Community AI Notes | He built a US-stock watchlist app with AI — all his pre-market homework in one place

Pre-market homework — check quotes, flip through the macro calendar, look up earnings, read the options chain

The data is scattered across different sites — hunting around for it is a headache 🤕

Turns out, one app can handle it all!

Community member @乐某 used AI and the Longbridge Open Platform to turn this whole workflow into one app — "Market Compass" —

pre-market, intraday and after-hours information, all pulled into one screen.

👉 Original post 👉 I built an iOS US-stock workbench with Longbridge and AI (@乐某 )

👉Open source 👉 https://github.com/Lemou-Memo/market-compass

App overview: pre-market to after-hours, seven screens on one thread

From the market backdrop down to single-stock structure, each of the seven screens does its own job:

Data sources: watchlist groups, quotes and trading-session data, candlesticks, financials and consensus estimates, the options chain and each contract's open interest (OI) all come from the Longbridge Open Platform; external sources fill in just a few things — interest rates, credit spreads, and the financial conditions index — pulled from FRED, the US Treasury, and similar sources.

Where the real work went: the data

✅ Pull everything at the same moment. The Longbridge Open Platform returns multiple tickers in one call, so the macro indicators on the home screen are a cross-section from the same instant — that's the only way a side-by-side comparison is actually valid. The sector-rotation screen scales up that same logic: three windows — 1, 5, and 20 days — sit side by side, which is the only way to tell whether today is a broad rally or just a few sectors leading.

✅ Bake the data conventions into the code. Year-over-year only ever compares comparable reporting periods, and a swing from loss to profit never gets expressed as a percentage; the change is always calculated against the last regular-session close, tagged with whether it's overnight, pre-market, intraday, or after-hours — because the Longbridge Open Platform's quote API returns each session as its own quote block, each with its own timestamp and previous close, so there's no guessing which number belongs to which session.

✅ Macro events get archived, not just forecast. "Upcoming" and "released" sit on the same timeline, and every entry keeps its release time, the period it covers, the forecast, the prior reading, and the actual — plus the original rationale. That turns the screen from a preview table into a record you can actually look back on.

Working with AI — the division of labor is what matters

The community member laid out the division of labor clearly:

• He defines the trading scenarios, judges whether data is actually useful, and verifies the financial logic;
• AI looks up API docs, writes the SwiftUI, organizes the data models, implements the indicator calculations, tracks down crashes and network issues, and rearranges screens based on usage feedback.

He owns the scenarios and the logic; AI owns the code and the implementation.

Want to try it yourself? Head to the Longbridge Open Platform!

Quotes, candlesticks, financials, options chains — the Longbridge Open Platform can pull all of it ✅

Want to check data fast? One command with Longbridge CLI;

Want to wire it into your own app? Use Longbridge OpenAPI

Start here — the next "US Stock Compass" is yours to build.

Come play: try it, and submit your own build

👉 Build it yourself: there are more ways than ever to turn an idea into something real — the Longbridge API, the CLI, Skills, the Agent platform — one of them will fit you.

👉 Got an idea? Run with it: tag @LongbridgeAI to submit your build — membership cards and merch are up for grabs.

🎁 Comment to win

Tell us in the comments what new use case you'd want to build~

Every valid entry gets 288 task coins, and we'll also draw 1 winner at random from all valid entries to receive an AI Pro membership card worth SGD 25 (30 days).

Campaign period: from today until 7pm on 17 September 2026. Rewards will be issued within 15 working days of the campaign closing.

Thanks to @乐某 for sharing 🙌

This post covers a research tool independently developed by a community member; quoted content is used with the author's authorization and does not constitute investment advice.