Workflow: Run Deep Research
Use Deep Research when a ticker needs coordinated specialist analysis rather than a short summary.
1. Enter the ticker
Section titled “1. Enter the ticker”Enter the ticker you want to investigate in the Deep Research workspace. Make sure it identifies the intended listed instrument before starting the run.
2. Choose Depth
Section titled “2. Choose Depth”- Quick: a rapid structured investigation.
- Standard: balanced multi-agent coverage.
- Deep: a more extensive investigation. For an eligible U.S. equity, it can request and use the 51-section U.S. Equity Master when applicable.
Depth is requested scope, not a guarantee of the execution profile. The actual execution and output profile is capability- and policy-dependent, and the run view is authoritative. Do not assume every Deep run uses Master.
3. Choose Horizon
Section titled “3. Choose Horizon”Select intraday, days, weeks, or months. The horizon changes which
evidence, catalysts, and risks are most relevant; it does not predict an
outcome.
4. Follow the profile in use
Section titled “4. Follow the profile in use”The standard team profile uses plan, analysis, debate, strategy, risk, and synthesis. Review the run view for the profile and work that actually ran.
An eligible Deep Master run uses a separate 51-section section-writer swarm. It is not the six-phase standard workflow. After the sections are drafted, the run can remain active for final synthesis, citation verification, quality review, and report rendering.
5. Read the completed result
Section titled “5. Read the completed result”When available, start with the completed summary card: its main finding, confidence where shown, risks, coverage, and source preview. Then open the canonical HTML report for the complete output. PDF and DOCX downloads are available only when their artifacts are ready.
An artifact marked as preparing is still being assembled. An unavailable artifact is not a substitute report; read the available summary and run status. The full U.S. Equity Master is normally read in its HTML report, not inline in the workspace.
6. Interpret it as research
Section titled “6. Interpret it as research”Read sources and evidence quality before relying on the final view. Compare the positive and skeptical cases for a standard run, review limitations and invalidation conditions, and keep the selected horizon in mind. The result is a probabilistic research input, never an instruction to buy, sell, allocate, or execute a trade.