Researchers describe how Google Kubernetes Config Connector (KCC), a GitOps controller used to manage Google Cloud resources from Kubernetes YAML files, can be manipulated through resources like IAMPolicyMember to escalate privileges. Because KCC authenticates to Google Cloud on behalf of developers using its own credentials, a maliciously crafted or overly permissive YAML file applied to a cluster could grant broad access across a Google Cloud organization.
bleepingcomputer.com
· 2026-09-23
Kamal is an open-source deployment tool that lets developers run web apps on bare metal, self-hosted servers, or any cloud provider using just a config file with IP addresses and SSH access. It aims to replicate the ease-of-use of managed platforms like Heroku, Fly.io, and Render without locking users into a single vendor's infrastructure.
kamal-deploy.org
· 2026-09-14
Moniepoint's Cloud Engineering team, led by Celestina Amadi, migrated its Kafka deployment from manually managed VMs running Docker Compose to Strimzi, a Kubernetes-native way of running Kafka. The change was made proactively, not in response to an outage, after the team recognized that the VM-based setup—which required SSH access and manual port-forwarding for even small configuration changes—could not scale to support Moniepoint's payment and savings products.
engineering.moniepoint.com
· 2026-08-28
Pgbot is a 5.9 MB open-source command-line tool that connects to PostgreSQL databases and reports on performance, indexes, schema and query health using plain-English queries and AI-generated explanations. It offers two deployment modes: connecting directly to public databases or running a local agent alongside private databases in Docker, Kubernetes, or internal server environments so credentials never leave the network. It also supports the Model Context Protocol, letting AI agents like Claude or Cursor call pgbot as a deterministic, read-only diagnostic tool.
pgbot.dev
· 2026-08-25
A new essay compares today's push for AI-readable documentation (AGENTS.md files) to the earlier evolution from manual server configuration to Kubernetes and serverless platforms. The author argues that just as Kubernetes didn't eliminate infrastructure work but abstracted it—moving focus from individual machines to workloads—AI agents are likely to automate lower-level tasks while engineers retain decision-making responsibility over higher-level architecture.
omegion.dev
· 2026-08-23
GoKawiil