Monorepo vs Multi-Repo: Why AI Agents Tip the Scale
AI coding agents optimize for full-codebase context, atomic changes, and connected dependency graphs — properties that monorepo provides by default.
Read moreWriting on AI in engineering — agentic workflows, LLM tooling, context engineering, and the practical implications for software and infrastructure teams.
AI coding agents optimize for full-codebase context, atomic changes, and connected dependency graphs — properties that monorepo provides by default.
Read moreclaude-code-scheduler bridges Claude Code to launchd and crontab for persistent, observable, stateful task automation — no Desktop app required.
Read moreAGENTS.md files can hurt AI coding agent performance by over 20%. The fix isn't removing them — it's treating them as caches for ambiguity resolution.
Read moreAI agents with direct write access to production reintroduce the risks we eliminated with IaC — state drift, broken audit trails, and silent mutations.
Read more