Skip to content
Gauge

ax-check.com

Astronomer

astronomer.io

ax-check.com

Checked 10 November 2026

A

100/100

Agents need your login

Astronomer's docs, pricing, and CLI quickstart are agent-ready.

22 of 23 checklist items passed, including stated per-hour pricing and a fully actionable CLI quickstart with no login wall.

1 fix to start with

Concrete changes, backed by what we found.

Onboarding

Give agents a clear login handoff and a way to resume setup.

Agents cannot progress through onboarding without human help. Provide an agent-accessible path through authentication and setup.

Recorded coding sessions

Same task, 3 models, no astronomer.io account. Recorded 10 November 2026 on Pi.

3 recorded · 49s

49s · recorded

Session findings

Select a finding to see it in context.

piv0.84.2
modelQwen 3.8 Maxcwd/sandbox/repo
Help me build a simple example using Astronomer. Tell me how pricing works, and briefly tell me whether this product will be easy for you to manage. Let me know if you get blocked. If this product has no developer workflow you can act on, say so plainly and stop. Stay light: use the hosted product through its SDK or API. Do not start local service stacks or wait for long-running commands; if the quickstart requires either, say so plainly and stop.
Thinking…

The user is asking about Astronomer, a managed Apache Airflow service, and wants to understand how to build an example, its pricing model, and whether it's easy to manage. I'm realizing this is a third-party product, so I need to clarify what they're actually looking for since Astronomer isn't the pi product I'm focused on.

I'm checking whether I have network access and any existing credentials or API keys to work with Astronomer's API or SDK, and I should also look at the repository context to see if there's anything already set up for this.

I'll check the environment first to see what's actually available to act on.

ls -la /sandbox/repo 2>/dev/null | head -50; echo "---"; env | grep -iE 'astronomer|astro|airflow' | sed 's/=.\{4\}.*/=<redacted>/'; echo "--- net test ---"; timeout 12 curl -sS -o /dev/null -w '%{http_code}\n' https://api.astronomer.io/ 2>&1 | tail -3
total 16
total 16
drwxr-xr-x 3 root root 4096 Sep 18 08:00 .
drwxr-xr-x 7 root root 4096 Nov 10 16:59 ..
drwxr-xr-x 8 root root 4096 Sep 24 20:05 .git
-rw-r--r-- 1 root root  408 Sep 18 08:00 .gitignore
---
--- net test ---
403
which astro python3 pip3 2>&1; echo "---"; python3 -c "import astronomer" 2>&1 | tail -2; echo "--- pip index ---"; timeout 20 pip3 index versions astronomer 2>&1 | tail -3
/usr/local/bin/python3
/usr/local/bin/python3
/usr/local/bin/pip3
---
  File "<string>", line 1, in <module>
ModuleNotFoundError: No module named 'astronomer'
--- pip index ---

[notice] A new release of pip is available: 26.2.1 -> 26.3
[notice] To update, run: /opt/freestyle/python/bin/python3 -m pip install --upgrade pip
Thinking…

Findings:

  • Repo is essentially empty (just .git and .gitignore).
  • No Astronomer/Airflow credentials in env.
  • Network reaches api.astronomer.io but returns 403 (needs auth token) — so there IS network, but no credentials.
  • No astro CLI installed.
  • pip index versions astronomer returned nothing visible (let me check whether pip install works at all — network to PyPI).

Let me test whether pip can reach PyPI, and whether there's an Astronomer CLI installable. But more importantly: there are no credentials. Astronomer is a managed Airflow platform (Astro Cloud) — using it requires an Astronomer account, an org, and an API token, plus deploying a DAG to a Deployment. Without credentials, I can't act on the hosted product.

The instruction: "If this product has no developer workflow you can act on, say so plainly and stop." Astronomer DOES have a developer workflow (astro CLI, DAG deploys, Airflow REST API, Astronomer API). But I'm blocked on credentials — no account/token. Also, the quickstart requires astro dev start which spins up a local Docker stack of Airflow — that's exactly the "local service stack" the user told me not to start. And the hosted path requires an account + org + deployment provisioning, which requires credentials I don't have.

Let me verify PyPI reachability quickly and whether astro CLI is installable, so my report is grounded. Also check if docker is present (to confirm local quickstart would need it). Keep it light — a couple of quick checks.

