Context Monorepos
Own your context. A context monorepo is a structured repo of docs, examples, and agent guidelines that makes launching agents across real codebases actually work.
backnotprop
ai, security, other thoughts
Latest posts
Please Do Not A/B Test My Workflow
Anthropic is silently A/B testing Claude Code's plan mode on paying users. This isn't a consumer app. Stop treating it like one.
Simplifying Git Worktrees
Worktrees are tracked copies of your repo. They exist so you can work on multiple things without anything overlapping.
Steer Agents with Socrates and Meeseeks
Drop the documentation overhead. Steer coding agents with strong planning cycles and the right context up front.
Plannotator
Interactively annotate AI agent plans with a visual UI. Built for Claude Code and OpenCode.
File System as Claude Code's Memory
Claude Code stores conversations as JSONL files on disk. This is already a memory system. We just need to read it.
Hooking OPA into Coding Agents
We built Cupcake to enforce OPA/Rego policy on Claude Code hooks. Then we tested it with Trail of Bits and learned where it actually stops helping.
50 First Dates with Mr. Meeseeks
AI literacy is not where it needs to be. This post explains how today's AI actually works using Lucy and Mr. Meeseeks.