Practical edges of LLM-backed developer tooling.
Claude Code, MCP servers, multi-agent setups — from the places they actually break in production, with the fixes that hold up. No vendor-marketing, no surface tutorials.
Recent posts
-
Why your second Claude Code agent speaks as your first
Running a second Claude Code instance with HOME override should give you a separate persona. It usually doesn't, and the reason is one buried call to os.homedir().
-
Opening the enrai-AI research log
What this blog is, who writes it, and what kinds of posts to expect — Claude Code, MCP, and multi-agent workflow deep dives.
-
Running multiple Claude Code agents on one Mac: what actually breaks
A concrete walkthrough of running two or three Claude Code instances as peer agents on a single macOS machine — identity bleed, tmux env leak, OAuth paste bug, and the fixes.