MCP Apps Ships July 28: What a Rollout Dashboard Instead of Raw JSON Means for Deploy-From-Chat
MCP Apps lets a tool return a live, sandboxed dashboard instead of a JSON blob the model has to paraphrase. Here's exactly what that changes for a deploy-from-chat platform's rollout and rollback tools.
Neon's 2026 Pricing Reset: What $0.106/CU-Hour Serverless Postgres Really Costs Against Self-Hosted CNPG on Hetzner
Neon's compute and storage rates have dropped repeatedly since the Databricks deal. A line-by-line recompute of what a branching-heavy Postgres workload actually costs on Neon's 2026 pricing versus CloudNativePG on Hetzner — including the labor cost of manual branch refreshes that most comparisons leave out.
Nomad vs Cluster API: Why a 1-3 Week Setup Win Still Loses for a Multi-Tenant PaaS
Nomad reaches production in 1-3 weeks against Kubernetes's 2-6 months — a real win for running your own app, but the wrong axis for a git-push PaaS whose product is hiding the orchestrator from tenants.
Northflank's Vibe-Coding Playbook Has Four Stages. bex Only Ships Two of Them by Default
Northflank's enterprise vibe-coding guidance reduces to four jobs a platform owes an AI-generated app: audit, harden, observe, sandbox. Here's which two bex ships as defaults today, and which two are still the tenant's job — checked against the actual bex.yml schema.
Porter's Two Control-Plane Models, One Codebase
Porter markets itself as a PaaS that runs in your own cloud, but the control plane most customers actually get is Porter's, not theirs. A tier-by-tier look at what real self-hosting costs versus what the marketing implies.
Heroku Invented the Preview Environment for Humans. In 2026, AI Agents Are the Ones Clicking the Link.
Devin Review catches bugs before a human opens the PR, and GitHub Copilot's agentic review closes the loop even faster. Heroku's 2015 preview-environment primitive was never built for that reader — here's what has to change.
Why Talos Linux's Own Creator Didn't Build Omni on Cluster API
Sidero Labs, the team behind Talos Linux, built its own Omni product without Cluster API. Here's a point-by-point read of its six objections against what a CAPH/CAPD-based fleet like bex actually looks like.
Deploying vLLM as Just Another Git-Push App: What GPU Scheduling Actually Adds
Placement mostly reduces to a manifest field and a node-pool label. Routing doesn't — a worked example against bex's actual Hetzner GPU inventory shows exactly where 'just another git-push app' breaks down for a model server.
12-Factor Agents Is the New 12-Factor App: A Deploy Tool's Field Guide
Dex Horthy's 12-Factor Agents mirrors Heroku's 2011 methodology almost line for line. Here's what each of the twelve factors concretely costs a platform building MCP-exposed deploy and rollback tools for AI agents.