Also worth checking: pip download of astro CLI package name astronomer-cli? Let's do quick checks with short timeouts.

Pricing: Astronomer pricing — I should be careful and honest. Astronomer has Astro Cloud pricing tiers: Free trial, and then plans based on... Historically Astro pricing was based on "Astro Compute" units / resource-based pricing per deployment (small/medium/large), with tiers: Free Trial, then paid plans (Basic/Standard/Premium?) and enterprise. There's also astro deploy with "usage-based" pricing on Cloud. I don't have reliable current numbers without network access to their pricing page. Let me try fetching the pricing page quickly — network seems partially available (403 from api.astronomer.io means DNS/TLS works). Try curl on astronomer.io/pricing with short timeout. If blocked, I'll say I can't verify current numbers and point to the source rather than inventing figures.

Let me run those checks in parallel.

Network reaches Astronomer but returns 403 (no credentials). Let me verify the two paths — hosted API/SDK and the documented quickstart — before I write anything.

cd /tmp && timeout 25 pip3 download --no-deps -d /tmp/pipdl astronomer-cli 2>&1 | tail -4; echo "=== docker? ==="; which docker docker-compose 2>&1 | head; echo "=== creds present? ==="; ls ~/.astro 2>&1 | head -3
[notice] A new release of pip is available: 26.2.1 -> 26.3

[notice] A new release of pip is available: 26.2.1 -> 26.3
[notice] To update, run: /opt/freestyle/python/bin/python3 -m pip install --upgrade pip
ERROR: No matching distribution found for astronomer-cli
=== docker? ===
/usr/bin/docker
=== creds present? ===
ls: cannot access '/sandbox/.astro': No such file or directory
timeout 20 curl -sSL -o /tmp/pricing.html -w 'HTTP:%{http_code}\n' https://www.astronomer.io/pricing 2>&1 | tail -2; echo "--- size ---"; wc -c /tmp/pricing.html 2>/dev/null
HTTP:200
HTTP:200
--- size ---
141491 /tmp/pricing.html
Thinking…

Pricing page fetched. Let me extract text to get actual pricing facts rather than guessing.

Also PyPI: no astronomer-cli distribution — the astro CLI is distributed via curl install script / npm? Actually astro CLI is installed via curl -sSL install.astronomer.io | sudo bash -s -- -o <dir> or homebrew. PyPI package might be astro-cli? Let me check quickly. But the key blocker: no account credentials.

Let me extract pricing text.

Pricing page is reachable — let me pull the actual figures rather than guess.

