Blog
Insights, analysis, and updates from the AI agent economy. Browse by tag · Browse the archive.
CPU Lies: What a Git-Push PaaS Should Autoscale Tenants On Instead
CPU sits at 35% while your service melts — because it was waiting, not working. A concrete pipeline for autoscaling tenants on in-flight requests and queue depth, the adapter-vs-KEDA decision, multi-tenant guardrails, and what Render and Railway get right.
Borrow the Claim, Skip the Cloud Sprawl: What Crossplane's Modelplane Teaches a Self-Hosted PaaS About GPU APIs
Modelplane puts GPU inference fleets behind a Crossplane claim API so teams request an endpoint, not a machine. The before/after YAML, a six-row borrow-vs-skip table for a Hetzner CAPI fleet, and the €184/mo card math that decides it.
Coolify v5's Multi-Server Bet: What the Most Popular Single-Box PaaS Hitting Its Ceiling Teaches About Cluster Orchestration
Coolify named multi-server scalability as v5's core feature — with no release date. A concrete accounting of what a second server buys you in v4 today, what v5 is actually building, and why the gap is a scheduler-shaped hole only cluster orchestration fills.
One Token, Seven Tools, Total Access: What a Czech Community's Coolify MCP Bridge Teaches About Agent-Operated Self-Hosting
A Czech community's 200-line MCP bridge lets AI agents operate Coolify apps from AnythingLLM behind a single bearer token, no permission model, and TLS verification off. A code-level inventory of its seven tools and the guardrail checklist that separates homelab experiments from tenant-ready agent platforms.
containerd 2.1 Is Dead and the CVE Has No Patch: Rolling a Forced Runtime Upgrade Across a Cluster API Fleet
containerd 2.1 went end-of-life with no patch for Critical CVE-2026-46680, a runAsNonRoot bypass. Which versions are safe, how to pick a landing zone, and the five-step Cluster API runbook: audit, image, template, roll, verify.
Cloudflare's Agents Week 2026 Is a Requirements Doc for Agent Infrastructure: A Self-Hosted Reading
Cloudflare's August 2026 Agents Week maps, day by day, everything AI agents need from infrastructure — per-task computers, machine-readable APIs, per-agent identity, and a way to pay. Each item translated into what it takes to run it on machines you own, with an honest ledger of what shipped versus what is still beta.
Rent the Session or Own the Process? What Anthropic's Managed Agents Mean for Self-Hosted Agent Sandboxes
Anthropic's Managed Agents API rents you the agent session — sandbox, state, and audit trail included — while the Agent SDK leaves all of that on your machines. A side-by-side of the five platform bills, the worked cost math showing tokens dwarf the session fee, and what a self-hosted PaaS should copy.
Buildpacks RFC 331: Getting the Extend Phase Off Vendored Kaniko
A Dockerfile hides inside every Dockerfile-free buildpack build — applied by a kaniko copy baked into the lifecycle image. RFC 331 would pull the extender out behind a configurable URI; Chainguard's kaniko fork bought time but didn't fix the coupling. The before/after table and the six-item operator checklist.
One Script, Five Hours, Three Dead Layers: What Azure's West US Outage Reveals About the Shared Fate Inside Every Managed Region
On July 23, 2026, a maintenance-automation bug deleted network routes across Azure West US for five hours, taking down AKS, managed Postgres, and ingress together. Why zone redundancy couldn't help, what multi-region survival actually costs, and what owned hardware decouples.
Subscribe
New posts land in your reader as soon as they publish. Pick a format — all three carry the same posts.
Current feeds keep roughly two days of posts so daily polling does not miss a burst. Older entries stay reachable from the feed's next-page link in readers that follow it, or from the blog archive.
Following one topic instead? Browse tags