Supported Data for AI Strategy Prompts on Everstrike

When writing a strategy prompt, knowing what data the AI can actually access is essential for building reliable strategies. Reference this page to understand exactly which market data, account context, and technical indicators are available during each evaluation, and which data types to avoid referencing in your prompts. Data availability directly limits what the AI can reason about.

​

Available Market Data

The AI has access to the following real-time market data during every evaluation:

• Real-time price data, including OHLCV (open, high, low, close, volume)
• L1 order book data showing the best bid and ask
• Order flow data, including recent buys and sells
• Temporal data, including current time, date, and weekday
• Statistical data (see Statistical Data below)
• Technical indicators (see Technical Indicators below)
• Extra data sources, opt-in per strategy: macro and stock indices, commodities, the crypto market, tokenized stocks and ETFs, Everstrike derivatives, Bitcoin network, DeFi liquidity and sentiment (see Extra Data Sources below)

​

Unavailable Data

​

Account Context

The AI receives the following account context automatically with each evaluation: Default context (always included):

• Open positions
• Open orders
• Past trades
• Account equity
• Available cash

Optional context (can be enabled per strategy):

• Liquidation prices
• Stop losses and take profits
• Active strategies
• Previously submitted strategies

All account context is anonymized before the AI receives it. Your account identity is never exposed to the model.

​

Technical Indicators

The AI supports the following technical indicators. These use the same inputs as TradingView and the Everstrike UI.

• Trend: MACD, ADX, Ichimoku Cloud, Keltner Channels, KST, TRIX
• Momentum: RSI, CCI, Force Index, Rate of Change, Stochastic Oscillator, Stochastic RSI, William’s R
• Volume: ADL, MFI, OBV, VWAP
• Volatility: Bollinger Bands, ATR
• Price channels: PSAR
• Moving averages: SMA, EMA, and WMA with periods of 9, 20, and 50

By default, the AI receives the 5 most recent values for each indicator. Enabling historical indicator insights expands this to up to 100 values per indicator, but significantly increases credit consumption and is disabled by default.

​

Statistical Data

The following statistical values are available for each evaluated market:

• Open, high, low, and close
• Open interest, daily volume, and daily change
• Spread, funding rate, index price, and mark price

​

Extra Data Sources

Besides market data and indicators, strategies can read 297 extra data sources:

• Macro: rates and yields, inflation (CPI, core PCE and inflation expectations), growth (GDPNow), credit spreads, the dollar and FX, Fed liquidity, jobs, the S&P 500 and Nasdaq Composite, and days until the next CPI, FOMC and payrolls dates
• Commodities: oil and gas prices, and tokenized gold, silver and uranium
• Crypto market: total market cap, dominance and major coin prices
• Tokenized stocks and ETFs: on-chain tokens tracking US shares and ETFs, trading around the clock
• Everstrike derivatives: funding, basis, open interest, depth, volatility and put/call ratios on Everstrike’s own markets
• Bitcoin network: hashrate, fees, mempool and miner economics
• DeFi, stablecoins and liquidity: TVL, DEX volume, stablecoin supply and pegs
• Sentiment: the Crypto Fear & Greed Index
• Data quality: the health of the data feed itself

To add an extra to a strategy:

• Open the strategy’s Advanced settings → Input data, then search by name or key (searching “vix” or “dollar” works), or browse the Metrics sections (Macro, Commodities, Crypto market, Tokenized stocks and ETFs, Everstrike derivatives, Bitcoin network, Liquidity, Sentiment, Data quality). Your strategy keeps all its default inputs.
• Or write it into your prompt (for example “only buy when the VIX is below 20”). The prompt check flags it, and Add puts it in the input data for you.
• When the AI assistant writes or fills in a strategy that relies on an extra, it adds it to the input data itself.

In the prompt, refer to an extra by name, by key, or both, for example “VIX (macro.vix)”.

​

Units and update frequency

• Latest value only. Each live evaluation receives the current value of an extra, with no history: the lookback and historical insights settings don’t apply to extras. Rules that need a trend, such as “the VIX has risen for three days”, can’t be evaluated. Where a ready-made change figure exists, use it instead (for example liquidity.defi_tvl_change_prev_7d_pct, liquidity.dex_volume_change_7d_pct, or a _change_24h_pct key).
• Backtests use daily history. A Fast Estimate replays an extra from its daily history, so a backtest sees one value per day, while the live strategy sees the latest value at each evaluation.
• Keys ending in _pct are percentages (4.78 means 4.78%), and keys ending in _usd are US dollars.
• Update pace varies. Crypto, tokenized, derivatives and Bitcoin-network figures update every few minutes or live. Many macro series update daily, weekly or monthly (jobless claims weekly; CPI, PCE, payrolls, job openings and unemployment monthly), and the S&P 500 and Nasdaq values are daily closes. Use the slow ones as background filters, not fast triggers.
• Keys ending in _bps are basis points (25 means 0.25%). Flags such as macro.curve_inverted_10y_2y are 1 (yes) or 0 (no). days_to_* keys count whole days, with 0 meaning today (US Eastern).
• Tokenized stocks and ETFs are priced per token and follow the ETF or share, not the index level: tokenized.sp500_price_usd tracks an S&P 500 ETF, not the index’s points (use macro.sp500_index for those). Thinly traded tokens can lag their underlying; check their 24h volume.
• Everstrike funding is capped at ±0.3% per hour on perpetuals and ±10% on options. Option volatility is the underlying’s 14-day annualised historical volatility, one value for every strike. Order volume is the size of orders placed, in the market’s own units, not dollars.
• macro.nonfarm_payrolls is the total number of jobs, not the monthly change.
• network.btc.hashrate is in TH/s, Bitcoin fee keys are in sat/vB, and network.btc.transaction_fees and network.btc.miner_revenue_usd are 24-hour totals.
• macro.usd_broad_index is the Federal Reserve’s broad dollar index (around 120), not DXY.
• sentiment.fear_greed runs from 0 to 100, with 50 as neutral.

​

All extra data sources

​

Options and Position Data

​

Options Data

When evaluating options markets, the AI can access:

• Underlying price
• Historical volatility
• Black-Scholes price

​

Position Data

For any open positions on an evaluated market, the AI can access:

• Market
• Size
• Profit
• Entry price

​

Data Limits

Keep the following constraints in mind when writing prompts:

• The AI uses UTC as its default timezone for all temporal data.
• Order flow includes the 50 most recent trades per market. The AI cannot access trade history beyond this window.
• Extra data sources are only sent to strategies that include them, carry their latest value only (no lookback), and update at their source’s own pace (see Units and update frequency).