Frequently asked questions.

PredictFlow is a non-custodial trading terminal for prediction markets on Solana. It routes orders through DFlow and adds conditional order types — limit, stop-loss, take-profit, and DCA — that aren't natively available on most prediction-market venues.

PredictFlow surfaces every active event on DFlow's prediction-market network, grouped by category: Sports, Politics, Crypto, Economics. When Kalshi CFTC-regulated markets are available, those appear alongside the rest.

Those venues execute market orders natively but don't offer real conditional order types. PredictFlow adds a client-side monitor that polls the live feed every five seconds and fires limit / stop-loss / take-profit orders when your conditions are met — while keeping custody of funds in your own wallet throughout.

Phantom, Solflare, and Backpack — all via browser-injected providers. No WalletConnect or Ledger yet. Mobile users on Phantom or Solflare can deep-link into the app directly.

No. PredictFlow never reads or transmits key material. Every transaction is decoded, program-whitelist-checked, and simulated locally; your wallet does the actual signing. The terminal only stores your public key (for reconnect on reload) and hashes it before any analytics event.

Four defenses, in order: payload size limit → @solana/web3.js decode → program-whitelist assertion (only System, SPL Token, ATA, ComputeBudget, Memo, and the DFlow router can execute) → RPC simulation. Any failure blocks the signing prompt entirely.

The preflight simulation fails closed — we do not sign a transaction we cannot verify. Configure multiple RPC endpoints (comma-separated in VITE_SOLANA_RPC_ENDPOINTS) and the client will try them in order.

No. Conditional orders run client-side — close the tab and the monitor stops. A persistent banner in the app warns you whenever pending limit/stop-loss/take-profit orders or active DCA strategies exist. A server-side scheduler is on the roadmap.

Every 5 seconds against the configured live-price source (WebSocket preferred; REST fallback on VITE_LIVE_PRICE_URL). Triggers are evaluated synchronously on each tick.

Yes. Once a market has a filled position in local storage, both tabs unlock and you can place them independently. Each has its own trigger price and execution record.

Pick a total budget, a number of purchases, and a frequency. PredictFlow schedules each purchase and runs it through the same signing pipeline as a manual market order. Execution history is kept per strategy; pause/resume/cancel at any time.

No protocol fee today. You pay Solana network fees (sub-cent) and whatever spread DFlow quotes for routing — identical to trading DFlow directly.

Only for CFTC-regulated markets (Kalshi). For those, PredictFlow integrates with Proof — a click-through verification. Other markets need no KYC. Browsing and reading market data are never gated.

PredictFlow never stores identity documents. The Proof flow redirects you off-site and returns a verification status only. When self-hosted, the optional KYC backend is under your own control.

The markets list falls back to a built-in demo catalog so you can still browse the UI. A yellow "Demo mode" banner makes the state unambiguous — no trades can route against the demo data.

For production use, yes. The public mainnet-beta endpoint rate-limits aggressively and will cause preflight failures under normal traffic. Helius, Triton, and QuickNode all work; drop the URL in VITE_SOLANA_RPC_ENDPOINTS.

Yes. The terminal is a static SPA — clone the repo, set your env vars, run npm run build, and drop the output on any static host. Cloudflare Pages, Vercel, Netlify, and Nginx configs ship in the repo.

Set VITE_SENTRY_DSN for error reporting and VITE_ANALYTICS_PROVIDER (+ write key) for product analytics. Both integrations dynamic-import lazily so the base bundle stays small.