# Octagon CLI | Trade prediction markets from your terminal AI-powered Kalshi trading CLI. Natural-language agent, edge detection vs. live order book, half-Kelly sizing, diversified baskets — all from your terminal. Powered by the Octagon Research API. Prediction Markets CLI · Powered by Octagon # Trade prediction markets from your terminal. An AI agent that researches every Kalshi market, finds edge against the live order book, and executes trades — all behind a single CLI. Half-Kelly sizing. Five-gate risk engine. Diversified basket construction in one call. $ bunx kalshi-trading-bot-cli@latest Copy [ View on GitHub ](https://github.com/OctagonAI/kalshi-trading-bot-cli) [ Read the guide → ](https://github.com/OctagonAI/kalshi-trading-bot-cli/blob/main/GUIDE.md) 21 Octagon endpoints wired in 5 Risk gates before each fill 1 Command to build a basket ~/kalshi-bot — bun · kalshi $ bunx kalshi-trading-bot-cli@latest Welcome to Kalshi Trading Bot CLI Type help for commands, or just ask a question. > > search crypto KXBTC-26APR-B95000 Bitcoin above $95k by Apr 30 $0.58 12,841 KXBTC-26APR-B100000 Bitcoin above $100k by Apr 30 $0.31 8,203 KXETH-26APR-B2000 Ethereum above $2k by Apr 30 $0.72 5,419 3 markets found > analyze KXBTC-26APR-B95000 Octagon Research Report — KXBTC-26APR-B95000 ──────────────────────────────────────── Model Probability 72% Market Price 58% Edge +14.0% (very_high confidence) Top Drivers 1. Bitcoin ETF inflows accelerating impact: high 2. Halving cycle momentum impact: high 3. Macro risk-on sentiment impact: moderate Kelly Sizing Recommended: 3 contracts YES at $0.58 Risk gates: ✓ Kelly ✓ Liquidity ✓ Correlation ✓ Concentration ✓ Drawdown What it does ## A research analyst, risk manager, and order router — in one binary. ### AI agent loop Natural-language queries trigger up to 10 tool-use iterations across Kalshi and Octagon endpoints. Bring your own LLM key — OpenAI, Anthropic, Google, xAI, OpenRouter, or local Ollama. ### Edge detection Octagon's model probability minus the live order book = edge. Surfaced per market, ranked across the universe, and sized with half-Kelly behind five risk gates before a dollar is risked. ### Diversified baskets One-call basket construction with thematic-cluster caps and pairwise correlation caps. Backtest a theme, Kelly-size it, validate cluster overlap — all server-side. Live commands ## From "I have an idea" to a filled order in four lines. 01 ### Discover Server-side full-text + semantic search across the Kalshi universe. > search "bitcoin price" --category crypto --min-volume 10000 Ticker Title Last Volume KXBTC-26APR-B95000 Bitcoin above $95k by Apr 30 $0.58 12,841 KXBTC-26APR-B100000 Bitcoin above $100k by Apr 30 $0.31 8,203 KXETH-26APR-B2000 Ethereum above $2k by Apr 30 $0.72 5,419 3 markets found 02 ### Analyze Per-event research report with model probability, drivers, and Kelly sizing. > analyze KXBTC-26APR-B95000 Octagon Research Report — KXBTC-26APR-B95000 ──────────────────────────────────────── Model Probability 72% Market Price 58% Edge +14.0% (very_high confidence) Top Drivers 1. Bitcoin ETF inflows accelerating impact: high 2. Halving cycle momentum impact: high 3. Macro risk-on sentiment impact: moderate Kelly Sizing: 3 contracts YES at $0.58 Risk gates: ✓ Kelly ✓ Liquidity ✓ Corr ✓ Concentration ✓ Drawdown 03 ### Build a diversified basket One call: candidate universe → cluster caps → correlation caps → Kelly sizing. > basket build --label fed,cpi,fomc,gdp,jobs -n 5 --max-corr 0.4 Basket built · 5 legs · max pairwise corr 0.32 ──────────────────────────────────────── KXFEDDECISION-MAR YES $0.61 Kelly $122 cluster: fed-policy KXCPIYOY-FEB NO $0.44 Kelly $98 cluster: inflation KXFOMC-DOTS-25Q4 YES $0.55 Kelly $87 cluster: fed-policy KXU3-MAR NO $0.39 Kelly $76 cluster: labor