How to Build a Polymarket Trading Bot with Slippage and Position Sizing in Python

·1 views

A developer tutorial on DEV Community outlines how to build a Python-based trading bot for Polymarket, a prediction market platform. The guide covers connecting to Polymarket's CLOB and Gamma APIs to discover active markets and read live order books. It explains how to calculate trading edge by comparing a model's probability estimate against the current market price, and applies fractional Kelly criterion with a hard per-market exposure cap for position sizing. The tutorial also addresses slippage by checking available order book depth before placing trades, recommending resting limit orders over market orders in thin books. The author notes that smarter API polling — scaled to market activity rather than fixed intervals — would improve efficiency in a production setup.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Engineer Builds Custom Transpiler to Migrate 4,000 Chatbot Tests to Playwright

A software engineer at an unnamed company built a custom transpiler pipeline to migrate over 4,000 legacy QA tests to the Playwright framework after joining an internal team tasked with improving the QA pipeline. The existing test suite stored all test cases in a spreadsheet workbook using BDD-style keywords, which caused serious issues including binary file conflicts in Git, no tooling support, and difficulty debugging or running tests in parallel. A full LLM-based migration was ruled out due to token cost constraints and the need for determinism and traceability. Instead, the engineer designed a multi-stage transpiler pipeline in TypeScript — moving from workbook extraction through semantic analysis to code emission — completed within a self-imposed two-week deadline. Key libraries used included xlsx for reading the workbook and ts-morph for programmatically generating the output TypeScript test files.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

Why 'Open to Offers' Listings Kill Pre-Revenue App Sales on Marketplaces

Listing a pre-revenue app with no asking price and an 'open to offers' note is widely cited as the most common reason such sales fail, as it forces buyers — who have the least information — to do the pricing work themselves. Without monthly recurring revenue, there is no standard valuation multiple, leaving buyers with no default answer to the question of what the asset is worth. Industry data from micro-startup marketplace Microns shows roughly 90 pre-revenue projects sold over three years at an average of around $4,300, with its founder noting that buyers rarely pay well for pre-revenue SaaS. Analysts suggest valuing such products based on replacement cost — estimated developer hours multiplied by an hourly rate, then discounted 40–70% to account for unproven demand — with additional assets like a domain, waitlist, or user base pushing the price higher. Research in auction economics supports fixed or floor-price listings over open negotiations, particularly when asset valuation is highly uncertain, as competitive bidding tends to discover price more effectively than one-on-one bargaining.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

How Three Root Files Can Make AI Coding Agents Production-Ready

AI coding tools like Lovable and Cursor are powerful but share a common weakness: they lack persistent context about a project's conventions, architecture, and library choices. Sandbox agents generate fast prototypes but lose all project-specific decisions the moment a session ends, while file-system agents read real codebases but often invent conflicting conventions without proper guidance. The core problem is that language models operate within a sliding context window, forcing them to guess at naming patterns, component APIs, state management, and architectural rules when no reference exists. A practical fix involves placing a small set of structured files — such as CLAUDE.md, .cursorrules, and AGENTS.md — at the repository root, where agents can automatically read them at the start of every session. Written once and maintained alongside the codebase, these files give AI agents the project-specific map they need to bridge the gap between prototype and production.

0 comments Read more at DEV Community

ProgrammingDEV Community ·

DevTools Launches Browser-Side Password Generator Using Web Crypto API

DevTools has released a browser-based password generator that uses the Web Crypto API's crypto.getRandomValues() method to produce cryptographically secure passwords and passphrases. Unlike tools relying on Math.random(), it draws from the operating system's entropy sources, making output resistant to prediction or reproduction. Users can customize password length, character sets, and filtering options, while a passphrase mode offers a memorable yet secure alternative. All processing happens entirely client-side, meaning no password data is ever sent to a server. The tool is part of a suite of over 200 free, no-signup privacy-first utilities available through DevTools.

0 comments Read more at DEV Community