# Mwaa

> Inspect Amazon MWAA (Managed Airflow) environments, mint web-login/CLI tokens, and run the Airflow CLI over the MWAA REST endpoint via the `datus mwaa` CLI

- Skill: `datus-ai/mwaa` (Agent Skill)
- Install (CLI): `npx skillmds@latest add datus-ai/mwaa`
- Raw SKILL.md: https://api.skillmd.com/api/skills/datus-ai/mwaa/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: DevOps & Infra
- Author: datus-ai (https://skillmd.com/u/datus-ai)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/datus-ai/mwaa

---


# MWAA

`datus mwaa` inspects Amazon MWAA environments and reaches their Airflow. It
complements the dedicated `datus airflow` plugin: MWAA manages the
**environment**; for fine-grained, permission-classified DAG operations point
`datus airflow` at this environment instead of using `cli run`. Global usage:

```
datus mwaa [--profile <env>] <group> <subcommand> [args...]
```

## Environments

```
datus mwaa environments list
datus mwaa environments get <name>
```

## Tokens

```
datus mwaa token web-login [<name>]   # one-time Airflow UI login URL
datus mwaa token cli [<name>]         # CLI token + web server hostname
```

`<name>` defaults to the profile's `environment`.

## Airflow CLI over REST

```
datus mwaa cli run '<airflow cli command>' [--env <name>]
# e.g. datus mwaa cli run 'dags list' --env prod
```

`cli run` mints a CLI token and POSTs the command to
`https://<hostname>/aws_mwaa/cli`, printing the decoded stdout (stderr goes to
stderr). It is an **opaque passthrough** — the wrapped command could be
destructive (`dags trigger`, `variables delete`, ...) — so the agent always
confirms it. Not every Airflow CLI command is supported over MWAA's REST
endpoint (e.g. `dags backfill` is restricted); prefer `datus airflow` for rich
operations.

## Exit codes

`0` success · `1` runtime/API error (also: MWAA CLI HTTP error) · `2` usage
(also: no environment) · `3` config error.