilderFee, referral

POST https://api.hyperliquid.xyz/exchange
  - action: order (with optional builder field)
  - signed by user-associated trading identity
    

Rollout Plan

STAGE 01

Internal Dogfood

Single-asset happy path (e.g., BTC perp), market orders only, low notional guardrails, full observability.

STAGE 02

Allowlist Beta

Limited user set, error-budget tracking, builder attribution verification, funding funnel tuning.

STAGE 03

General Availability

Expanded assets, richer order types, paper/live mode toggle, and venue-level kill switch controls.

05

Options Platforms

Third market type: crypto options. Asymmetric risk/reward and volatility exposure. All read endpoints free, no auth.

Deribit — Gold Standard (~90% of Crypto Options Volume)

www.deribit.com/api/v2 · CEX · 20 req/sec unauthed

Endpoint	Returns
`GET /public/get_instruments?currency=BTC&kind=option`	All BTC options (strike, expiry, type)
`GET /public/ticker?instrument_name={name}`	Price, IV, greeks, OI, volume, bid/ask
`GET /public/get_order_book?instrument_name={name}`	Full orderbook
`GET /public/get_book_summary_by_currency?currency=BTC&kind=option`	Summary of ALL BTC options
`GET /public/get_historical_volatility?currency=BTC`	Historical volatility

TICKER DATA INCLUDES: mark_price, mark_iv (implied vol %), greeks (delta, gamma, vega, theta, rho), open_interest, bid/ask, volume_24h, underlying_price

INSTRUMENT FORMAT: BTC-28MAR25-80000-C (asset-expiry-strike-type)

Aevo — Hybrid DEX (OP Stack L2)

api.aevo.xyz · DEX · ~100 req/min

Endpoint	Returns
`GET /markets?asset=BTC&instrument_type=OPTION`	All BTC options markets
`GET /orderbook?instrument_name={name}`	Full orderbook
`GET /index?asset=BTC`	Current index/spot price
`GET /statistics`	Exchange-wide OI and volume

Derive — DEX (Formerly Lyra V2)

api.derive.xyz · DEX · ~10 req/sec · Mirrors Deribit API design

Endpoint	Returns
`GET /public/get_instruments?currency=BTC&kind=option`	List options (Deribit-style)
`GET /public/get_ticker?instrument_name={name}`	Price, IV, greeks, OI
`GET /public/get_order_book?instrument_name={name}`	Full orderbook

Why Options Matter for Capacitr

Each market type answers a different question about the same narrative:

Predictions

"Will X happen?" — Binary outcome

Perpetuals

"Which direction?" — Directional exposure

Options

"How volatile?" — Asymmetric risk, defined downside

Example: "Fed rate decision tomorrow" →
Polymarket: "Will Fed cut rates?" (68% Yes)
HyperLiquid: BTC-PERP (directional bet)
Deribit: BTC options at nearest expiry (volatility play, defined risk)

Other Platforms Evaluated

Platform	Type	Status	Verdict
OKX Options	CEX	Active	Good supplementary API — `/public/opt-summary` with greeks/IV
Delta Exchange	CEX	Active	Smaller CEX, decent REST API
Premia Finance	DEX	Subgraph only	No REST API — requires GraphQL
Hegic	DEX	Low volume	Subgraph only, declining usage
Opyn / Squeeth	DEX	Deprecated	Wound down late 2024
Zeta Markets	DEX	Pivoted	Solana — pivoted to perps only

06

X / Twitter API

𝕏

Official API

$100/MO BASIC

10K tweet reads/mo, search recent, single tweet fetch. Sufficient for link-drop and limited search.

⚠ TRENDING TOPICS = PRO TIER ($5,000/MO)

GET /2/tweets/search/recent GET /2/tweets/:id

🔶

Third-Party Alt

FOR PROTOTYPE

SocialData.tools: $49/mo — full Twitter-like API, no rate pain.
Apify scrapers: Pay-per-use — scraper actors for tweets and trends.

