TRADING DESK

AUTONOMOUS BOT // OWNER CONSOLE
ACCESS TOKEN

TRADING DESK

AUTONOMOUS // LEARNS FROM EVERY TRADE
Equity
Day P&L
Cash
Market
Portfolio equity
My portfolio — what I own now
SymbolSharesAvg costPriceMarket valueUnreal. P&L
Trade ledger
SymbolStatusEntryExit / MarkSharesP&L $P&L %Opened
Meta controls — set these two, the knobs follow
RISK TOLERANCE50
AGGRESSIVENESS50

Hard knobs — enforced in code, per run
Investing strategies — educational overview & selection. The bot trades whatever mix you set; every trade still passes the risk engine.
Regime fit reflects how each approach has historically behaved in markets like the current one. It is generic market-state information, not personalized investment advice, and past behavior does not guarantee future results.
Staged trades — waiting on their trigger; every armed plan still passes the risk engine
Watchlist & research history — every stock the bot has looked at
Discuss & evolve the bot's strategy
BOT
This is the strategy channel. Tell me how you want to invest — stay on news/sentiment, shift toward swing trading, screen for impact criteria, hold longer. When we land on a change, I'll propose a DIRECTIVE and you commit it to my memory. I read that memory before every trading run.
Closed
Hit rate
Realized P&L
Cumulative realized P&L by trade
Today's research brief
Prediction calibration — measured hit-rates of the bot's own calls
Playbook — what the bot has learned
Owner directives — what you've taught it
Edge verification — pre-registered hypotheses (docs/experiments/), scored by arithmetic against real prices. Verdicts are code-assigned; kills and scale-ups still need your sign-off.
IDHypothesisnKey statVerdict
Forecasts scored
Hit rate (all)
Daily IC
Daily NAV — bot vs its no-LLM twin vs SPY vs the forecast shadow (all start at 1.00)
Paired daily excess — differencing removes the shared market move
PairDaysMean / day95% CI
Forecast accuracy by driver & liquidity tier — no rate is quoted below n=30
Recent forecasts
DateSymHp↑DriverSourceResult