Atarus Certifications
Practical certifications, built in the open.
Each Atarus credential is one hands-on exam in a real environment, graded against a rubric we publish before anyone sits it. The AOCP is the first credential. Its founding list is open now. Its 25 founding seats are coming soon. Later credentials are in development.
Atarus Offensive Cloud Professional
Eight hours in a real AWS account, four flagged objectives, and a report an engineer can reproduce every finding from, graded against the published rubric with a person confirming every result. The final blueprint and standards are in development.
- Founding list
- Open now
- Founding seats
- Coming soon
- Public exam
- Coming soon
Credentials
The credential family.
Each credential gets a public blueprint and a standards page before anything goes on sale. No tile carries a date.
- Founding list open
AOCP
Atarus Offensive Cloud Professional
AWS accounts, identity, escalation, secrets, logging and the report.
Final blueprint and standards: in development. Founding seats: coming soon.
- In development
AONP
Atarus Offensive Network Professional
Internal networks, Active Directory and Windows, from a foothold to the domain.
Coming soon.
- In development
AOWP
Atarus Offensive Web Professional
Web applications and APIs: authentication, authorization, business logic and the report.
Coming soon.
- In development
AOIP
Atarus Offensive Identity Professional
Entra ID and Okta, sessions and tokens, help desks and the controls around them.
Coming soon.
- In development
AOAP
Atarus Offensive AI Professional
LLM applications and agents: prompt injection, tool permissions, retrieval boundaries and reporting.
Coming soon.
What is tested
What every Atarus credential tests.
The nouns change with the credential. In the cloud exam the identity is an IAM role and the log is CloudTrail; in the network exam it is a domain account and the event log. The six areas stay the same, and each blueprint says how much weight its exam puts on each one.
-
Identity enumeration and access mapping
Who the credentials you hold belong to, what applies to them, and every trust or path that leads somewhere else.
-
Escalation paths through misconfiguration
Find the path, prove it with the smallest action that shows it works, and quote the setting that grants it.
-
Secrets, storage and data exposure
Where secrets sit, who can read them, and what the system exposes to the outside. Names and locations, never values.
-
Guardrails, logging and the defender's view
What the logs recorded for each step you took, what should have alerted, and the gap that let you through.
-
Persistence and proving impact without damage
Footholds found and left in place for the owner. Impact shown with a count, one record and a hash, not a dump.
-
Evidence and reporting
A report an engineer can reproduce every finding from, with the fix and a one-command retest.
Who they are for
People who test systems for a living, or run them.
You should be comfortable on a command line in the area the credential covers. You do not need the paid course to sit an exam: everything an exam rests on is on the public Learn pages its blueprint cites.
-
Penetration testers.
You test for a living and want a practical credential for the work you do, starting with cloud. The exam is the work itself, in a real environment, with a report at the end.
-
Engineers who run the systems.
You run the account, the domain, the application or the identity provider. You want to know what an assessment finds, how it proves it, and what the fix looks like in policy and code.
-
Detection and response.
You read the logs for a living. You want the attacker's side of them: which step produced which event, and what should have alerted.
The AOCP exam
Eight hours, four objectives, one report.
The AOCP is the first exam in the family, and this is how it runs. The objectives in a set follow one shape: from a CI deploy key, map the identity and reach a second role through a trust policy misconfiguration; find the secret names exposed in function configuration and user data across two regions without printing a value; find the guardrail gap that let the escalation go unseen and show what the defender would have seen; prove impact without damage by reading one named object and one named record.
- Where
- A leased AWS account, built for your sitting and wiped after it. AWS only in version 1.
- How long
- Eight hours of lab time. Your credentials are a role session that expires when the lease ends, not later.
- What you do
- Four flagged objectives plus the report. Each objective has a flag you submit and an evidence requirement you meet in the report.
- The report
- Due 48 hours after the lab window closes. Every finding carries reproduction steps, where the evidence is, why the severity is what it is, the fix and the defender's view.
- When
- Weekday sittings. The start times are published with the final blueprint.
- Your layout is your own
- Each sitting is generated from one of three templates. Names, account ids, role names, bucket names, parameter paths and flag values differ per sitting.
- AI assistants
- Allowed in the lab window and for the report, as long as your report declares each one and what you used it for.
- No recording
- No webcam, screen or room recording, and no identity or selfie check. Your evidence and the lab's logs carry the exam.
Extra time and breaks are agreed before you book, on request, and never appear on the credential.
How it is built and graded
A published rubric and a conjunctive standard.
Your work is scored against a rubric of 100 points. To pass you need all of the following: at least three of the four objectives, every must-pass report item (no secret values printed, every finding reproducible, no out-of-scope action in the telemetry), and a total at or above the cut. A high score does not make up for a printed secret or an action outside scope.
