How do I install the Nixtla Prod Pipeline Generator skill?

Run npx skillmds@latest add intent-solutions-io/nixtla-prod-pipeline-generator-2 in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Nixtla Prod Pipeline Generator safe to use?

This skill has not completed SkillMD's automated safety review yet. Independent scanners report: SkillSpector: PASS, Skill Scanner: PASS. Capability flags: executes scripts, makes network calls. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Nixtla Prod Pipeline Generator?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Nixtla Prod Pipeline Generator free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.

Who published Nixtla Prod Pipeline Generator?

intent-solutions-io (@intent-solutions-io) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Nixtla Prod Pipeline Generator by intent-solutions-io

Name: Nixtla Prod Pipeline Generator
Rating: 4.2 (5 reviews)
Author: intent-solutions-io

Nixtla Prod Pipeline Generator

Transforms forecasting experiments into production-ready inference pipelines with Airflow, Prefect, or cron orchestration. Generates ETL tasks, monitoring, error handling, and deployment configs. Activates when user needs to deploy forecasts to production, schedule batch inference, operationalize models, or create production pipelines.

File contents

Nixtla Production Pipeline Generator

Transform validated forecasting experiments into production-ready inference pipelines with proper orchestration, monitoring, and error handling.

Overview

This skill productionizes Nixtla forecasting workflows by generating complete deployment artifacts:

Airflow DAGs: Enterprise orchestration with dependencies and monitoring
Prefect Flows: Modern Python-native pipelines with better local testing
Cron Scripts: Simple single-machine batch processing

All pipelines implement: Extract -> Transform -> Forecast -> Load -> Monitor

Prerequisites

Required:

Python 3.8+
Completed experiment in forecasting/config.yml
One of: Airflow, Prefect, or cron access

Environment Variables:

NIXTLA_API_KEY: TimeGPT API key (if using TimeGPT)
FORECAST_DATA_SOURCE: Production data connection string
FORECAST_DESTINATION: Output destination for forecasts

Installation:

pip install nixtla pandas statsforecast  # Core
pip install apache-airflow  # For Airflow
pip install prefect  # For Prefect

Instructions

Step 1: Read Experiment Config

Load experiment from forecasting/config.yml:

python {baseDir}/scripts/read_experiment.py --config forecasting/config.yml

Step 2: Select Orchestration Platform

Choose based on requirements:

Airflow: Enterprise, complex dependencies, extensive monitoring
Prefect: Python-native, better local testing, modern error handling
Cron: Simple single-machine, no dependencies, quick setup

Step 3: Generate Pipeline

python {baseDir}/scripts/generate_pipeline.py \
    --config forecasting/config.yml \
    --platform airflow \
    --output pipelines/

Step 4: Add Monitoring

python {baseDir}/scripts/add_monitoring.py \
    --pipeline pipelines/forecast_dag.py \
    --metrics smape,mase

Step 5: Deploy

Follow generated pipelines/README.md for deployment instructions.

Output

pipelines/forecast_dag.py: Main pipeline file (Airflow/Prefect/Cron)
pipelines/monitoring.py: Quality checks and fallback logic
pipelines/README.md: Deployment instructions
pipelines/requirements.txt: Dependencies

Error Handling

Error: Config file not found Solution: Run nixtla-experiment-architect first to create config
Error: NIXTLA_API_KEY not set Solution: Export your TimeGPT API key or use StatsForecast baselines
Error: Database connection failed Solution: Verify FORECAST_DATA_SOURCE connection string
Error: Forecast quality check failed Solution: Pipeline auto-falls back to baseline models

Examples

Example 1: Airflow DAG

python {baseDir}/scripts/generate_pipeline.py \
    --config forecasting/config.yml \
    --platform airflow \
    --schedule "0 6 * * *" \
    --output pipelines/

Output:

Generated: pipelines/forecast_dag.py
Schedule: Daily at 6am
Tasks: extract -> transform -> forecast -> load -> monitor

Example 2: Simple Cron Script

python {baseDir}/scripts/generate_pipeline.py \
    --config forecasting/config.yml \
    --platform cron \
    --output pipelines/

Resources

Scripts: {baseDir}/scripts/
Templates: {baseDir}/assets/templates/
Nixtla Docs: https://nixtla.github.io/

