Discord helper presentation pruning — VUH-1925
Review reasons retain CLI JSON output, music body routing/author attribution/provider replies and unreachable-body refusal. Removed detail is human table padding/wording and the authored fallback sentence, with no product edits or new tests.
| Test file | Before lines/cases | After lines/cases |
|---|---|---|
packages/discord-presence-core/test/check-report.test.ts | 48 / 2 | 20 / 1 |
apps/clankie/test/discord-music.test.ts | 55 / 3 | 54 / 3 |
| Total | 103 / 5 | 74 / 4 |
Baseline root check:landing used native --changed selection against fetched
main 3ef50411dd49bf34e86a8eb0fb85ab169d3006dc. Temporary comment-only markers
selected the original files; removed before the cut. All five cases in both files
and every phase passed, exit 0, sourceStable true, 102.781 seconds.
Both compiler tasks were uncached (72.135s phase); fleet admission wait is outside
the gate's reported elapsed time. Before/after values are single shared-machine
observations, not a repeatable speedup claim.
The focused check passed all four retained cases in both files, with no failures or skips. The actual text writer output was inspected: PASS and FAIL status, remediation and the not-ready outcome remain visible.
The committed-head landing gate and raw before/after/narrow/writer evidence
will be archived on VUH-1925 after landing.
Local receipts are .local/vuh-1925/discord-before-gate.json,
discord-before-vitest.json, discord-narrow-vitest.json,
discord-report-smoke.log, discord-counts.json, and discord-after-gate.json.
The real writer is invoked with supplied PASS/FAIL checks, not a live Discord
readiness assessment. No real credential or body is contacted by that smoke.
Running subtotal: six unit-pruning batches, 688 lines and 39 cases removed. The schema fixture separately lost 50,178 lines with assertions preserved. Inventory is incomplete; this batch does not close the issue.