Research topic
Workload security
Runtime context decides whether a control matters.
Field guidance for Kubernetes, machine identity, service boundaries, and modern application infrastructure.
View every topicFrom the desk
38 publicationsStop pull_request_target workflows from running fork code
GitHub blocks pull_request_target in public repositories by default from November 2, 2026. Find the workflows that run fork code with secrets, split them, pin a checkout that refuses unsafe refs, and allow the rest deliberately.
Know what service mesh mTLS proves about the caller
A mesh handshake proves that a peer held a key for a namespace and service account. It says nothing about the user or the operation, and permissive defaults, probes and waypoints open paths around it.
Tune managed WAF rules without opening a bypass
Fix a managed WAF false positive by changing one rule for one route, then decide what happens to the request body bytes AWS WAF, Azure WAF and Cloud Armor never inspect.
Use DNS egress filtering without mistaking it for a firewall
Resolver-layer filters in AWS, Azure and Google Cloud decide which names a workload can resolve, not where it connects. What each provider says its DNS control misses, and the connection control to pair with it.
Separate API gateway JWT checks from backend authorization
A gateway that accepts a JWT has checked its signature, issuer, audience and time window. It has not checked token type, revocation, cached decisions or object ownership. This mapping shows which side owns each decision.
Test PostgreSQL row-level security with the application role
Row-level security only filters queries from roles that do not bypass it. Test with the production login role, through the production pooler, with transaction-local tenant context and a patched minor release.
Plan the move from ingress-nginx to Gateway API
Ingress-nginx is archived and will not be patched again. Contain the old controller, inventory the behaviors your clients depend on, and cut over to Gateway API with evidence and a rollback path.
Pin third-party GitHub Actions and limit what they can reach
A tag can be moved under your workflow overnight. Pin every action to a commit SHA, enforce it across the organization, and cut the token and secret access a malicious action would inherit.
Publish npm packages with trusted publishing instead of long-lived tokens
npm revoked classic tokens and capped write tokens at 90 days. Move CI releases to trusted publishing with staged approval, then protect the workflow, tags and environment that now hold publish authority.
Encrypt Kubernetes Secrets at rest with a KMS v2 provider
KMS v2 keeps etcd copies and backups unreadable without an external key, but it does not stop API readers. Configure it, compare what EKS, AKS and GKE already do, and finish key rotations with a rewrite.
Build a cryptographic inventory for the post-quantum transition
Record where public-key cryptography protects long-lived data and which library, service or vendor must change it. Then rank entries by function and data lifetime, not by how many RSA keys a scanner found.
Scan uploaded files for malware before your application trusts them
Hold every upload where nothing can read it, promote only files with a clean verdict, and decide in advance what happens to the files GuardDuty, Defender or a ClamAV deployment cannot scan.
Keep an S3 bucket private with Block Public Access
Enable S3 Block Public Access while preserving approved readers. Review the four settings, private CloudFront origins, and practical access tests.
Remove public SSH access from an EC2 security group
Remove internet-wide EC2 SSH rules after testing Session Manager or restricted SSH. Check all groups, IPv6, fresh sessions, and recovery access.
Keep a Google Cloud Storage bucket private
Use public access prevention and uniform bucket-level access for Cloud Storage, then test IAM readers, signed links, and application dependencies.
Require authentication between Cloud Run services
Configure Cloud Run service-to-service authentication with a dedicated caller identity, the right ID token audience, and useful negative tests.
Check and stop anonymous access to Azure blobs
Understand account and container settings and prove both anonymous denial and intended app access.
Restrict SSH and RDP access with an Azure network security group
Inspect effective rules, preserve the approved management path and test a fresh connection.
Keep Terraform plans and state inside the change boundary
Protect Terraform plans and state as sensitive artifacts, and bind production approval to the specific plan, dependencies, workspace and apply identity that will be used.
Separate webhook authenticity from permission to process it
Verify webhook authenticity at intake, then make durable acceptance, duplicate handling and permission to change business state separate decisions.
Keep build credentials out of the image and its evidence
A temporary BuildKit secret mount controls credential delivery, not everything a build command can do with the credential or leave in its outputs.
Design RDS IAM authentication around the connection pool
Generate IAM database tokens for the physical connections that need them, and keep token validity separate from the lifetime of an already-established SQL session.
Finish S3 multipart uploads with verifiable object integrity
Keep an owned part manifest and verify the complete object result, because successful part transfers and an initial HTTP 200 are not sufficient completion evidence.
Give Kubernetes admission webhooks an explicit failure contract
Treat an admission webhook as a control-plane dependency with explicit behavior for denial, call failure, mutation and the changes needed to repair it.
Questions answered
What does Cloud Security Desk cover under Workload security?
Field guidance for Kubernetes, machine identity, service boundaries, and modern application infrastructure.
Supporting context
Runtime context decides whether a control matters.
Why does the workload security topic matter to cloud security?
Runtime context decides whether a control matters. This topic applies that principle to practical control evidence and review decisions.
Which control questions belong to Workload security?
Research belongs here when its central evidence, failure mode, or operational decision falls within the workload security boundary described on this page.
How many workload security publications are available?
This page currently lists 38 publications assigned to Workload security.
How should readers use the Workload security research?
Start with the publication closest to the control, platform, or evidence gap under review, then follow its findings, limitations, numbered sources, related work, and downloadable material.
Can workload security overlap other research topics?
Yes. A publication can appear under several topics when one evidence path crosses multiple control boundaries; each publication page identifies its complete scope.
Supporting context