A Claude Code agent twice weekly covers about 15 positions and a watchlist with news synthesis and bull/bear cases, while a human sets risk limits and approves trades.
A Claude Code agent twice weekly covers about 15 positions and a watchlist with news synthesis and bull/bear cases, while a human sets risk limits and approves trades.
An anonymous operator gave a Claude Code agent (Opus 4.6) a one-line instruction to beat the S&P 500 on a $50,000 real-money account. The agent produces research on about 15 positions plus a candidate list, covering news, bull and bear cases and scenarios, twice a week. The operator wrote the risk rules (size limits, a turnover ceiling, a written thesis that can be proven wrong) and says the final picks are theirs.
The operator reports +19.04% against +12.24% for the S&P 500 from March 3 to August 4, 2026, with $27 million copying the trades. The article says five months is too short to show skill, and the figures are unaudited.
Cost was only described as more than zero and less than an analyst; the full reply is cut off. The author notes this is a marketing funnel for a copy-trading product.
Act as my research analyst for these positions: [tickers] and this watchlist: [tickers]. Twice a week, summarize recent news for each, give a bull case and a bear case, and flag anything that breaks my risk limits: [limits]. Do not place or suggest executing trades without my approval. You're done when I have a first report in [file or email format].
An app built with Claude Code extracts vendor, date, total, tax and line items from receipts, including foreign-currency ones at historical exchange rates, for freelancers and small businesses.
A tenant used ChatGPT and Gemini to review a landlord's apartment deposit dispute, identify legal violations, and avoid losing $4,200.
The poster used an AI assistant to summarize a denial PDF to draft an appeal, but it invented an appeal deadline.