aws-forecast
Configures AWS Forecast workflows. Use this for dataset management and model training in the cloud.
Install
mkdir -p .claude/skills/aws-forecast && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/18830" && unzip -o skill.zip -d .claude/skills/aws-forecast && rm skill.zipInstalls to .claude/skills/aws-forecast
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Forecast via AWS CLI v2 (`aws forecast`). Time-series forecasting, datasets, AutoPredictors, forecasts, what-if analysis, explainability, monitors, queries.Key capabilities
- →Create time-series datasets
- →Import data into datasets from S3
- →Train AutoPredictors for automated model selection
- →Generate forecasts with specified forecast types
- →Query forecasts for specific items
- →Perform what-if analysis with transformations
How it works
The skill executes AWS CLI v2 commands to manage Amazon Forecast resources. It handles dataset creation, data import, predictor training, forecast generation, and what-if analysis.
Inputs & outputs
When to use aws-forecast
- →Create time-series dataset
- →Train AutoPredictor
- →Generate forecasting model
- →Set up what-if analysis
About this skill
Amazon Forecast CLI Reference
Amazon Forecast is a time-series forecasting service that uses machine learning to generate accurate forecasts from historical time-series data. It supports multiple domains (retail, inventory, workforce, web traffic, EC2 capacity, metrics) and provides AutoPredictor for automated model selection.
Note: Amazon Forecast is no longer available to new customers as of June 2024. Existing customers can continue to use the service.
Common Workflows
Create a Dataset and Import Data
# Create a dataset
aws forecast create-dataset \
--dataset-name "retail_demand" \
--domain RETAIL \
--dataset-type TARGET_TIME_SERIES \
--data-frequency "1D" \
--schema 'Attributes=[{AttributeName=item_id,AttributeType=string},{AttributeName=timestamp,AttributeType=timestamp},{AttributeName=demand,AttributeType=float}]'
# Create a dataset group
aws forecast create-dataset-group \
--dataset-group-name "retail_group" \
--domain RETAIL \
--dataset-arns "arn:aws:forecast:us-east-1:123456789012:dataset/retail_demand"
# Import data from S3
aws forecast create-dataset-import-job \
--dataset-import-job-name "import_20240101" \
--dataset-arn "arn:aws:forecast:us-east-1:123456789012:dataset/retail_demand" \
--data-source S3Config={Path=s3://my-bucket/data/,RoleArn=arn:aws:iam::123456789012:role/ForecastRole} \
--timestamp-format "yyyy-MM-dd"
Train a Predictor and Generate Forecast
# Create an AutoPredictor
aws forecast create-auto-predictor \
--predictor-name "retail_predictor" \
--forecast-horizon 30 \
--forecast-frequency "1D" \
--data-config '{"DatasetGroupArn":"arn:aws:forecast:us-east-1:123456789012:dataset-group/retail_group"}' \
--optimization-metric WAPE
# Check accuracy metrics
aws forecast get-accuracy-metrics \
--predictor-arn "arn:aws:forecast:us-east-1:123456789012:predictor/retail_predictor"
# Create a forecast
aws forecast create-forecast \
--forecast-name "retail_forecast_q1" \
--predictor-arn "arn:aws:forecast:us-east-1:123456789012:predictor/retail_predictor" \
--forecast-types "0.10" "0.50" "0.90"
# Query the forecast
aws forecastquery query-forecast \
--forecast-arn "arn:aws:forecast:us-east-1:123456789012:forecast/retail_forecast_q1" \
--filters '{"item_id":"product_123"}'
# Export forecast to S3
aws forecast create-forecast-export-job \
--forecast-export-job-name "export_q1" \
--forecast-arn "arn:aws:forecast:us-east-1:123456789012:forecast/retail_forecast_q1" \
--destination S3Config={Path=s3://my-bucket/output/,RoleArn=arn:aws:iam::123456789012:role/ForecastRole}
What-If Analysis
# Create a what-if analysis
aws forecast create-what-if-analysis \
--what-if-analysis-name "price_increase" \
--forecast-arn "arn:aws:forecast:us-east-1:123456789012:forecast/retail_forecast_q1"
# Create a what-if forecast with transformations
aws forecast create-what-if-forecast \
--what-if-forecast-name "price_up_10pct" \
--what-if-analysis-arn "arn:aws:forecast:us-east-1:123456789012:what-if-analysis/price_increase" \
--time-series-transformations '[{"Action":{"AttributeName":"price","Operation":"MULTIPLY","Value":1.10}}]'
# Query the what-if forecast
aws forecastquery query-what-if-forecast \
--what-if-forecast-arn "arn:aws:forecast:us-east-1:123456789012:what-if-forecast/price_up_10pct" \
--filters '{"item_id":"product_123"}'
Monitoring and Explainability
# Create a monitor for predictor drift
aws forecast create-monitor \
--monitor-name "retail_monitor" \
--resource-arn "arn:aws:forecast:us-east-1:123456789012:predictor/retail_predictor"
# Create explainability
aws forecast create-explainability \
--explainability-name "retail_explain" \
