# Amazon MWAA adds AI agent skills for workflow creation

> Amazon Managed Workflows for Apache Airflow now lets AI assistants author, debug, test and upgrade pipelines via natural‑language commands.

Oossa · 2026-10-15 · https://oossa.com/en/amazon-mwaa-adds-ai-agent-skills-for-workflow-creation

Amazon Managed Workflows for Apache Airflow (MWAA) now includes four AI agent skills. The skills—authoring, debugging, testing, and upgrading—let any AI coding assistant that supports agent skills interact with MWAA in plain language. You can describe an ETL pipeline and get a ready‑to‑run DAG, or ask the assistant to debug or upgrade an environment. The features work with both provisioned and serverless MWAA and are available in every AWS Region where MWAA runs.

## The facts

- Announcement date: 2026-10-15
- Four new skills: authoring‑mwaa‑workflow, debugging‑mwaa‑workflow, testing‑mwaa‑workflow, upgrading‑mwaa‑environments

## Why it matters

Data engineers can cut manual coding time by using natural‑language prompts to build and maintain Airflow pipelines.

## Sources & references

1. [Amazon Managed Workflows for Apache Airflow now offers AI agent skills](https://aws.amazon.com/about-aws/whats-new/2026/10/amazon-mwaa-ai-agent-skills/) – AWS, 2026-10-15

Last updated: 2026-10-15