SX Bet vs
SimpleFunctions.

SimpleFunctions indexes Kalshi and Polymarket, covering 48K+ active contracts across political, economic, sports, and financial event categories. SX Bet operates a separate on-chain P2P exchange on the SX Network Arbitrum rollup, primarily sports-focused. The two platforms serve different market ecosystems — SF does not expose SX Bet markets, and SX Bet does not expose Kalshi or Polymarket contracts. If you need both, you would need both APIs.

POST /api/thesis/create with a plain-language sentence. SF decomposes it into a causal tree of testable sub-claims, assigns probabilities to each node, and automatically scans Kalshi and Polymarket for contracts that represent tradeable edges. A continuous evaluation heartbeat runs news scans, price refreshes, milestone checks, LLM evaluations, and confidence updates on each active thesis. You can inject new signals at any time via /api/thesis/{id}/signal. Public theses are forkable by other users.

Yes. SF ships a 56-tool MCP server. Add it with: `claude mcp add simplefunctions --url https://simplefunctions.dev/api/mcp/mcp`. It works with Claude Code, Cursor, and any MCP-compatible client. Tools cover market search, thesis management, portfolio monitoring, orderbook depth, cross-venue arbitrage pairs, calibration data, and more — all accessible to the agent without any additional authentication for read operations.

SX Bet offers REST and WebSocket APIs for programmatic access to its P2P sports orderbook, making it usable for bots that target sports markets on its Arbitrum rollup. SF's Portfolio Autopilot is a different class of tool: a 1M-context LLM-driven agent with 13 data sources and a 7-gate risk cascade (kill switch, position limits, drawdown gate, regime check, liquidity score, conviction threshold, concentration limit) that operates on Kalshi and Polymarket. The two are not substitutes — they target different venues and different design philosophies.

GET /api/calibration returns SF's own Brier scores broken down by venue, category, and price bucket, computed over the past 90 days. Current figures: Kalshi 0.20, Polymarket 0.12 on T-24h price. Brier score measures mean squared error of probabilistic forecasts — lower is better, 0.25 is chance on a 50/50 market. SF publishes this audit publicly so you can verify the accuracy of its data layer before building on it, and re-verify it at any time with a curl call.

SF pre-computes six indicators across the full 48K+ active contract universe: IY (implied yield — annualised return if the market resolves YES), CRI (cliff risk index — probability of sharp price movement near settlement), LAS (liquidity availability score), EE (event overround — the market-maker margin embedded in prices), τ-days (time to settlement), and a regime label (adverse-selection classification). All are available at /screen without authentication.

SimpleFunctions connects to Kalshi and Polymarket, which use their own settlement mechanisms — Kalshi is CFTC-regulated with USD settlement; Polymarket settles on Polygon. SF itself is a data and agent layer and does not manage custody. It does not connect to the SX Network or Arbitrum. If crypto-native, non-custodial P2P settlement on an Arbitrum rollup is a hard requirement, SX Bet is the appropriate tool and SF is not a substitute.

Portfolio Autopilot uses a 7-gate risk cascade that must be fully cleared before any trade is submitted: kill switch (global halt), position limits (per-contract cap), drawdown gate (portfolio-level loss threshold), regime classification (adverse-selection check), liquidity availability score (LAS gate), conviction threshold (minimum signal strength), and portfolio concentration limit. The agent draws from 13 data sources and a 1M-context LLM. All gates are configurable; the system is designed to operate without per-trade human approval while remaining auditable.