Technicals
Execution guides for prediction market agents. How to set up, size positions, manage risk, and run 24/7 trading operations.
Kalshi vs Polymarket: A Developer's Comparison of APIs, Orderbooks, and Liquidity
A data-driven comparison of the two largest prediction markets from a developer and trader perspective.
Build a Prediction Market Research Crew with CrewAI + SimpleFunctions
Use CrewAI multi-agent architecture to build a prediction market research and trading team.
Build a Prediction Market Agent with LangChain + SimpleFunctions
Step-by-step guide to building an autonomous prediction market agent using LangChain and SimpleFunctions API.
Connect Claude Code to Prediction Markets: MCP Server Setup Guide
One command to give your AI agent access to Kalshi and Polymarket data.
How to Scan Prediction Market Orderbooks: Spread, Depth, and Liquidity Analysis
A practical guide to reading and analyzing orderbook data from Kalshi and Polymarket.
Heartbeat architecture: how to monitor 50+ prediction markets in real-time
Inside the 10-step monitoring loop that watches Kalshi, Polymarket, and traditional markets on a 15-minute cycle for $0.61/thesis/day
Automating thesis lifecycle: create, monitor, evaluate, trade
The full agentic loop in code: six API calls from raw thesis to executed trade, with complete request/response examples.
Piping prediction market signals into your existing trading system
Three integration patterns for teams that already have infrastructure: cron polling, agent middleware, and thesis-as-filter.
Connecting your AI agent to prediction market data in 5 minutes
Three integration paths — MCP, REST, CLI — each with working code you can ship today.
Quantitative Orderbook Analysis for Prediction Markets: Signals, Metrics, and Code
The practical companion to orderbook theory. Concrete formulas, real data, and working code for extracting actionable signals from prediction market orderbooks — depth ratios, coherence checks, liquidity scoring, and slippage estimation.
Automated Prediction Market Trading: Architecture and Cost Breakdown
The real numbers behind running an automated prediction market system. LLM costs per evaluation, Tavily search budgets, Kalshi fees, total cost per thesis, and when each interface (CLI, API, MCP, agent) makes sense.
Your First Prediction Market Trade: End-to-End CLI Walkthrough
From npm install to your first filled order. Every command, every output, every decision point. The definitive zero-to-first-trade tutorial for prediction market trading with the SimpleFunctions CLI.
Understanding Prediction Market Orderbooks: A Complete Guide
How to read a Kalshi orderbook from the raw API response to executable trading decisions. Covers yes_dollars vs no_dollars, bid/ask computation, slippage algorithms, depth analysis, and liquidity scoring.
The Evaluation Cycle: How Automated Thesis Monitoring Works
Inside the heartbeat loop that powers continuous thesis monitoring: news scanning, price refreshes, milestone checks, LLM evaluation, confidence updates, and smart scheduling that adapts to market volatility.
Edge Calculation in Prediction Markets: From Theory to Execution
Theoretical edge means nothing if you can't execute it. This article covers the full edge stack: theoretical edge, spread cost, slippage, depth-adjusted edge, and when to walk away from a trade entirely.
How Causal Tree Decomposition Works in Prediction Market Trading
The core methodology behind structured prediction market analysis: decompose a thesis into a tree of testable sub-claims, assign probabilities, propagate them, and find where the market disagrees with your model.
How to Backtest a Prediction Market Strategy
Binary outcomes and clear settlement make prediction markets unusually good for backtesting. Here is how to build a calibration curve, avoid common pitfalls, and use settlement data to track realized returns.
Reading Prediction Market Orderbooks: Liquidity, Spread, and When to Enter
Price tells you what the market thinks. The orderbook tells you how confident it is, how much it will cost you to trade, and whether the price can be trusted at all.
Building Real-Time Prediction Market Alerts with Webhooks
Polling wastes resources and misses events. Here is how to build a webhook-based alert system for prediction market price moves, confidence shifts, and strategy signals.
Cross-Venue Edge Detection: Kalshi vs Polymarket
The same event priced differently across venues. Why it happens, how to detect it programmatically, and why thesis-informed cross-venue trading beats pure arbitrage.
Kalshi API Quick Start: JavaScript and Python in 5 Minutes
From zero to your first API call in both languages. Authentication, market data, placing orders — then how SimpleFunctions collapses it all into one command.
Running a 24/7 Trading Agent: Architecture, Costs, and What to Watch
The real operational picture. Heartbeat cron jobs, Tavily news search costs, OpenRouter LLM spend, Kalshi API quirks, and why this whole system runs for ~$100/month vs. a quant fund's $50K/month data bill.
From Thesis to Execution: How SimpleFunctions Manages the Full Trading Loop
The complete walkthrough. From "I think Iran will cause a recession" to "my agent detected CPI data at 3am and updated the causal tree." Every step is a product feature wrapped in a real trading decision.
Position Sizing for Prediction Markets: Kelly Criterion Meets Causal Models
Prediction market contracts have a $1 cap, binary settlement, and clear expiry. Kelly criterion applies directly — but the critical input is your estimated true probability. Here's how causal model confidence feeds into the formula.
Setting Up Your First Prediction Market Agent with SimpleFunctions
From zero to a running agent in 15 minutes. MCP configuration, your first scan, your first thesis, your first edge — with real screenshots and every decision point explained.
5 Patterns That Kill Prediction Market Traders (and How Agents Fix Them)
Not a textbook. These are real trading mistakes every prediction market trader makes — anchoring, news overload, asymmetric fear, frequency illusion, confirmation bias — and how an automated agent eliminates each one.