Locked-in design decisions for llm-tutor. Each entry pairs a what (the decision itself) with a why that cross-links back to the research catalog when the decision was empirically backed.
| Date | Topic | Backing research |
|---|---|---|
| 2026-06-03 | Statusline integration architecture — why the wrapper pattern, and why rendering logic lives in llm-tutor rather than in any statusline | (engineering call, no external research) |
| 2026-06-02 | /tutor-start and gamification — 5 decisions covering calibration depth, path surfacing format, state schema, feedback flow, and the rotating-question bank |
llm-tutor design foundations |
When a design call is locked in, drop a markdown file in this directory using a YYYY-MM-DD-topic-slug.md filename. Each decision should explain:
- Context — what was the situation that required a decision?
- Decision — what was chosen?
- Why — what evidence or reasoning backed it? If research informed it, link to
docs/research/<slug>.md. - What's still open — any sub-decisions deferred to future work.
The "why" cross-link is what gives the decision durability — six months from now, someone asking "why did we pick this?" should be able to follow the link to the evidence.