https://playdeepkeep.com/blog Building a deterministic tactical roguelite almost entirely with an AI coding agent — the design decisions, the invariants, and the process. en Wed, 01 Jul 2026 09:00:00 GMT https://playdeepkeep.com/blog/shipping-in-small-steps https://playdeepkeep.com/blog/shipping-in-small-steps Wed, 01 Jul 2026 09:00:00 GMT Why building with an AI agent came out as ~145 small pull requests instead of a few big ones — the cadence, the phased features, and the replay discipline that made each step verifiable. The series wrap-up. https://playdeepkeep.com/blog/directing-the-agent https://playdeepkeep.com/blog/directing-the-agent Wed, 01 Jul 2026 09:00:00 GMT If the agent writes the code and the gates catch the disasters, what's left for the human? Setting intent, and reshaping vague or dangerous requests into the form that fits the project's rules — the highest-value work in AI-assisted development. https://playdeepkeep.com/blog/when-the-agent-was-wrong https://playdeepkeep.com/blog/when-the-agent-was-wrong Wed, 01 Jul 2026 09:00:00 GMT Three times the AI agent shipped a plausible, well-argued change that was quietly broken — an infinite loop, a difficulty-flattening item, an unkillable build — and the gate that caught each one. https://playdeepkeep.com/blog/trust-but-verify https://playdeepkeep.com/blog/trust-but-verify Wed, 01 Jul 2026 09:00:00 GMT An LLM will assert that its code is correct, balanced, and safe with total fluency and no grounds. The load-bearing practice of the whole project was converting every such claim into a check that passes or fails the build. https://playdeepkeep.com/blog/external-memory-across-sessions https://playdeepkeep.com/blog/external-memory-across-sessions Wed, 01 Jul 2026 09:00:00 GMT An AI agent starts every session with a blank slate. Two living documents — a backlog and a design log kept honest — became its external memory, so no session re-litigated a settled decision or re-built a finished feature. https://playdeepkeep.com/blog/the-constitution-claude-md https://playdeepkeep.com/blog/the-constitution-claude-md Wed, 01 Jul 2026 09:00:00 GMT How a short, blunt file of non-negotiable rules at the root of the repo kept an AI coding agent from drifting — the cheapest, highest-leverage practice in the whole project. https://playdeepkeep.com/blog/building-a-game-with-an-ai-agent https://playdeepkeep.com/blog/building-a-game-with-an-ai-agent Wed, 01 Jul 2026 09:00:00 GMT A dev-process retrospective: what it's really like to build a non-trivial game with an AI coding agent doing most of the typing, and the practices that kept ~145 pull requests coherent instead of collapsing into mush.