Build a stock research agent for updated company reports
They made a stock-research tool that refreshes reports on tracked companies using live market data instead of stale chatbot answers.
- What they did
- They built a custom stock-research agent because they did not trust ChatGPT for investing when its information lagged behind the market. The tool sits on paid live-data feeds, including Polygon for stocks and options, and can run code to handle screening tasks such as spotting inside-day setups. They also released it with a command-line interface and a free web version, while asking users to check the results and flag mistakes.
- How it went
- It gave them newer research and let them ask market-data questions that a standard chatbot could not answer well. It was still limited, with the author saying it mostly worked for U.S. names; one user pointed out a bad CHF/USD conversion and poor non-U.S. support, and the open-source label was disputed because the backend was not fully open.
- Tips
- If you copy this setup, expect to pay for live data access. Have the agent pause before any action that sends, submits, or changes anything, and review its outputs carefully before trusting them.
Try it yourself with Claude Code
You are my stock-research assistant. For each ticker in [tickers], pull the latest available price data, recent company news, and recent filings from [data sources], then write an updated report in [output folder]. Compare each report with the previous version and call out what changed, the most important numbers, and any risks or open questions. First, test this on [ticker 1] and [ticker 2], then stop and show me the draft reports for approval before you refresh anything else or save over existing files. If any step would send, submit, delete, or overwrite data, ask me first.
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.
Related agent uses
Build a dashboard from a semantic data layer
Use a coding agent with Graphene to define metrics, query your warehouse, and publish a report with charts and screenshots.
Use a research agent for systematic literature reviews
A Reddit thread points to tools that can help with systematic reviews, paper notes, and keeping a project organized over time.
Research a topic and verify every citation
A multi-agent setup gathers sources, drafts a brief, and checks each citation before delivering the final write-up.
Build a reusable stock-analysis workspace
A custom agent kept one investment project alive, saving models and notes so later sessions could pick up where the last run stopped.
Use a browser agent to find and save papers
A commenter uses a custom browser agent for a narrow set of research and data-entry jobs, including finding scientific papers and downloading bank CSVs.
Use an agent to trace papers from a seed set
An agent can start with a few papers, follow citations both ways, and keep narrowing a review until the results stop changing.