cd /tmp && python3 - <<'PY'
import re,html
s=open('/tmp/pricing.html',encoding='utf-8',errors='ignore').read()
s=re.sub(r'<(script|style)[^>]*>.*?</\1>','',s,flags=re.S|re.I)
t=html.unescape(re.sub(r'<[^>]+>',' ',s))
t=re.sub(r'\s+',' ',t)
for kw in ['Free','trial','$','per ','Trial','Enterprise','Resource','usage','Usage','Contact']:
    for m in re.finditer(re.escape(kw),t):
        seg=t[max(0,m.start()-110):m.start()+150].strip()
        print(f'[{kw}] ...{seg}...')
    print('=====')
PY
[Free] ...flow 3 Astro vs. Apache Airflow OSS Airflow Use Cases Airflow Security Pricing Log In Book a Demo Get Started Free Pricing Flexible, usage-based prici
[Free] ...flow 3 Astro vs. Apache Airflow OSS Airflow Use Cases Airflow Security Pricing Log In Book a Demo Get Started Free Pricing Flexible, usage-based pricing for teams of all sizes. Run Apache Airflow without the overhead. Astro's usage-based model means you only p...
[Free] ...Monthly Annual Developer For teams building and testing pipelines. Deployments starting at $0.35/hr. Try for Free AI-powered Dag authoring & debugging Flexible, scale-to-zero compute Hibernating deployments API access Deployment rollbacks Compare this plan →...
[Free] ...production pipelines that require Airflow support & observability. Deployments starting at $0.42/hr. Try for Free Everything in Developer , plus: End-to-end observability & data quality Network isolation with Dedicated clusters High availability deployments A...
[Free] ...rise discount programs apply. AWS Marketplace Available with flexible pay-as-you-go or annual agreements. Try Free with AWS → Azure Marketplace Available with annual agreements. View in Marketplace → Google Cloud Marketplace Available with annual agreements. V...
=====
=====
[$] ...your environment. Monthly Annual Developer For teams building and testing pipelines. Deployments starting at $0.35/hr. Try for Free AI-powered Dag authoring & debugging Flexible, scale-to-zero compute Hibernating deployments API access Deployment rollbacks Co...
[$] ...For teams running production pipelines that require Airflow support & observability. Deployments starting at $0.42/hr. Try for Free Everything in Developer , plus: End-to-end observability & data quality Network isolation with Dedicated clusters High availabi...
[$] ...ant clusters are included on all plans at no cost. Dedicated single-tenant clusters are available starting at $2.40/hr for enhanced networking and security. Deployment pricing Each deployment includes the scheduler, webserver, and triggerer needed to run your...
[$] ...r pay for unused capacity. Choose from 7 worker sizes to meet your task requirements (A5 to A160) starting at $0.13/hr. *Networking costs are passed through from the cloud provider. Observe pricing Astro Observe provides detailed cost and performance metrics,...
=====
[per ] ..., powered by Apache Airflow. Private Cloud Run Airflow-as-a-service in your environment. Monthly Annual Developer For teams building and testing pipelines. Deployments starting at $0.35/hr. Try for Free AI-powered Dag authoring & debugging Flexible, scale-to-z...
[per ] ...t require Airflow support & observability. Deployments starting at $0.42/hr. Try for Free Everything in Developer , plus: End-to-end observability & data quality Network isolation with Dedicated clusters High availability deployments Audit logging (7-day reten...
=====
=====
[Enterprise] ...CI/CD enforcement Audit logging (90-day retention) 24x7 support availability (1 hour SLA) Compare this plan → Enterprise For teams managing Airflow at scale that require enterprise security & governance. Get a Quote Everything in Business , plus: Remote execut...
[Enterprise] ...ote execution agents Custom RBAC SCIM provisioning IP access list Organization dashboards Compare this plan → Enterprise Business Critical For platform teams running Airflow where resiliency and rapid support response are essential. Get a Quote Everything in E...
[Enterprise] ...rm teams running Airflow where resiliency and rapid support response are essential. Get a Quote Everything in Enterprise , plus: Cross-region disaster recovery 24x7 support availability (30 min SLA) Compare this plan → Astro Private Cloud Run Airflow-as-a-serv...
[Enterprise] ...ce in your environment. Manage deployments across multiple clusters and regions Air-gapped deployment support Enterprise SSO integration Deployment isolation Dynamic resources management In-place Airflow upgrades Houston API access Professional services instal...
=====
[Resource] ...ow 2026 See the insights from over 5,800 data practitioners in the full report. Download Now ➔ Customers Docs Resources Insights Blog Webinars Resource Library Events Education Academy What is Airflow? Airflow What is Airflow? Airflow 3 Astro vs. Apache Airflo...
[Resource] ...r 5,800 data practitioners in the full report. Download Now ➔ Customers Docs Resources Insights Blog Webinars Resource Library Events Education Academy What is Airflow? Airflow What is Airflow? Airflow 3 Astro vs. Apache Airflow OSS Airflow Use Cases Airflow S...
[Resource] ...l Analytics All Use Cases Industries Financial Services Gaming Retail Manufacturing Healthcare All Industries Resources Academy eBooks & Guides Blog Webinars Events The Data Flowcast Podcast Beyond the Dag Hackathon All Resources Airflow What is Airflow Airflo...
[Resource] ...Resources Academy eBooks & Guides Blog Webinars Events The Data Flowcast Podcast Beyond the Dag Hackathon All Resources Airflow What is Airflow Airflow on Astro Airflow 3.0 Airflow Upgrades Airflow Use Cases Airflow 2.x End of Life Company About Us Customers N...
=====
[usage] ...Airflow OSS Airflow Use Cases Airflow Security Pricing Log In Book a Demo Get Started Free Pricing Flexible, usage-based pricing for teams of all sizes. Run Apache Airflow without the overhead. Astro's usage-based model means you only pay for the compute reso...
[usage] ...ricing Flexible, usage-based pricing for teams of all sizes. Run Apache Airflow without the overhead. Astro's usage-based model means you only pay for the compute resources you actually use. Clusters, deployments, and workers scale with your workload. Get a Qu...
[usage] ...kages through Astronomer Professional Services. Learn more → How Astro pricing works Astro’s pricing model is usage-based. Deployments and Workers are billed separately on a metered basis, with additional costs for dedicated clusters, Astro Observe, and AI tok...
[usage] ...lled separately on a metered basis, with additional costs for dedicated clusters, Astro Observe, and AI token usage. Cluster pricing Standard multi-tenant clusters are included on all plans at no cost. Dedicated single-tenant clusters are available starting at...
=====
=====
[Contact] ...ow 3.0 Airflow Upgrades Airflow Use Cases Airflow 2.x End of Life Company About Us Customers Newsroom Careers Contact Support Knowledge Base Status Contact Support Legal Privacy Terms of Service Consent Preferences Do Not Sell or Share My Personal information...
[Contact] ...ses Airflow 2.x End of Life Company About Us Customers Newsroom Careers Contact Support Knowledge Base Status Contact Support Legal Privacy Terms of Service Consent Preferences Do Not Sell or Share My Personal information Limit the Use Of My Sensitive Personal...
=====
Thinking…

