The capstone, in brief

About this track: Clawdemy uses the open-source TradingAgents framework (by TauricResearch; Yijia Xiao, Edward Sun, Di Luo, and Wei Wang; arXiv:2412.20138; Apache-2.0) as a real-world example of how an agentic AI system is structured. Our goal is to teach the architecture: how specialist agents, a research debate, a trader role, and a risk layer are coordinated by an orchestrator. We are not teaching you how to invest, trade, or pick stocks, and we make no claim that this or any AI system is profitable or predicts markets. This content is for education only and is not investment, financial, or trading advice. Rules vary by country, and this is not advice anywhere. The Trading Agents Lab capstone is a simulation you explore yourself with your own AI provider key; it does not place real trades. The framework’s own authors state it plainly: “It is not intended as financial, investment, or trading advice.”

What you’ll learn

This is the capstone, where everything you read across the track runs at once. You will watch the full pipeline execute end to end: the four analysts gather, the bull and the bear debate, the research manager rules, the trader plans, the three risk seats stress-test, and the portfolio manager makes the final call. You will see the system write seven plain reports, one per stage, and learn to map each report back to the agent and the lesson that produced it. The two judgment points stand out exactly as the track described them: the research manager turns the debate into a ruling, and the portfolio manager (the deep-model gate) reconciles the risk arguments into a final written recommendation on the system’s five-point scale. Then, if you want, you run it yourself: it is the same open-source code you studied, pinned to one snapshot, in a safe simulation that writes a research opinion and places no trades. The lasting takeaway is the architecture, not the domain: a pattern you can carry into your own AI projects.

Where this fits

This is the eighth and final lesson of the track. Lessons 1 through 7 built and explained the team part by part; this lesson runs the whole thing and ties the architecture together. There is no next lesson; this is where the track lands.

Before you start

Prerequisites: the rest of the track (lessons 1 through 7), since the capstone assumes you recognize each agent when you see it run. To run the system yourself (optional) you need a computer with Python and your own AI provider key; to simply read and watch, you need nothing.

By the end, you’ll be able to

Trace how control flows through the full pipeline from analysts to the final gate
Identify each of the seven reports with the agent and lesson that produced it
Explain why the two managers run on the deep model and the rest on the faster one
Run the open-source system yourself in a safe simulation (optional)
Apply the multi-agent design pattern to your own AI projects

Time and difficulty

Read time: about 9 minutes
Practice time: about 20 minutes to read and reflect; longer if you choose to run the system yourself
Difficulty: deep (this track is the advanced end of the catalog; the foundational agent tracks are the on-ramp)