I Put Omarchy on an M1 MacBook Pro
There's a one-command path to Omarchy on an Apple Silicon Mac. Installing it is the easy part. Here's what happened after, days in and still dual-booting.
Writing, page 4 of 12
All 138 posts, newest first. The latest are on page 1, and the topic directory groups them by subject.
I Read Oban's Source Code So You Don't Have To
How Oban actually fetches jobs, notifies workers, and elects a leader, plus the VACUUM cost nobody mentions. Read from oban-bg/oban v2.23.1, not the README.
Building an LLM Audit Trail That Passes SOC 2
An append-only Ecto schema for LLM calls, the fields an auditor actually samples, and why provider request IDs are what makes an entry defensible.
Throttling LLM Calls in Elixir Before You Hit 429
LLM providers cap tokens per minute too, and you can't count a response's tokens before it lands. Size a Hammer bucket that never sends the call that 429s.
Agents Are Already Writing a Language You Can't Read
Compression toward machine-preferred notation isn't a future bet. It's already landing in context formats, agent protocols, and the files your harness reads.
The Agent Evidence Pack for a Security Review
The two documents an enterprise reviewer wants for your AI agent, the allow-list and the tool-call log, specified field by field and generated from code.
Wiring LiveView Uploads to Claude Vision
Boundary-validate uploads in allow_upload before they cost tokens, then wire consume_uploaded_entries into Claude vision and PDF calls from LiveView.
Most Security Controls Assume a Team You Don't Have
Security checklists only ever grow, because nobody who writes them pays the cost of a control a small team cannot staff. Which to defer, and which never wait.
Structured LLM Output in Elixir: Which Library, When
Four Elixir options for schema-validated LLM output, scored on maintenance, provider coverage, retry behavior, and escape hatches.
What Elixir Gives a Coding Harness for Free
José Valim's case for Elixir as a coding-agent harness holds up on the runtime. The tax is the ecosystem around it: sandboxing, plugins, and SDKs.
Metering LLM Usage Per Customer in Phoenix
ReqLLM emits token-usage telemetry; turning it into a per-customer invoice line needs an append-only ledger, integer-cent math, and idempotent writes.
Your Junior Dev Leans on the Agent. Good.
Telling a junior to use less AI is advice with an expiry date. Move the quality bar off style and onto verification. Here's what that looks like.