Analysis updated 2026-05-18
Study seven example trading strategies that combine AI classification with deterministic execution rules.
Backtest trading strategies offline against synthetic market data with zero API keys.
Learn a staged, safety-checked process for moving a trading bot from paper trading to live capital.
| regardo911/seven-ai-trading-bots | 0xallam/my-recipe | 0xhassaan/nn-from-scratch | |
|---|---|---|---|
| Stars | 0 | — | 0 |
| Language | Python | Python | Python |
| Last pushed | — | 2022-11-22 | — |
| Maintenance | — | Dormant | — |
| Setup difficulty | easy | moderate | moderate |
| Complexity | 4/5 | 2/5 | 4/5 |
| Audience | developer | general | developer |
Figures from each repo's GitHub metadata at analysis time.
Runs fully offline on synthetic data with no API keys required to try the demo.
This repository is the companion code for a book called Use Claude to Build 7 AI Trading Bots. It contains seven example trading strategies built on a shared framework, where the AI model Claude helps interpret information like news or earnings call transcripts, while plain, predictable Python code makes the actual buy or wait decisions. The idea is that Claude reads and classifies things, and ordinary code, not the AI, handles the math and the trading rules. The project is clearly labeled as educational software, not financial advice. Trading carries real risk of loss, and the built in backtests run on made up sample data rather than real market history. Simulated paper trading mode is turned on everywhere by default, and switching to real trading requires a deliberate change in the code itself, not just a setting or command line flag, specifically to prevent accidentally trading with real money. The seven bots cover different strategies: a trend following bot using price breakouts, a pairs trading bot based on two related assets moving together, an earnings reaction bot where Claude reads transcripts and scores them, a news reaction bot that classifies headlines, an options flow bot, a funding rate arbitrage bot for cryptocurrency exchanges, and a portfolio allocator that spreads risk across the others. Only two of the seven, the earnings and news bots, call Claude while actually trading, and the estimated cost of each of those calls is factored into the backtests so results reflect real expenses. Getting started needs no API keys, no brokerage account, and no internet access beyond installing Python packages, since everything runs offline using made up test data. The project includes a guided quickstart with a series of commands that check the setup, walk through example decisions, and run the full test suite. All configuration settings, including the Claude API key, are optional, with sensible offline stand ins used when they are missing. The book itself lays out a cautious, staged path for anyone who eventually wants to trade with real money, moving from paper trading up to full capital only gradually, with specific checks at every stage.
Companion code for a book teaching how to build seven AI-assisted trading bots, where Claude classifies data and deterministic Python executes trades.
Mainly Python. The stack also includes Python, Claude API, Anthropic.
Use freely for any purpose, including commercial use, as long as you keep the copyright notice.
Setup difficulty is rated easy, with roughly 5min to a first successful run.
Mainly developer.
This repo across BitVibe Labs
Verify against the repo before relying on details.