Bypass Trending Paywall

Link Drop: Tweet Extraction

// 01. Extract tweet ID from URL
const match = url.match(/status\/(\d+)/)

// 02. Fetch via X API with entity expansion
GET /2/tweets/{tweetId}?expansions=author_id
    &tweet.fields=text,entities,context_annotations

// 03. context_annotations → free topic/entity tags
// Use tweet text + annotations for market matching
    

07

News & Article Scraping

News Aggregation APIs

API	Free Tier	Endpoint	Best For
NewsAPI.org	100 req/day	`/v2/top-headlines?category=`	Headlines by category
GNews.io	100 req/day	`/api/v4/top-headlines`	Backup source
CryptoPanic	Free w/ key	`/api/v1/posts/?filter=hot`	Crypto + sentiment
CoinGecko	30 req/min	`/api/v3/search/trending`	Trending coins

Article Content Extraction

⚡

Jina Reader

RECOMMENDED // FREE

https://r.jina.ai/{url} — returns clean markdown. Handles JS rendering, anti-bot. Zero config.

🔧

Firecrawl / Readability

PRODUCTION ALT

Self-host Mozilla Readability for speed. Firecrawl API (generous free tier) for complex pages.

08

Trend Sources

Source	Endpoint	Auth	Cost
Google Trends	`serpapi.com/google-trends-api`	API key	Free 100/mo
Reddit	`GET /r/{sub}/hot.json`	Optional	Free
Farcaster	`api.neynar.com/v2/farcaster/feed/trending`	API key	Free 300 req/min
Hacker News	`hn.algolia.com/api/v1/search`	None	Free
CoinGecko	`api.coingecko.com/api/v3/search/trending`	Optional	Free 30 req/min
DeFi Llama	`api.llama.fi/protocols`	None	Free

Farcaster via Neynar

GET /v2/farcaster/feed/trending          // trending casts
GET /v2/farcaster/user?fid={fid}         // user profile + social data
GET /v2/farcaster/feed/user/{fid}/popular // user's popular casts
GET /v2/farcaster/channel/trending        // trending channels
GET /v2/farcaster/cast/search?q={query}   // search casts
    

09

Privy Authentication

Single SDK. Social + wallet auth. Web and React Native.

Supported Login Methods

Email OTP SMS Google Apple X / Twitter Discord Farcaster MetaMask WalletConnect Coinbase Wallet

Data Returned Post-Auth

interface PrivyUser {
  id: string;
  createdAt: Date;
  linkedAccounts: LinkedAccount[];

  twitter?: {
    subject: string;      // Twitter user ID
    username: string;     // @handle
    name: string;         // Display name
    profilePictureUrl: string;
  };

  farcaster?: {
    fid: number;          // Farcaster ID
    username: string;
    displayName: string;
    pfp: string;
    bio: string;
  };
}
    

⚠

Critical Limitation

SOCIAL GRAPH NOT INCLUDED

Privy returns identity (username, pfp) but NOT social graph (followers, interests, liked posts). Workaround:

• Farcaster: Neynar API with FID → follower graph + activity

• X: Bearer token → public timeline (limited on Basic)

10

Database Schema

7 tables. PostgreSQL. Users, content, and market data.

USERS

id · privy_id · display_name · avatar_url
x_username · x_user_id
farcaster_fid · farcaster_username
created_at · updated_at

USER_INTERESTS

id · user_id (FK)
category [crypto|politics|sports|ai|...]
weight (adjustable via engagement)

TRENDING_TOPICS

id · source · category · title · summary
image_url · source_url · source_data (JSONB)
keywords[] · entities[]
relevance_score · expires_at

CACHED_MARKETS

id · source [polymarket|hyperliquid]
external_id · market_type [prediction|perp]
title · category · current_price
volume_24h · open_interest · metadata (JSONB)

TOPIC_MARKET_MATCHES

id · topic_id (FK) · market_id (FK)
match_score (0-1)
match_reason (LLM explanation)

