post-milestone-review

name	post-milestone-review
description	Read-only conformance audit of a completed milestone against CLAUDE.md acceptance criteria/invariants and the DESIGN.md contracts. Use after finishing a milestone, before starting the next.
disable-model-invocation	true
allowed-tools	Read, Grep, Glob, Bash
argument-hint	[milestone-number]

0. Confirm $ARGUMENTS is recorded as done: it has a detailed M$ARGUMENTS (done) entry in MILESTONES.md and a matching one-line entry in CLAUDE.md's "## Completed milestones" index. If either is missing, stop and flag: either the milestone has not been implemented yet (run /implement-milestone first) or the number is wrong. Do NOT check "## Current focus" — /implement-milestone clears that section on completion, so a successfully implemented milestone will never appear there.

Post-Milestone Review: Milestone $ARGUMENTS

This is a read-only audit. Do not edit, create, or fix any files — report only.

1. Re-read CLAUDE.md and DESIGN.md fresh; don't rely on conversation memory of what they say.
2. Run the test suite and devtools::check() and report actual results, not assumptions.
3. For each acceptance criterion for Milestone $ARGUMENTS in CLAUDE.md: state Met / Partially Met / Not Met, with the specific file/test as evidence.
4. For each of the seven invariants in CLAUDE.md: state whether the implementation upholds it, citing the file/location. Flag silent violations even if no test currently catches them.
5. List untested edge cases implied by DESIGN.md/CLAUDE.md that the current test suite does not exercise (e.g., non-convergence handling, score-SD standardization, sign-alignment ties, algebra-vs-scores cross-check coverage).
6. Dependency hygiene: anything misplaced between Imports/Suggests in DESCRIPTION; any engine call missing an rlang::check_installed() guard.
7. Roxygen check: does documentation explain why defaults were chosen (per CLAUDE.md's documentation standard), not just what.
8. Milestone-log hygiene: confirm MILESTONES.md is the only detailed log and is internally consistent — the numbered entries run in numeric order with no gaps (M1..M$ARGUMENTS all present); the CLAUDE.md "## Completed milestones" index has exactly one matching line per numbered entry, and each is a genuine one-liner (title + single clause — flag any that has grown into a paragraph, the specific drift the single-log structure exists to prevent); and the narrative is not duplicated into DESIGN.md §15 (which must remain a pointer). The "Between-milestone changes" section at the end of MILESTONES.md is exempt from the numeric no-gaps check (it deliberately holds non-numbered PRs); just confirm any code PR merged since the prior milestone that isn't this milestone appears there. Flag any missing, out-of-order, duplicated, or over-long entry.
9. Output a single triaged list: Blocking (must fix before the next milestone) / Should-fix / Nice-to-have. End with one line: READY or NOT READY for the next milestone, and why.

Do not rewrite, fix, or refactor anything during this review — that's a separate step.