A human factors reading of LLM token limits, and the plain-text vault I built so my AI tools stop forgetting what matters.
Sep 2026