voyager67/marketpilot-ai
MarketPilot AI combines 17 cross-model agent skills with a Python research engine for market scans, news and filings, entry/exit plans, position sizing, risk checks, scenario replay, walk-forward testing, alerts, journaling and portfolio review. Built for ChatGPT, Claude, Grok, Gemini and Codex; transparent, testable and paper-only; traders decide.
Create and manually check one-shot local price alerts with explicit symbol, threshold and direction.
Summarize reported SEC XBRL metrics and recent filing metadata while preserving periods, units, and missing-data caveats.
Prepare and record local paper-only tickets after a passing risk check and explicit paper confirmation; this skill has no live-order capability.
Calculate whole-share position size from account equity, per-trade risk, stop distance, slippage, fees and a notional cap.
Audit OHLCV data integrity, timestamps, missing volume and source completeness before any signal or simulation.
Index public U.S. House Periodic Transaction Reports and explain disclosure lag, transaction ranges and source limitations.
Check whether a specified long setup threshold has triggered on a completed, sufficiently fresh market bar.
Build a transparent long-only ATR stop and reward/risk target plan from user-specified entry and historical bars.
Summarize user-supplied recorded trades and identify process patterns without presenting unaudited journal statistics as verified performance.
Collect recent public headlines and SEC filing metadata for a user-supplied watchlist; classify catalysts without turning headlines into predictions.
Compute supplied cash and position weights to surface concentration; no broker account is connected.
Describe broad trend and volatility state from historical bars without forecasting direction.
Validate the exact proposed long-only order against user-defined loss, exposure, trade-count, reward/risk and data-freshness constraints.
Scan a supplied symbol list or public Nasdaq Trader universe for predefined chart setups. Use after the user asks for a scan, watchlist, breakout candidates, or oversold names.
Run a chronological training/holdout breakout study and audit for look-ahead, sample size and parameter overfitting.
Run reproducible historical and bootstrap scenarios for a user-specified entry, stop, and target.
Review an already completed trade against its original plan, realized fill, slippage and process adherence.