Run any Skill in Manus with one click

general-time-series-forecasting

Stars1

Forks0

UpdatedJuly 1, 2026 at 23:33

Plan and review general time-series forecasting with classical models, ML, foundation models, backtesting, leakage checks, and uncertainty. Use for non-omics forecasting, sensor streams, operations, finance, clinical vitals, demand, or irregular temporal data.

Installation

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

Run Skill in Manus

Source

BaranziniLab

BaranziniLab/biorouter-skills

View GitHub Repository View Creator Repositories

Download

Run Skill in Manus

Related occupationsSOC

Based on SOC occupation classification

Data ScientistsComputer and Mathematical Occupations·SOC 15-2051

SKILL.md

readonly

name	general-time-series-forecasting
description	Plan and review general time-series forecasting with classical models, ML, foundation models, backtesting, leakage checks, and uncertainty. Use for non-omics forecasting, sensor streams, operations, finance, clinical vitals, demand, or irregular temporal data.
user-invocable	true
license	Apache-2.0

General Time-Series Forecasting

Source adaptation: BioRouter-original synthesis inspired by Scientific Agent Skills time-series and forecasting patterns, rewritten under Apache-2.0.

When to Use

Use this skill for general forecasting or temporal prediction outside BioRouter's omics-specific temporal-genomics BioSkills. Use it for sensors, operations, finance, clinical vitals, weather, demand, events, or multivariate temporal panels.

Intake

Collect:

Forecast target, horizon, cadence, and decision use.
Entity structure: single series, panel, hierarchy, spatial-temporal, or event stream.
Exogenous covariates and when they are known.
Missingness, irregular sampling, seasonality, interventions, and regime changes.
Train/validation/test dates and backtesting expectations.
Required uncertainty: intervals, quantiles, scenarios, or calibrated probabilities.

Model Routing

Baseline first: seasonal naive, moving average, ETS, ARIMA/SARIMA, or Prophet-style decompositions.
Use ML when feature history and panel size justify it: gradient boosting, random forests, neural nets.
Use deep/foundation models only after checking horizon, frequency, scaling, covariate availability, and license.
Use causal methods when the task asks about interventions, not just forecasts.
Use financial-accounting-econometrics for market/event-study tasks and clinical-decision-support-review for patient-impacting forecasts.

Backtesting Gates

Use time-ordered splits only.
Prevent leakage from future covariates, normalization, target encoding, or imputation.
Compare against simple baselines.
Report horizon-specific metrics.
Check calibration if intervals or probabilities are reported.
Preserve timestamp timezone, calendar, and missingness decisions.

Output Shape

Return:

Forecast framing.
Baseline and candidate models.
Backtest plan.
Leakage checklist.
Metrics and uncertainty plan.
Minimal reproducible code or commands if requested.

More from this repository

same repository

computational-fluid-dynamics

BaranziniLab/biorouter-skills

Plan and review computational fluid dynamics workflows with meshes, boundary conditions, turbulence models, OpenFOAM/FluidSim-style solvers, validation cases, convergence checks, and reproducible simulation artifacts. Use when users mention CFD, fluid dynamics, Navier-Stokes, meshing, turbulence, OpenFOAM, or flow simulation.

2026-07-011

arbor-optimization

BaranziniLab/biorouter-skills

Run evaluator-driven hypothesis trees for improving research artifacts, prompts, analyses, and designs. Use when a task needs branch-and-prune exploration, rubric scoring, iterative refinement, or explicit comparison of competing scientific/workflow alternatives.

2026-07-011

autoskill-workflow-discovery

BaranziniLab/biorouter-skills

Detect repeated user workflows and draft native BioRouter skills, hooks, or checklist updates from observed work patterns. Use when users ask to turn repeated work, screen-history patterns, project routines, or recurring guardrails into reusable BioRouter skills.

2026-07-011

bids-neuroimaging

BaranziniLab/biorouter-skills

Plan and audit BIDS neuroimaging datasets, derivatives, validation, metadata, and reproducible preprocessing. Use when users mention BIDS, fMRI, MRI, EEG/MEG/iEEG, DICOM conversion, dataset_description.json, participants.tsv, or neuroimaging derivatives.

2026-07-011

clinical-ai-modeling

BaranziniLab/biorouter-skills

Plan and review clinical AI modeling workflows with EHR cohorts, PyHealth-style tasks, validation, bias checks, calibration, safety boundaries, and no-PHI handling. Use when users ask about clinical prediction models, risk scores, treatment-response ML, medical reports, or clinical model evaluation.

2026-07-011

hugging-science-resource-discovery

BaranziniLab/biorouter-skills

Find and evaluate Hugging Face scientific models, datasets, Spaces, and papers with license, provenance, benchmark, and fit checks. Use when users ask for scientific ML resources, model cards, datasets, embeddings, checkpoints, or reproducible AI assets from Hugging Face.

2026-07-011

name	general-time-series-forecasting
description	Plan and review general time-series forecasting with classical models, ML, foundation models, backtesting, leakage checks, and uncertainty. Use for non-omics forecasting, sensor streams, operations, finance, clinical vitals, demand, or irregular temporal data.
user-invocable	true
license	Apache-2.0

General Time-Series Forecasting

Source adaptation: BioRouter-original synthesis inspired by Scientific Agent Skills time-series and forecasting patterns, rewritten under Apache-2.0.

When to Use

Intake

Collect:

Forecast target, horizon, cadence, and decision use.
Entity structure: single series, panel, hierarchy, spatial-temporal, or event stream.
Exogenous covariates and when they are known.
Missingness, irregular sampling, seasonality, interventions, and regime changes.
Train/validation/test dates and backtesting expectations.
Required uncertainty: intervals, quantiles, scenarios, or calibrated probabilities.

Model Routing

Baseline first: seasonal naive, moving average, ETS, ARIMA/SARIMA, or Prophet-style decompositions.
Use ML when feature history and panel size justify it: gradient boosting, random forests, neural nets.
Use deep/foundation models only after checking horizon, frequency, scaling, covariate availability, and license.
Use causal methods when the task asks about interventions, not just forecasts.
Use financial-accounting-econometrics for market/event-study tasks and clinical-decision-support-review for patient-impacting forecasts.

Backtesting Gates

Use time-ordered splits only.
Prevent leakage from future covariates, normalization, target encoding, or imputation.
Compare against simple baselines.
Report horizon-specific metrics.
Check calibration if intervals or probabilities are reported.
Preserve timestamp timezone, calendar, and missingness decisions.

Output Shape

Return:

Forecast framing.
Baseline and candidate models.
Backtest plan.
Leakage checklist.
Metrics and uncertainty plan.
Minimal reproducible code or commands if requested.