obai-strategy

Designs and backtests quantitative equity trading strategies via the OBaI backtest MCP server.

11|6|Updated Mar 18, 2026
One-click install
npx skills add https://github.com/sixteen-dev/obai --skill obai-strategy-sixteen-dev
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: obai-strategy
Source: https://github.com/sixteen-dev/obai/tree/main/skills/obai-strategy
Command: npx skills add https://github.com/sixteen-dev/obai --skill obai-strategy-sixteen-dev

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? It turns trading ideas into executable strategy JSON and validates them with real backtest evidence instead of untested prose, so strategy decisions rest on measured Sharpe, drawdown, and trade statistics. ## Core Features & Use Cases - Strategy Design and Backtesting: Converts user-specified entry and exit rules into valid strategy JSON and runs single-range backtests with backtest_run_strategy_tool. - Iterative Optimization: Runs a build-test-analyze-adjust loop with parameter sensitivity comparisons and train/out-of-sample validation. - Walk-Forward Robustness: Uses backtest_walk_forward_tool for multi-period overfitting checks on strategies with four or more years of data. - Use Case: Ask to backtest a 5-minute RSI mean-reversion on NVDA's opening hour; the skill builds the JSON, executes the backtest, and returns a verdict with full metrics and the final executable strategy. ## Quick Start Backtest a long-only RSI mean-reversion strategy on AAPL and MSFT over the last five years and give me the verdict with the final strategy JSON.

Frequently Asked Questions about obai-strategy

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I backtest a trading strategy with entry and exit rules?▼

State your entry and exit conditions in plain language along with the ticker symbols. The skill converts them into strategy JSON and runs `backtest_run_strategy_tool`, returning metrics like Sharpe, max drawdown, win rate, and profit factor.

How do I check if my trading strategy is overfit?▼

Use walk-forward validation via `backtest_walk_forward_tool`, which tests one fixed strategy across expanding train/test windows. A consistency score below 60 percent or degradation above 0.5 indicates the strategy does not hold out-of-sample.

What indicators and operators does the backtest engine support?▼

The engine supports trend, volatility, volume, statistical, and candlestick indicators such as RSI, SMA, ATR, DONCHIAN, VWAP, and CDL patterns, with operators like greater_than, less_than, crosses_above, and crosses_below. Run `backtest_get_supported_indicators_tool` for the full list.

Does the backtest engine support intraday strategies?▼

Yes, it supports 5min, 15min, and 1hour timeframes alongside daily bars. Intraday data is limited to 2 years for 5min and 15min bars and 5 years for 1hour bars, and portfolio allocation mode only works on daily timeframes.

Why did my backtest return zero trades?▼

Zero trades usually means a predicate never fired, an indicator never primed, or a data gap exists. Check `signal_diagnostics` and warnings before loosening thresholds, since a threshold misclassified as a crossover is a common cause.

What are the limitations of the strategy backtest engine?▼

The engine cannot represent cross-sectional ranking, custom rebalance engines, portfolio-level circuit breakers, or earnings blackout logic. When ideal logic is unsupported, the skill builds the closest valid approximation and discloses what it captures and misses.