Related Skills:

nixtla-experiment-architect: Creates experiments to productionize
nixtla-timegpt-finetune-lab: Fine-tuned models for pipelines
nixtla-usage-optimizer: Cost-effective routing strategies

---name: nixtla-prod-pipeline-generator-2description: Transforms forecasting experiments into production-ready inference pipelines with Airflow, Prefect, or cron orchestration. Generates ETL tasks, monitoring, error handling, and deployment configs. Activates when user needs to deploy forecasts to production, schedule batch inference, operationalize models, or create production pipelines.license: MIT---# Nixtla Production Pipeline GeneratorTransform validated forecasting experiments into production-ready inference pipelines with proper orchestration, monitoring, and error handling.## OverviewThis skill productionizes Nixtla forecasting workflows by generating complete deployment artifacts:- **Airflow DAGs**: Enterprise orchestration with dependencies and monitoring- **Prefect Flows**: Modern Python-native pipelines with better local testing- **Cron Scripts**: Simple single-machine batch processingAll pipelines implement: Extract -> Transform -> Forecast -> Load -> Monitor## Prerequisites**Required**:- Python 3.8+- Completed experiment in `forecasting/config.yml`- One of: Airflow, Prefect, or cron access**Environment Variables**:- `NIXTLA_API_KEY`: TimeGPT API key (if using TimeGPT)- `FORECAST_DATA_SOURCE`: Production data connection string- `FORECAST_DESTINATION`: Output destination for forecasts**Installation**:```bashpip install nixtla pandas statsforecast # Corepip install apache-airflow # For Airflowpip install prefect # For Prefect```## Instructions### Step 1: Read Experiment ConfigLoad experiment from `forecasting/config.yml`:```bashpython {baseDir}/scripts/read_experiment.py --config forecasting/config.yml```### Step 2: Select Orchestration PlatformChoose based on requirements:- **Airflow**: Enterprise, complex dependencies, extensive monitoring- **Prefect**: Python-native, better local testing, modern error handling- **Cron**: Simple single-machine, no dependencies, quick setup### Step 3: Generate Pipeline```bashpython {baseDir}/scripts/generate_pipeline.py \ --config forecasting/config.yml \ --platform airflow \ --output pipelines/```### Step 4: Add Monitoring```bashpython {baseDir}/scripts/add_monitoring.py \ --pipeline pipelines/forecast_dag.py \ --metrics smape,mase```### Step 5: DeployFollow generated `pipelines/README.md` for deployment instructions.## Output- **pipelines/forecast_dag.py**: Main pipeline file (Airflow/Prefect/Cron)- **pipelines/monitoring.py**: Quality checks and fallback logic- **pipelines/README.md**: Deployment instructions- **pipelines/requirements.txt**: Dependencies## Error Handling1. **Error**: `Config file not found` **Solution**: Run `nixtla-experiment-architect` first to create config2. **Error**: `NIXTLA_API_KEY not set` **Solution**: Export your TimeGPT API key or use StatsForecast baselines3. **Error**: `Database connection failed` **Solution**: Verify `FORECAST_DATA_SOURCE` connection string4. **Error**: `Forecast quality check failed` **Solution**: Pipeline auto-falls back to baseline models## Examples### Example 1: Airflow DAG```bashpython {baseDir}/scripts/generate_pipeline.py \ --config forecasting/config.yml \ --platform airflow \ --schedule "0 6 * * *" \ --output pipelines/```**Output**:```Generated: pipelines/forecast_dag.pySchedule: Daily at 6amTasks: extract -> transform -> forecast -> load -> monitor```### Example 2: Simple Cron Script```bashpython {baseDir}/scripts/generate_pipeline.py \ --config forecasting/config.yml \ --platform cron \ --output pipelines/```## Resources- Scripts: `{baseDir}/scripts/`- Templates: `{baseDir}/assets/templates/`- Nixtla Docs: https://nixtla.github.io/**Related Skills**:- `nixtla-experiment-architect`: Creates experiments to productionize- `nixtla-timegpt-finetune-lab`: Fine-tuned models for pipelines- `nixtla-usage-optimizer`: Cost-effective routing strategies

Frequently asked questions