FEED_ITEMS

id · user_id (FK) · topic_id (FK)
position · seen · engaged
created_at

SCRAPED_LINKS

id · url (unique) · content_type [tweet|article] · title · body_text · author · keywords[] · entities[] · summary

11

API Endpoints

Auth & Users

Method	Endpoint	Auth	Purpose
POST	`/api/auth/verify`	—	Verify Privy JWT, upsert user
GET	`/api/users/me`	Bearer	Current user + interests
PUT	`/api/users/me/interests`	Bearer	Set interest categories
POST	`/api/users/me/enrich`	Bearer	Fetch social data, infer interests

Link Drop — Scrape & Match

Method	Endpoint	Auth	Purpose
POST	`/api/scrape`	Optional	Scrape URL → text, keywords, entities
POST	`/api/match`	Optional	Keywords/entities → matched markets
POST	`/api/analyze-link`	Bearer or Skill Key	Combined scrape + match; reserves one quota unit before analysis (default 5/day UTC)

Feed

Method	Endpoint	Auth	Purpose
GET	`/api/feed?page=&limit=&category=`	Bearer	Personalized feed with markets
GET	`/api/feed/refresh`	Bearer	Force-refresh

Markets — Direct Query

Method	Endpoint	Purpose
GET	`/api/markets/polymarket/trending`	Top predictions by volume
GET	`/api/markets/polymarket/search?q=`	Search predictions
GET	`/api/markets/hyperliquid/assets`	All perps with prices
GET	`/api/markets/hyperliquid/:coin`	Detailed perp data

12

Background Jobs

5 pipelines. BullMQ on Redis. Continuous charge cycle.

📡

Poll

*/15 * * * *

🧹

Dedup

fuzzy match

🏷️

Classify

gpt-4o-mini

✍️

Summarize

+ image gen

🎯

Match

markets

01. Trend Polling — Every 15 Min

Parallel fetch from all sources:

X API Search NewsAPI Reddit /hot Farcaster Trending CryptoPanic CoinGecko Google Trends

Dedup via fuzzy title match (Levenshtein < 0.3). Classify via GPT-4o-mini. Store in trending_topics.

02. Summarize + Image Gen

GPT-4o writes 2-3 sentence summary. Extracts keywords. DALL-E 3 generates editorial illustration per feed card.

03. Market Matching

Search Polymarket by keywords/entities. Map HyperLiquid assets via entity→ticker lookup. LLM scores relevance 0-1. Keep > 0.5.

04. Price Refresh — Every 5 Min

Batch update cached market prices. Polymarket via CLOB endpoint. HyperLiquid via single allMids call.

05. Feed Generation — Every 30 Min

Per-user: filter by interests, join market matches, rank, apply diversity rules.

MS

MiroShark Integration

Capacitr is built on / using MiroShark as its social and prediction-market simulation layer. When a user finds a tradeable narrative in Capacitr, the app can convert that market thesis into a MiroShark-native simulation prompt, submit it through MiroShark's x402 run endpoint, track the upstream job, and surface the hosted report when the simulation completes.

🦈

How Capacitr Uses MiroShark

Scenario simulation layer

Prompt enrichment: Capacitr expands terse trade cards into richer social-market scenarios, preserving source context such as scraped article text, entities, keywords, tickers, categories, trade direction, position vehicle, horizon, and risk framing.
x402 job launch: enriched prompts are sent to MiroShark's paid /x402/run flow. Capacitr stores the upstream payload and IDs so simulations can be audited and debugged.
Status tracking: Capacitr polls MiroShark status JSON while keeping user-facing UI focused on progress, unread report notifications, and completed simulation summaries.
Report handoff: when MiroShark returns a hosted share_url, Capacitr links users to that report instead of exposing raw status endpoints as final results.

Input to MiroShark

Structured market thesis, source context, entities, sentiment, proposed trade type, direction, size, and horizon.