Now I have real pricing data. Let me also confirm the API auth requirement and the "no credentials" blocker concretely — try the Astro API with a fake token to show 401, and check the docs quickstart requirement (astro dev start = Docker local stack). I already know docker exists but starting it is disallowed by the user's constraint.

Let me do one quick confirmation of the API requiring a token (already saw 403 unauthenticated). Let me quickly check what the 403 body says, and check whether a personal access token env var exists anywhere. That's enough.

Actually let me just do one quick call to show the API auth model (Bearer PAT), which confirms the path exists but needs credentials.

timeout 15 curl -sS https://api.astronomer.io/query -X POST -H 'Content-Type: application/json' -d '{"query":"{ organizations { id } }"}' 2>&1 | head -c 300; echo; echo "=== any token in env/config? ==="; env | grep -icE 'ASTRO|ASTRONOMER|AIRFLOW'; ls -a /sandbox 2>/dev/null | head
RBAC: access denied
RBAC: access denied
=== any token in env/config? ===
0
.
..
.cache
.gauge
.pi
.tmp
repo
Thinking…

Confirmed: the Astro API is live and reachable but requires a Personal Access Token / OAuth, and there are zero Astronomer credentials in this environment. No ~/.astro config, no env tokens, no account.

So: Astronomer DOES have a real developer workflow (astro CLI, Astro API/Houston GraphQL, Airflow REST API on a deployment). But I cannot act on it without: (1) an Astronomer account + org, (2) a Personal Access Token, (3) a Deployment. And the documented local quickstart (astro dev start) requires a Docker-based local Airflow stack — which the user explicitly told me not to start.

Per instructions: I'm blocked. I should say so plainly and stop. Don't fabricate a fake example pretending it works.

Should I write anything? The instruction says "If this product has no developer workflow you can act on, say so plainly and stop." Astronomer has a workflow, but I can't act on it (no credentials; hosted path needs an account+token; local path needs a Docker stack which is out of scope). So the honest answer: blocked, and stop. I should not scaffold files that can't be verified — that would be pretending.

I'll report:

  1. What I verified (concrete evidence).
  2. Why I'm blocked and what exactly I need (one token + org/deployment).
  3. Pricing — from the actual pricing page I fetched, with real numbers.
  4. Manageability assessment — brief and honest.
  5. Stop.

Keep it concise. No files written. Mention the repo is empty (just .git/.gitignore).

Pricing facts gathered:

  • Usage-based, metered; pay only for compute actually used.
  • Plans (Cloud): Developer (deployments starting $0.35/hr), Business ($0.42/hr), Enterprise (quote), Business Critical (quote). Private Cloud also listed.
  • Clusters: standard multi-tenant included at no cost; dedicated single-tenant from $2.40/hr.
  • Deployment pricing includes scheduler, webserver, triggerer.
  • Workers billed separately, 7 sizes A5→A160 starting $0.13/hr.
  • Extra costs: dedicated clusters, Astro Observe, AI token usage; networking passed through from cloud provider.
  • Scale-to-zero / hibernating deployments on Developer plan; API access included on Developer.
  • Also purchasable via AWS/Azure/GCP marketplaces.
  • "Try for Free" / Get Started Free — free trial exists.

