You do not need to Dockerize your entire app to run multiple agent workspaces
Why coding-agent workspaces can keep application processes local while using Docker only for shared infrastructure such as databases, queues, caches, and object storage.
Posts
I share my thoughts on some software engineering topics.
Introducing Agent-Up: run multiple agent-built versions of your web app at once
Agent-Up is an open-source desktop app that gives each AI coding agent its own isolated runtime environment — separate processes, ports, Docker services, and browser session — so you can review multiple agent-built versions of a web app simultaneously.
AI agents changed my opinion on vertical slices.
A post about why vertical slices make feature ownership, reviews, tests, and coordination clearer when working with multiple AI agents.
One AI agent is a chatbot. Three AI agents start looking like a small development team.
A post about why running multiple AI coding agents needs visibility, isolated working directories, branches, pull requests, and code review.
I built xcepto-skill so AI agents can write better Xcepto tests
Why Xcepto needed a dedicated Codex skill before AI agents could write readable, maintainable, behavior-focused tests.
How do you safely let other developers deploy code on your infrastructure?
A HiveShard deployment pipeline post about compile-time restrictions, SDK enforcement, and trusted build artifacts.