TRADE THESIS SOURCE CONTEXT SCENARIO PROMPT

Output to Capacitr

Simulation status, hosted report URL, summary metadata, and notification state for returning users.

STATUS JSON SHARE URL REPORT SUMMARY

Why MiroShark Fits Capacitr

Capacitr's discovery layer answers "what can I trade from this signal?" MiroShark adds a second step: "how might this narrative evolve once agents, communities, and prediction-market participants react?" That makes the integration useful for pre-trade context, narrative stress testing, and turning raw market matches into richer decision support.

Capacitr signal → enriched market thesis → MiroShark x402 run → status polling → hosted simulation report → Capacitr notification / report link

13

LLM Strategy

Task	Model	Tokens	Frequency	Cost/Cycle
Topic categorization	`gpt-4o-mini`	~200	100/cycle	$0.003
Summary generation	`gpt-4o`	~500	100/cycle	$0.250
Keyword extraction	`gpt-4o-mini`	~200	100/cycle	$0.003
Market scoring	`gpt-4o-mini`	~300	500/cycle	$0.015
Image prompts	`gpt-4o-mini`	~200	50/cycle	$0.001
Link analysis	`gpt-4o`	~1000	on-demand	$0.005/req

COST SUMMARY

$0.30/cycle · $7.20/day @ 4 cycles/hr · ~$10-15/day with caching

14

Matching Algorithm

Polymarket Matching Pipeline

01. Keyword Search

Top 5 keywords → GET /public-search?q={keyword}

02. Entity Search

Named entities (person, org, event) → Polymarket search

03. Deduplicate

Remove duplicates by conditionId

04. Filter

Active only. Volume > $1,000

05. LLM Score

GPT-4o-mini rates relevance 0-1 per topic-market pair

06. Return Top 5

Score > 0.5, sorted descending

HyperLiquid Matching

// Entity → Ticker lookup + LLM fallback
"bitcoin" → BTC   "ethereum" → ETH   "solana" → SOL
"nvidia"  → NVIDIA "tesla"   → TSLA  "gold"   → GOLD

// "Bitcoin ETF" topic → match BTC perp
// Unknown entity → ask LLM to extract ticker
    

Feed Ranking Formula

score = category_weight × 0.4   // user interest strength
      + relevance_score × 0.3   // trending intensity
      + freshness       × 0.2   // decay over 24h
      + market_volume   × 0.1   // tradeability signal
    

15

Implementation Phases

CURRENT STATE // SHIPPED

Unified Discovery + Market Context (Live)

app.capacitr.xyz feed + market surfaces are live with saved search, sharing pages, and cross-source market context. markets.capacitr.xyz remains discovery-forward while execution moves toward app-native flows.

App Feed Saved Search Sharing Polymarket + Hyperliquid Context

IN PROGRESS // WEEKLY PRIORITY

Execution Foundation in app.capacitr.xyz

Privy wallet onboarding, funding rails, Hyperliquid builder approval flow, and first app-native order execution path.

Simulated execution mode, positions/PnL tracking, and clear live vs paper separation to improve onboarding and conversion confidence.

AFTER STABILIZATION

Venue Expansion + Automation

Polymarket execution path hardening, multi-venue orchestration, and deeper ranking/attribution feedback loops.

16

Cost Projections

Prototype — ~$130/mo

X API Basic

$100

OpenAI

$30

Polymarket

$0

HyperLiquid

$0

Production @ 10K Users — ~$1,328/mo

OpenAI (LLM)

$450

NewsAPI

$449

X API

$100

Neynar

$99

SerpAPI

$75

DALL-E 3

$60

Infra

$95

17

Key Tradeoffs

LLM vs. Keyword Matching

Keyword search hits too many false positives. "Trump" matches dozens of markets. LLM understands a tariff tweet → tariff market, not primary market. Cost: ~$0.015/500 pairs.

