待翻譯：Show HN: An AI agent that trades inside limits you set, starting on paper

AI 服務暫時不可用，以下為來源正文，待恢復後補全翻譯。

QuantSignals FST — Supervised AI Trading from Research to Execution FST connects proprietary QS Research signals to account-aware decisions, hard risk limits, order preview and approval, monitored execution, and a reviewable session record. Key Takeaways QuantSignals FST is a supervised AI trading agent, not a promise of hands-off profits. The operating loop is screen, signal, trade, monitor, park, and learn. Every order is constrained by the selected account, certified broker capabilities, risk policy, preview, approval settings, and reconciliation. New users can evaluate behavior in QS paper trading before connecting an eligible live account. FST runs across the QS web app, desktop and CLI surfaces, iOS, and Android. What QuantSignals FST Does QuantSignals FST is a supervised AI trading agent that carries a market thesis from QS Research into a bounded brokerage workflow. It screens opportunities, reads the supporting and conflicting evidence, checks portfolio fit, calculates a risk budget, previews an order, follows the account's approval policy, monitors the resulting position, and records what happened. FST does not remove market risk and does not guarantee a profitable outcome. Its purpose is to make the decision and execution process more consistent, observable, and accountable while keeping hard controls outside the language model. The product is one connected system across QS Research, FST, and QS Brokerage. Research supplies time-stamped evidence; FST turns that evidence into an explicit thesis and policy decision; Brokerage owns account capabilities, risk enforcement, execution, reconciliation, and the authoritative order record. How It Works Choose a destination — Select QS paper or an account whose current capabilities have been verified for the requested operation. Start a supervised session — FST screens the configured universe, reads QS signal evidence, and narrates the market and thesis state. Apply policy and risk — The agent checks portfolio fit, calculates bounded exposure, and rejects plans that violate account or risk rules. Preview, approve, and reconcile — Orders are previewed, follow the configured approval policy, route to the exact account, and reconcile against the broker response. Monitor and park — FST tracks price, thesis, stop, target, and market state until the position is exited, reduced, added to, or held under policy. Frequently Asked Questions What is QuantSignals FST? QuantSignals FST is a supervised AI trading agent that connects QS Research signals to account-aware, risk-governed brokerage execution and continuous position monitoring. Is FST fully autonomous? FST can run an automated session inside preconfigured boundaries, but it remains supervised: account eligibility, capability checks, hard risk limits, approval policy, reconciliation, stop controls, and safe defaults constrain what it can do. Does FST guarantee profits? No. Trading involves risk, paper results do not represent future live results, and no AI trading agent can guarantee a profitable outcome. Can I test FST without a live brokerage account? Yes. QS paper trading is the default evaluation path so users can observe the complete session behavior before considering an eligible live account. Where does FST run? FST is available through the QS web app, desktop and command-line surfaces, iOS, and Android. Install FST 2.0 (desktop CLI & mobile app) · See QuantSignals V6