--resource-arn "arn:aws:forecast:us-east-1:123456789012:predictor/retail_predictor" \
--explainability-config TimeSeriesGranularity=ALL,TimePointGranularity=ALL
# Export explainability results
aws forecast create-explainability-export \
--explainability-export-name "explain_export" \
--explainability-arn "arn:aws:forecast:us-east-1:123456789012:explainability/retail_explain" \
--destination S3Config={Path=s3://my-bucket/explain/,RoleArn=arn:aws:iam::123456789012:role/ForecastRole}
Command Reference
| Group | File | Commands |
|---|---|---|
| Datasets | datasets.md | create-dataset, describe-dataset, list-datasets, delete-dataset, create-dataset-group, describe-dataset-group, list-dataset-groups, update-dataset-group, delete-dataset-group, create-dataset-import-job, describe-dataset-import-job, list-dataset-import-jobs, delete-dataset-import-job |
| Predictors | predictors.md | create-auto-predictor, create-predictor, describe-auto-predictor, describe-predictor, list-predictors, delete-predictor, get-accuracy-metrics |
| Forecasts | forecasts.md | create-forecast, describe-forecast, list-forecasts, delete-forecast, create-forecast-export-job, describe-forecast-export-job, list-forecast-export-jobs, delete-forecast-export-job |
| Explainability | explainability.md | create-explainability, describe-explainability, list-explainabilities, delete-explainability, create-explainability-export, describe-explainability-export, list-explainability-exports, delete-explainability-export |
| Monitors | monitors.md | create-monitor, describe-monitor, list-monitors, list-monitor-evaluations, delete-monitor |
| What-If | what-if.md | create-what-if-analysis, describe-what-if-analysis, list-what-if-analyses, delete-what-if-analysis, create-what-if-forecast, describe-what-if-forecast, list-what-if-forecasts, delete-what-if-forecast, create-what-if-forecast-export, describe-what-if-forecast-export, list-what-if-forecast-exports, delete-what-if-forecast-export |
| Backtest Exports | backtest-exports.md | create-predictor-backtest-export-job, describe-predictor-backtest-export-job, list-predictor-backtest-export-jobs, delete-predictor-backtest-export-job |
| Operations | operations.md | stop-resource, resume-resource, delete-resource-tree |
| Queries | queries.md | query-forecast, query-what-if-forecast (via aws forecastquery) |
| Tags | tags.md | tag-resource, untag-resource, list-tags-for-resource |
When not to use it
- →For new customers, as Amazon Forecast is no longer available to them as of June 2024
- →When the user does not have an existing AWS Forecast account
Prerequisites
Limitations
- →Amazon Forecast is no longer available to new customers as of June 2024
- →Requires existing AWS Forecast access
How it compares
This skill automates the interaction with Amazon Forecast via CLI commands, providing a programmatic way to manage forecasting workflows, unlike manual console operations.
Compared to similar skills
aws-forecast side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| aws-forecast (this skill) | 0 | 17d | Review | Intermediate |
| modal | 5 | 7mo | Review | Intermediate |
| mlops-engineer | 3 | 3mo | No flags | Advanced |
| bedrock | 1 | 6mo | Review | Intermediate |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by chrishuffman5
View all by chrishuffman5 →You might also like
modal
davila7
Run Python code in the cloud with serverless containers, GPUs, and autoscaling. Use when deploying ML models, running batch processing jobs, scheduling compute-intensive tasks, or serving APIs that require GPU acceleration or dynamic scaling.
mlops-engineer
sickn33
Build comprehensive ML pipelines, experiment tracking, and model registries with MLflow, Kubeflow, and modern MLOps tools. Implements automated training, deployment, and monitoring across cloud platforms. Use PROACTIVELY for ML infrastructure, experiment management, or pipeline automation.
bedrock
itsmostafa
AWS Bedrock foundation models for generative AI. Use when invoking foundation models, building AI applications, creating embeddings, configuring model access, or implementing RAG patterns.
aws-sagemaker
chrishuffman5
SageMaker via AWS CLI v2 (`aws sagemaker`). Training jobs, models, endpoints, processing, hyperparameter tuning, AutoML, pipelines, experiments, Feature Store, model registry, monitoring, notebooks, HyperPod clusters, Studio domains, MLflow, optimization, labeling, real-time/async inference.
azure-ai-ml-py
mk-knight23
|
llava
zechenzhangAGI
Large Language and Vision Assistant. Enables visual instruction tuning and image-based conversations. Combines CLIP vision encoder with Vicuna/LLaMA language models. Supports multi-turn image chat, visual question answering, and instruction following. Use for vision-language chatbots or image understanding tasks. Best for conversational image analysis.