Polling vs. Streaming

X streaming = Enterprise ($$$). Most sources lack streaming. 15-min poll is fresh enough. Exception: HL WebSocket for live prices.

Jina Reader vs. Self-Hosted

Zero infra. Handles JS/anti-bot. Free. Production: add Firecrawl backup, self-host Readability for speed.

Materialized vs. On-Demand Feed

Pre-compute per-user feeds. Lower latency, enables push notifications. Trade: up to 30 min stale.

18

End-to-End Example

Link Drop Flow

01. User Pastes Link

https://x.com/elonmusk/status/1234567890

02. Detect & Scrape

X post detected → fetch tweet → "Spoke with NVIDIA CEO about AI chip exports to China"

03. LLM Analysis

Keywords: nvidia, AI chips, China · Entities: NVIDIA, China, Musk · Sentiment: neutral

04. Polymarket Search

"US ban AI chip exports to China by 2025?" — 42% Yes ($2.1M vol)

05. HyperLiquid Match

NVIDIA-PERP: $145.20, +0.003% funding, $48M vol

06. Render Card

Feed card: tweet summary + prediction markets + perps

Sample Feed Card

🏦

BUSINESS // NEWSAPI // 12 MIN AGO

Federal Reserve Signals Rate Cut in September

The Fed indicated a likely 25bps rate cut at the September meeting, citing cooling inflation and labor market normalization.

Fed Cuts Rates in Sept?

88% YES · $4.2M vol

BTC Reaches $80K EOY?

34% YES · $1.8M vol

BTC-PERP

$67,400 · +0.01% funding

19

Future Enhancements

💰

One-Tap Trading

Trade matched markets via Privy embedded wallet

👥

Social Feed

Follow others. See what markets they watch.

🔔

Price Alerts

"Notify when this market moves 10%+"

🤖

AI Research Agent

Deep-dive any topic with citations

📊

Portfolio Tracking

Connect wallet. Show P&L tied to feed.

🎨

Creator Economy

Create and share market analysis cards

Trend-to-Trade Execution System

Thesis: Signal-to-Position

The Problem

The Insight

The Product

Who It's For

The Moat

Product Scope

System Architecture

Tech Stack

Polymarket Trading Integration Plan (App-Native)

Execution Architecture

Builder Program Requirements (Engineering Checklist)

Readiness State Machine (User UX)

Capacitr Services to Implement

API Surfaces Used

Rollout Plan

Quotient Intelligence

What It Does

API Endpoints

Response Fields

Integration Flow

Used In

Hyperliquid Trading Integration Plan (App-Native)

Execution Architecture

Builder Code Semantics (What Engineering Must Implement)

Readiness State Machine (User UX)

Backend Services (Capacitr App)

API Surfaces Used

Rollout Plan

Options Platforms

Deribit — Gold Standard (~90% of Crypto Options Volume)

Aevo — Hybrid DEX (OP Stack L2)

Derive — DEX (Formerly Lyra V2)

Why Options Matter for Capacitr

Other Platforms Evaluated

X / Twitter API

Link Drop: Tweet Extraction

News & Article Scraping

News Aggregation APIs

Article Content Extraction

Trend Sources

Farcaster via Neynar

Privy Authentication

Supported Login Methods

Data Returned Post-Auth

Database Schema

API Endpoints

Auth & Users

Link Drop — Scrape & Match

Feed

Markets — Direct Query

Background Jobs

01. Trend Polling — Every 15 Min

02. Summarize + Image Gen

03. Market Matching

04. Price Refresh — Every 5 Min

05. Feed Generation — Every 30 Min

MiroShark Integration

Why MiroShark Fits Capacitr

LLM Strategy

Matching Algorithm

Polymarket Matching Pipeline

HyperLiquid Matching

Feed Ranking Formula

Implementation Phases

Cost Projections

Prototype — ~$130/mo

Production @ 10K Users — ~$1,328/mo

Key Tradeoffs

End-to-End Example

Link Drop Flow

Sample Feed Card

Future Enhancements