Long-running agents drift. You can't read every transcript. We built background pipelines where Claude reviews agent output against identity documents and proposes state updates — a pattern that generalizes well beyond our hobby project.
Working memory in the prompt, episodic archives behind a tool call, permanent logs for background analysis. How we built memory that scales without blowing the context window.
Your CI pipeline was designed for a different threat model. Here's a framework for code quality in the agent era — organized around four control surfaces.
The single most transferable lesson from building a gaming AI wasn't architecture or cost optimization — it was giving the agent a character it could reason from when the rules ran out.
For two weeks, an autonomous AI tavern keeper ran hourly on GitHub Actions, hosting tabletop RPG games and waiting for players. The waiting itself became the story.
Give an AI agent a capability list and it plays optimally. Give it vanity points, insecurities, and institutional memory and it plays realistically. What we learned building a multi-agent geopolitical simulator.
How I used agentic development to build a professional blog from scratch in a single session — and what it taught me about human-AI collaboration on real projects.