eshlox

When AI-generated code breaks something, the real problem is often how we reviewed, understood, and trusted it.

AI turned changing my blog layout from a yearly event into six redesigns in one evening.

AI agents do almost all the work, while I control permissions, review every change, and decide what gets merged.

Every AI apology should refund the tokens it just used.

I let AI build a full side project in one run.

I write a Flutter app in a container on a Hetzner box and build it on my Mac.

I moved all my development off my MacBook and onto an always-on Hetzner VPS, reached over Tailscale.

Ask AI to expand a tiny side project and it may build an entire enterprise platform.

The Anthropic case reminded me that vendor lock-in is not just a technical problem.

I use AI every day, but I am not convinced that giving it access to every part of our lives is safe.

Why I wrote a small Bash wrapper around Lima to give each project its own Fedora VM, with VM-local SSH keys, an AI agent sandbox, and no host mounts by default.

A small zsh function on top of just that renames tmux windows, falls back to a global justfile, and shows project + global recipes in one fuzzy picker.

Subscribe to Google Calendar's secret iCal feed from any calendar app.

Developers went from distrusting everyone to handing AI agents access to everything.

A small LaunchAgent that runs hidutil at boot.

Mobile and desktop stores should support PWAs as first-class applications.

There are developers who sandbox everything and developers who will after their first incident.

Posting on multiple social networks is tiring.

When you host on a big cloud, you inherit every employee, every OAuth connector, and every co-tenant as part of your attack surface.

While developer tools move from terminals to GUIs and give AI access to everything, I am moving in the opposite direction.

The more software I install, the bigger the chance that I run something unsafe.

I discovered Polish folk music on YouTube that I genuinely enjoyed.

Using just recipes to rename tmux windows based on the running task, then restore the name when the command exits.

A Hammerspoon hotkey that saves clipboard images to the current tmux pane directory.

Claude Desktop adds MCP servers, Cowork, Computer Use, and Desktop Extensions.

A hardening checklist and decision matrix for secure local development with Infisical, Docker, and AI sandboxes.

Onboarding, offboarding, RBAC, CI/CD machine identities.

Practical day-to-day workflows for developing with Infisical, Docker, and AI sandboxes.

Comparing secrets managers for local development: HashiCorp Vault, OpenBao, Doppler, SOPS + age, and Infisical.

Keep Apple for personal, Linux for development.

Helix, lazygit, and other CLI tools can be sandboxed with nono or zerobox.

Rotate first, investigate second.

Set up infisical scan as a pre-commit hook.

Three ways to let Claude test endpoints without seeing secrets: exposed ports, shared Docker network, or an MCP proxy.

In the AI sandbox, compromised packages only see source code.

Every AI agent isolation option compared. sbx for maximum security, zerobox for ease of use, nono for irrevocable kernel restrictions.

Claude's trust levels, allowed actions, and permission popups are self-enforced and bypassable.

Expo cannot run in Docker because iOS Simulator needs direct Metro access.

Run your API and database in Docker Compose, inject secrets with Infisical, run AI agents in a separate sandbox with only source code.

Claude Code auto-loads .env files.

Install Infisical CLI, organize secrets by project and environment, inject them into processes without writing to disk.

1Password is great for personal passwords but bad for dev secrets.

AI coding agents read your .env files and send secrets to remote servers.

A quick guide to running Claude Code in a Docker sandbox using sbx.

A minimal Ghostty terminal config for macOS.

A single justfile for commands you need everywhere, not just in a project.

I use AI to help me write some of the blog posts.

A practical guide to protecting your JavaScript project from npm supply chain attacks using pnpm, Socket Firewall, and frozen lockfiles.

A minimal CSS framework that styles semantic HTML out of the box.

vibeship lets you describe what your code should do in plain English.