KXGDPNOW-Q1 YES $0.52 Kelly $64 cluster: growth Total exposure: $447 · expected return: +18.6% 04 ### Execute Confirm with one keystroke. Limit price in cents, side defaults to YES. > buy KXBTC-26APR-B95000 3 58 Confirm order? [y/n] y ✓ Order placed: BUY 3 YES @ $0.58 Order ID: abc-123-def Cash: $487.26 · Exposure: $1.74 · Open positions: 1 How it connects ## A thin terminal in front of a deep intelligence pipeline. You Your terminal Bun runtime · macOS, Linux, WSL → CLI kalshi-trading-bot-cli AI agent loop · slash commands → Octagon Octagon Research API 21 typed endpoints · model probabilities → Exchange Kalshi Order entry · positions · fills ### Which command hits which endpoint search · search edge → /kalshi/markets · /markets-with-edge similar · peers → /kalshi/markets/similar · /cluster-peers clusters → /kalshi/clusters · /behavioral-clusters correlate → /kalshi/markets/correlations basket build · size · backtest → /kalshi/baskets/* (5 endpoints) events · analyze → /prediction-markets/events Setup in 60 seconds ## No clone. No install. Just bunx. 01 ### Install Bun Bun ≥ 1.1 is required — the CLI uses bun:sqlite and runs .tsx directly. curl -fsSL https://bun.com/install | bash 02 ### Get a Kalshi API key From [kalshi.com](https://kalshi.com) — you'll need an API key plus an RSA private key. Use demo mode for paper trading. 03 ### Optional: Octagon key An [Octagon](https://app.octagonai.co) key unlocks AI edge analysis, semantic search, clusters, and basket construction. 04 ### Run the CLI The setup wizard handles the rest — keys land in ~/.kalshi-bot/.env. bunx kalshi-trading-bot-cli@latest Full command reference Click to expand · 40+ slash commands ### Discovery search [query] | Full-text or semantic search across markets | search edge | Rank markets by model edge from Octagon's latest run | similar | Semantic neighbors via embeddings | clusters [--label X] | Browse thematic K-means clusters | clusters --behavioral | Behavioral clusters by 30-day return vectors | peers | Markets in the same cluster as a ticker | events / events | Octagon events list + outcome ladder | catalysts upcoming --days N | Markets closing in the next N days, grouped by week | ### Baskets & portfolio correlate … | Pairwise Pearson correlation matrix | basket build | Diversified basket with cluster + correlation caps | basket size | Fractional Kelly sizing for picked legs | basket backtest | Total return / Sharpe / max DD on a basket | basket validate | Cluster overlap · pairwise corr · calendar clashes | themes report | 25-theme dashboard with SEO + liquidity | themes audit | Flag dead themes (high SEO, zero volume) | ### Trading & account analyze | Deep analysis: edge, drivers, Kelly sizing | watch | Live price and orderbook feed | buy / sell [price] | Place a confirmed order in cents | cancel | Cancel a resting order | portfolio | Positions, P&L, risk snapshot | backtest | Model accuracy scorecard + live edge scanner | setup / init / clear-cache | Wizard, fresh install, cache reset | Install ### Try it now — zero install No clone, no setup. The wizard prompts you for keys on first launch and writes them locally to ~/.kalshi-bot/.env. $ bunx kalshi-trading-bot-cli@latest Copy Build your own ### Or wire the API into your own stack Our CLI runs entirely on the public Octagon REST API. 21 typed endpoints — search, clusters, correlations, baskets, edge. Bring your own runtime. [Read API docs →](https://octagonai.co/docs/guide/rest-api/prediction-markets-kalshi) [Open Playground ↗](https://app.octagonai.co/playground) ## Attribution Policy When quoting, summarizing, or reproducing Octagon AI content, attribute it to Octagon AI and link to the Octagon source URL: https://octagonai.co/cli If a specific page was used, cite that page rather than only the site homepage.