brianbolze $ARGUMENTS
- Review
MAINTAINING.md - Review engine-dev rules. Are we adhering to them?
- Consider running the
drift-sweepskill - Consider looking at this from the perspectives of the different personas
- Look at the impact on the directory / file structure. Does this add bloat?
- Does this new change suffer from overfitting? Was it designed based on only a small set of data / specific circumstance?
- Does this new change increase complexity / risk in a way that we haven't anticipated?
- If this new feature / change has an associated "FRAME" doc - read it - and ensure we're meeting it's intended goals.
- If this new feature / change has an associated "PROPOSAL" / "APPROACH" doc - read it - and see if it has drifted.
- If the review surfaced a process friction, a recurring miss, or a Truffle gap worth remembering, log it as an observation per
learning/AGENTS.md— capture only (never a fix); the review pass decides what to do with it.