Run a paper-trading portfolio with a morning research routine
A tutorial for building a Claude Code agent that manages an Alpaca paper-trading portfolio with a scheduled morning research routine over a watchlist.
Done withClaude Code
- What they did
- The guide sets up a project with a CLAUDE.md rulebook, a watchlist.json, and Python helper scripts for Alpaca data and orders. Three scheduled Claude Code routines run on weekdays: morning research at 9:45, trading at 10:00, and an end-of-day journal. The agent checks market status, pulls bars and news, places limit orders, and writes a markdown journal.
- How it went
- It is a tutorial and reports no results from running it. It says paper results will look better than live ones because of slippage and partial fills, and that the agent executes a strategy rather than predicting markets.
- Worth knowing
- The author says Claude inference for three daily sessions typically costs a few dollars a month, rising with bigger watchlists. Cap tool calls and trim data to control it.
Try it yourself with Claude Code
Build me an agent that manages a paper-trading portfolio on Alpaca using my paper API keys stored in [env file or location]. Each morning at [time] it should research my watchlist of [tickers], write a short note on each, and propose trades with reasoning, then log everything to [file or folder]. Use paper trading only, never live. Pause and show me the proposed orders before placing any, and stop once a morning run completes end to end.
Want help getting this working? Join the Vibe Coding Academy ↗ Run by Alex Finn, who also runs this site.
Source: mindstudio.ai · Undated
Related use cases
Turn tickets into merged pull requests
Pulls tasks from GitHub Issues or Linear, then runs Claude Code or Codex in isolated Kubernetes pods to carry each one through to a PR.
Build a thyroid episode warning model from wearable data
Claude Code was given 9.5 years of Apple Watch and Whoop data and built an ML model that backtested as warning of a prior episode before labs confirmed it.
Turn GitHub issues into draft pull requests with long-running agents
The author used Dots for read-heavy long-horizon tasks like reading GitHub issues and making PRs, which were typically poor but gave a starting point.
Shop for umbrella insurance through your personal agent
Personal AI agents use Coverage Cat's Agent API/MCP to size umbrella coverage and compare carrier options with price ranges, backed by a licensed broker.
Custom agent
Fight a mattress warranty claim by email
A user's agent emailed back and forth with the mattress company's agent, citing laws and pushing aggressively, to resolve a warranty claim.
Custom agent
File an income tax return through the tax portal
Claude Desktop agent filed the user's ITR-1: cross-checked salary against AIS, handled an employer switch, fixed validation errors, and submitted.
Get 5 like this every morning
The best things people got an AI agent to do, each with the prompt to try it.