quantlogix-signals

QuantLogix Signals — interpretation & tool routing

QuantLogix is an AI-powered quantitative signal platform. This plugin connects Claude to the live QuantLogix engine via the quantlogix MCP server (you sign in with your QuantLogix account on first use; the tools you can call are gated by your subscription tier). Use these tools to answer markets questions with live data + a verified track record, never from memory.

Pick the right tool

The user wants…	Call
The long-term read on a stock/ETF ("is NVDA a buy?", "what's the signal on AAPL?")	get_signal (alias get_long_term_signal)
Today's intraday read (Pro+)	get_day_signal
A 2-day–4-week swing read with A/B/C/D grade (Pro+)	get_swing_signal
Order-flow / smart-money read (Pro+)	get_microstructure
"Does QuantLogix actually work?" / proof / verified win-rate	get_track_record
Find candidates by criteria ("Strong Buy tech under $50B, score ≥ 70")	screener (Pro+) — full multi-criteria; or get_signal_rankings (free) for a quick top-N
News + sentiment on a ticker	get_news / get_ticker_sentiment
Market overview (S&P/Nasdaq/Dow/Russell/VIX)	get_market_overview
Macro environment (FRED: rates, inflation, jobs, the macro score)	get_macro
Curated strategy baskets + their live track records	list_baskets / get_basket
Turn an investment THEME into a basket ("Quantum Computing")	list_basket_themes → resolve_theme_basket
Earnings setup / calendar / corporate events (Pro+)	get_earnings_setup / get_earnings_calendar / get_corporate_events
A fast price quote, crypto, FX, IPO calendar	get_quote / get_crypto / get_fx_signal (Pro+) / get_ipo_calendar

When the user names a ticker without specifying a horizon, default to get_signal (long-term) and offer the day/swing reads if they trade shorter horizons.

Read the 5-factor composite

get_signal returns a composite_score (0–100), a label, and a per-factor breakdown. Bands (gauge-aligned — the label and the score never disagree):

• ≥ 70 → Strong Buy (gauge "Strong")
• 50–69 → Buy (gauge "Bullish")
• 30–49 → Neutral (gauge "Mild")
• 16–29 → Sell (gauge "Bearish")
• ≤ 15 → Strong Sell

The five factors and their default weights: Technical 30% (RSI, SMAs, MACD, EMA), Momentum 25% (intraday change, VWAP, volume, ext-hours), Fundamental 15% (margins, ROE, growth, cash-flow quality), Options 10% (put/call, unusual flow, IV), Microstructure 20% (OFI, VPIN, smart-money, institutional flow). The composite also carries a bounded ±5 macro adjustment and a ±4 regime (HMM) adjustment, plus a pattern-prediction block (historical-analog win rates). When explaining a score, name the factors that drove it rather than just the number.

Always lead with the verified record

QuantLogix's differentiator is that every signal is logged immutably at issuance and resolved against live price — so the track record is audited, not self-reported. When a user is skeptical ("does this work?"), call get_track_record and quote the measured win rate, average return, and Sharpe by timeframe/sector. This is the single most credible thing you can show.

Theme baskets are tier-metered

resolve_theme_basket turns a theme into an AI-selected, equal-weight basket, but each plan may open a capped number of distinct themes per month (Free 5 / Starter 25 / Pro 100 / Institutional unlimited). Re-opening a theme you've already opened this month is free. If a resolve is refused for the cap, tell the user their plan's limit and that upgrading raises it — don't retry.

Honesty rules (non-negotiable)

• Never invent figures. Use the numbers the tools return. If a tool has no data for a ticker, say so.
• Signals and basket performance are quantitative model outputs / backtests, not guarantees and not personalized investment advice. State this when giving a stance.
• Basket constituents are AI-selected; any trailing performance shown is a transparent backtest, not a live money-managed track record.
• Tier-gated tools will return an upgrade message for free/under-tier users — relay it plainly; don't pretend the data is unavailable.