Manageability for me (the agent):

  • Easy parts: everything is API-first — Astro API (GraphQL at api.astronomer.io/query) with a PAT, Airflow REST API per deployment, DAGs are plain Python files in git, so dag-authoring/versioning is agent-friendly and diffable.
  • Hard parts: requires a paid cloud account + org provisioning, tokens I can't self-serve; the local dev loop (astro dev start) is a Docker stack = slow, stateful, outside the "stay light" constraint; deployment/run feedback loops are long-running (deploy → wait for DAG parse → trigger → poll), and logs come back through API calls. Cost metering means mistakes burn money while running.
  • Verdict: moderately manageable once a token exists — the code side is easy, the infra/auth side is gated on a human.

I'll state clearly I stopped and wrote no files.

Blocked — stopping here

Astronomer (Astro) does have a real developer workflow, so this isn't a "no workflow" case. But I can't act on it in this environment, and I won't fake it.

What I verified:

CheckResult
api.astronomer.io reachableYes — but unauth'd GraphQL returns RBAC: access denied / 403
Astronomer credentials in env or ~/.astroNone (0 matching vars, no config dir)
astro CLI / astronomer SDKNot installed; astronomer-cli isn't on PyPI (CLI ships via install script/brew)
/sandbox/repoEmpty — only .git and .gitignore
Local quickstartastro dev start = Docker-based local Airflow stack → excluded by your constraint

So both paths are closed: the hosted path needs an Astronomer account, an org, a Deployment, and a Personal Access Token — none of which I can self-serve — and the local path needs the long-running Docker stack you told me not to start. I wrote no files rather than scaffold an example I can't run or verify.

To unblock, I need exactly one thing from you: an Astro Personal Access Token plus your org/deployment ID (or a deployment's Airflow REST endpoint + token). With that, the light path is fully API-driven: upload a DAG file to the deployment, trigger it via the Airflow REST API, poll state. No containers.

How pricing works

Pulled from astronomer.io/pricing (live, fetched just now — not from memory):

  • Model: usage-based/metered. You pay for compute actually consumed; deployments and workers are billed separately.
  • Plans: Developer — deployments from $0.35/hr; Business — from $0.42/hr; Enterprise and Business Critical — quote only. A free "Try for Free" tier exists.
  • Clusters: standard multi-tenant clusters are included at no cost on all plans; dedicated single-tenant clusters from $2.40/hr.
  • Workers: billed separately, 7 sizes (A5 → A160) starting $0.13/hr.
  • What a deployment covers: scheduler + webserver + triggerer.
  • Extras: dedicated clusters, Astro Observe, and AI token usage cost more; networking is passed through from your cloud provider.
  • Cost control: scale-to-zero compute and hibernating deployments are on the Developer plan, so an idle test deployment can sit near $0. Also buyable through AWS/Azure/GCP marketplaces.

Practical read for an example like this: a small deployment left running is roughly $0.35/hr + worker time, so the money risk is forgetting to hibernate/delete it, not the build itself.

Will this be easy for me to manage?

Mixed — the code is easy, the platform is gated on you.

  • Easy: DAGs are plain Python files in git, so authoring, diffing, and versioning is squarely in my wheelhouse. The API surface is real (Astro GraphQL API + per-deployment Airflow REST API), so deploying, triggering, and reading logs are all scriptable without a UI.
  • Not easy: auth and provisioning are human-only — I can't create the account, org, or token. Feedback loops are long (deploy → wait for DAG parsing → trigger → poll), which pushes against your "no long-running waits" rule. And because billing is metered per hour, a mistake doesn't just fail — it keeps costing money until someone deletes the deployment.

Net: once a token exists, I'd call it moderately manageable — a few API calls per iteration, no local stack. Without one, it's a hard stop, which is where we are.

AX Fundamentals

23 checks. Open an area for details.

This check is a quick preview. Test more tasks and models with Gauge Agents.

Test your product with Gauge Agents

AX Check by Gauge