Next.js 16.3 Broke Every Standalone Adapter Deploy — and Some Victims Have No Config-Level Escape
Next.js 16.3 crashes any build combining a deployment adapter with output standalone — on Vercel and self-hosted alike. The root cause, who has an escape hatch, and a pin-and-qualify playbook.
When One Builder Bug Breaks Every Modern Laravel Deploy: Lessons From Nixpacks' Duplicate Nginx Block
A January 2026 Nixpacks bug emitted duplicate nginx location blocks for every Laravel app shipping both composer.json and package.json — green build, dead runtime, restart loop. The incident anatomy, 2026's Railpack-led builder shakeout, and two defenses for self-hosted PaaS operators: per-app builder pinning and a build-matrix CI check.
Railway's Three-Builder Roulette in 2026: What Auto-Detect Build UX Gets Wrong for Monorepos
Railway's 2026 builder lineup — Railpack by default, Nixpacks in maintenance, Dockerfile autodetect — plus root-relative config paths produces five concrete failure modes for monorepos. The detection-order table, the failure catalog with real-repo evidence, and the explicit-pinning checklist that stops the spin.
A Real Hetzner Decision Doc Picked Coolify v4 and Dokploy: What Teams Actually Reward in a Self-Hosted PaaS
A real team's Hetzner platform decision scored Coolify, Dokploy, CapRover, Kamal, Dokku, and Portainer on low ops burden — and picked its winners for operability, not feature count. Here is the reconstructed matrix and the machine-two question it deferred.
The Forge That Sketched Its Own PaaS, Then Cut It: What Cave's Deferred Podman Deploy Tells Teams Rebuilding Git-Push Without a Vendor
Cave's design PRD made podman deploy a launch pillar, then the author cut all container deploy functionality indefinitely. The spec, the cut, and a row-by-row accounting of what a forge answers versus what still needs a platform.
StackQL v0.11 Turns Every Agent Cloud Query Into an OpenTelemetry Record: What That Buys Your Fleet's Audit Trail
StackQL v0.11 speaks MCP revision 2026-07-28 and emits its agent audit log as OTLP records. The exact collector config that ingests it with no transform, the stateless-HTTP tradeoff, and what remains before inventory queries become deploy authority.
Versioning MCP Deploy Tools at 10,000-Server Scale: What Breaking Tool-Schema Changes Do to Agents Pinned Against Your Endpoint
Past 10,000 servers, your MCP tool schemas are a public API with pinned agent clients. A breaking-vs-safe change table, two worked deploy-tool examples — one loud, one silent — and a six-rule versioning discipline for self-hosted endpoints.
Your Deploy-from-Chat MCP Server Has Too Many Tools: What Anthropic's 98.7% Token Cut Means for Agent-Operated Infrastructure
Anthropic cut a multi-step agent task from 150,000 tokens to 2,000 by having agents write code against MCP tools instead of calling them one by one. Here is what that 98.7% saving means for deploy-from-chat servers, and why every mutation still needs its own approval and audit entry.
One Gateway or Fifty Servers? What AWS and Google's Opposite MCP Bets Mean for Your Deploy Surface
AWS ships one MCP server with a fixed tool set while Google ships more than fifty, one per service. This post compares both designs on context cost, auth, versioning, and audit, then recommends a concrete tool vocabulary for a deploy surface agents can operate.