Outside reviewers who do not sit the exam set the cut. It is checked against the beta results and labelled provisional for the first 50 sittings. Every failing report and every report near the cut is graded twice. Pass rates are published every quarter with the number of attempts behind them, bad quarters included.
| Part | Points | Anchor |
|---|---|---|
| Objectives and flags | 40 | 10 per objective; partial credit for a documented path that stops at the last step |
| Evidence quality | 20 | every claim tied to a command and output in the evidence package; no secret values printed |
| Severity and impact reasoning | 10 | severity matches what the evidence proves; impact in business terms |
| Remediation | 10 | the specific policy, setting or record, with the caveat to test in non-production |
| Defender's view | 10 | what CloudTrail showed for each step, and what would have caught it |
| Clarity and reproducibility | 10 | an engineer can reproduce every finding from the report alone |
Grading and integrity
AI helps the grader. A person decides.
Reports are graded against the published rubric with the help of an AI tool, and a person confirms every result. You can ask for a human re-grade at no cost. You may use AI assistants yourself if you declare them in your report, and every finding has to stand on your own lab evidence and the lab's logs.
-
The AI tool suggests, and shows its working.
It reads your report and evidence against the rubric one part at a time, suggests a score for each part and points to the passages it relied on. It never decides your result, and it does not check your flags: each flag is matched exactly against the value made for your sitting.
-
A person confirms every result.
A trained grader reads your whole report, checks every finding against your lab evidence and the lab's logs, and changes any suggested score they disagree with. No result is released until a person has confirmed it.
-
A human re-grade, if you want one.
Ask, and your report is graded again by hand, against the same rubric and without the AI's suggestions, at no cost. How to ask and the time limits are published with the final standards.
-
Your AI assistants, declared.
You may use AI assistants in the lab window and for your report. A section of your report lists each one and what you used it for, or says you used none. A declaration later shown to be false is grounds for revoking the credential.
-
Your evidence carries every finding.
Every finding must rest on your own lab evidence and the lab's own logs, which CloudTrail writes to an account you cannot reach (not the trail inside your lab account). A claim the logs do not support fails that item, however well it is written.
-
No recording, no identity check.
No webcam, screen or room recording, and no identity or selfie check. Integrity rests on the lab's logs, a layout generated for your sitting alone, flags made for your sitting, and grading that tests every finding against your evidence.
Timeline
The steps, in order.
In development. Each step happens after the one before it has shown there is demand for it.
Founding seats
25 founding seats: coming soon. The price is announced with the founding terms, before the sale opens.
A founding seat is the course plus a voucher for the beta exam, with the first renewal free. Founders sit the beta exam. Founders' credentials carry no special marking. The verifier shows the exam and blueprint version each holder sat, and the standards page describes the beta.
-
Blueprint and standards published
Blueprint v1.0, the standards page and the scoring rubric go live.
-
Free challenge opens
A cloud challenge on a recorded-response range: map the identity, find the path, write it up. Free with a reader account, and a signed badge delivered by email.
-
Founding seats on sale
25 founding seats: the course, a voucher for the beta exam and the first renewal free. The price is announced with the founding terms, before the sale opens.
-
Course opens to founders
Written walkthroughs, guided labs on real sandbox accounts, the evidence method and a practice exam with a graded sample report.
-
Graded sample report
A report on the practice template, graded against the rubric with every score explained.
-
Course on public sale
The course opens to everyone. You do not need it to sit the exam.
-
Beta exam
Founders sit the exam first. Outside reviewers set the provisional pass standard the week before; the results are published with it.
-
Public exam
The exam goes on public sale under the firm name.
Founding list
Hear when the blueprint, the challenge and the founding seats open.
The founding list is how we tell you when the blueprint is final, when the free challenge opens and when the 25 founding seats go on sale. We ask for your name, organization and work email, and a line on what you want from it. If you are not with an organization, put Independent. We write to you about this programme only, and the list is deleted if unused after 12 months.
You can also email info@atarussecurity.com with "AOCP founding list" in the subject.
What the list gets you
- The final blueprint and the standards page on the day they publish.
- The free challenge when it opens.
- The founding terms and the price, with the 25 seats, before the sale opens.
- Nothing else. The list is used for this programme and is not passed to anyone.
Joining the list commits you to nothing. See our privacy policy.
Issuer and accreditation
An Atarus credential records that the holder passed one practical exam against a named blueprint version on a given date, and the verifier page shows exactly that. We make no claim about how any employer values it, and we do not promise job or salary outcomes.
Atarus certifications are issued by Atarus Offensive Security LLC. They are not accredited by ANAB, ICE or any other accreditation body, and they are not approved under DoD 8140 or any government program. Our standards page shows exactly how each exam is built, graded and reviewed, and our pass rates are published every quarter. If our status changes, this page will say so, with the date and scope.