predict_forecast
ARIMA + Holt-Winters time series forecasting
How to use it
predict_forecast is exposed by the Oraclaw MCP server. Add the server to your MCP client (Claude Desktop, Cursor, Windsurf and others), and the predict_forecast tool becomes available to the model automatically. See the full listing for setup details and every tool this server provides.
Install Oraclaw
npm install @oraclaw/banditOther tools in Oraclaw (16)
PageRank, Louvain communities, bottleneck detection
VaR and CVaR (Expected Shortfall)
Z-Score + IQR anomaly detection
UCB1 / Thompson / Epsilon-Greedy arm selection
CMA-ES continuous black-box optimization
Context-aware LinUCB bandit
Genetic algorithm for discrete + multi-objective problems
A + Yen's k-shortest paths
Beta posterior update from weighted evidence
Multi-model consensus + uncertainty decomposition
Brier + log score for forecaster accuracy
Multi-source probability consensus (Hellinger)
Single-factor Monte Carlo (6 distributions)
What-if comparison + sensitivity ranking
LP / MIP / QP solver via HiGHS (provably optimal)
Energy-matched task scheduling