Cloud penetration testing
Cloud penetration testing looks for the ways an attacker could take over or abuse your AWS, Azure or Google Cloud environment. HYDN reviews identity and access policies, exposed storage and services, secrets handling, network controls and CI/CD pipelines, then tests what an attacker could reach from a leaked key or a compromised workload.
What we test
- Identity and access. Over-permissive roles and policies, privilege escalation paths, unused keys and cross-account trust.
- Storage. Buckets, blobs and databases that are public or readable by more people than intended.
- Secrets. Keys and credentials in code, environment variables, build logs and instance metadata.
- Compute and containers. Exposed services, container escapes, and Kubernetes RBAC and network policies.
- Network controls. Security groups and firewall rules, and what is reachable from the internet.
- CI/CD. Build pipelines that can be tampered with to ship malicious code to production.
- Logging. Whether an attacker’s actions would be recorded and noticed.
Why it matters for web3
Front ends, APIs and signing infrastructure for many crypto platforms run in the cloud. The Bybit theft in February 2025 was traced to a compromise of a wallet provider’s infrastructure, and security experts suspected a leaked or compromised AWS S3 or CloudFront key as the entry point. Read our post-mortem.
Frequently asked questions
Do we need permission from our cloud provider?
AWS, Azure and Google Cloud all allow customers to test their own resources within published policies, without prior approval for most services. We stay within those policies and agree the scope with you first.
Is a cloud configuration review the same as a penetration test?
A configuration review compares settings against benchmarks. A penetration test goes further and shows what an attacker can do by chaining weaknesses, such as turning a leaked read-only key into administrative access.
Published Updated