Use Cursor Agent to draft a trading bot script

A commenter had Cursor Agent with composer-1 inspect a codebase and make large coding changes for a personal trading bot.

Done with Cursor· Intermediate

What they did
They used Cursor Agent to automate the coding work for a personal trading script. The agent reviewed most of the existing code to collect context, then carried out a very large edit on its own, including one request that changed around 1,000 lines. The person mainly decided when to let it run and then checked the result after it finished.
How it went
They said the output was pretty good, but costly. One request used about 4.5 million uncached tokens, and they did not report any trading performance or deployment success.
Tips
This kind of workflow can move fast on large code changes, but token usage can balloon quickly. Keep the bot in backtest or paper-trading mode until you review the code and approve anything that would place live orders.

Try it yourself with Cursor

You are helping me build a personal trading bot in [language] for [asset or ticker]. First, inspect the relevant parts of the codebase and summarize the trading logic, data flow, and any missing pieces before editing anything. Then implement the bot so it can run in [paper-trading/backtest] mode using [data source or broker API], and add a backtest for [time period] plus a short README with run instructions. Do not place real trades or connect to live order submission yet. Before you add any code that would send orders, submit trades, delete files, or change live-money settings, stop and ask for my approval.

Fill in the [brackets]. Agents can make mistakes, so review anything before it is sent, paid or deleted.

Get 3 agent uses like this every morning

With the day's top AI news. Free, and one click to unsubscribe.