feat(linkedin-studio): N12 — editions-register CLI + phaseLog-telemetri (TDD) [skip-docs]

Two things were structurally invisible: WHICH editions are in flight (readable
only by opening every series folder's edition-state.json by hand) and HOW LONG an
edition takes (not recorded anywhere). At two editions a week both are
load-bearing — a production line cannot be dimensioned on numbers never collected.

Register (scripts/editions, new module beside distillate.ts): one row per edition
in ${LINKEDIN_STUDIO_DATA}/editions/register.json — series, edition, title, series
path, current phase, next action, slot, startedAt/completedAt. Data-dir placement
(M0) because the register spans ALL series; the distillate is per-series and stays
in the series root. Verbs: register-upsert / register-list / register-complete.

phaseLog: articles.NN.phaseLog[{phase, completedAt}] in edition-state, additive so
schemaVersion stays 1 — pre-N12 editions load unchanged and their absence reads as
"not measured", never as zero. Per-article, mirroring articles.NN.phase: lead time
is a property of an edition.

One call per transition, both writes: newsletter.md gains a phase-transition
protocol defined once beside the resumption table, and all 16 canonical phases
invoke it. register-upsert appends the phase-log entry AND mirrors the register
row. Deliberately one command, not two — telemetry the command layer must remember
to write separately is incomplete inside a week, and an incomplete log measures
nothing. Step 10 closes the row and prints the measured lead time.

Mirror discipline: resumption still reads edition-state.json and only that. Delete
the register and the next transition rebuilds it; a failed upsert is reported, never
a reason to stop the pipeline. startedAt is the one unrecoverable value, so it never
moves — re-upserting a completed edition reactivates the same row (what
/linkedin:pivot does), keeping the clock on real elapsed production time.

Deterministic: no clock in the core (now passed in, --at/--now at the edge, as the
distillate takes lockedAt). Idempotent where a re-run is legitimate (repeated
transition logs once; completing twice keeps the first completedAt), not where it is
real work (a phase recurring after another one is logged again — a pivot back through
cleared gates is production time that happened). Missing facts are refused at the
edge, not defaulted: a row naming the wrong phase is worse than no row.

TDD (Iron Law): all three test files written first and verified red before any
implementation existed (register.test.ts + editionState.test.ts failed on missing
modules, cli-register.test.ts on "unknown command: register-upsert").

Suites: editions 27 -> 72 (floor raised) · test-runner 173 -> 184 (Section 16s: 11
unconditional greps incl. a 16-phase coverage sweep + non-vacuity self-test;
anti-erosion floor 155 -> 166) · trends 300/0 · brain 134/0 · specifics-bank 45/0 ·
hooks 140/0 · tests 35/0 · render 60/0. tsc --noEmit clean.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01QxvWAjte7vPcF79QeSRvRJ

CLAUDE.md

@@ -13,6 +13,7 @@ Full-spectrum LinkedIn content engine — short-form feed posts, carousels, vide
 - **Analytics:** CLI `scripts/analytics/` (TypeScript, needs `tsx` + `npm install`); data `assets/analytics/` (gitignored)
 - **Specifics-bank:** `scripts/specifics-bank/` (TypeScript, needs `tsx` + `npm install`) — deterministic, topic-tagged store of the operator's lived specifics (real numbers, named cases, held opinions) in the per-user data dir; elicited/bound by `/linkedin:newsletter` Step 1.5 so drafts draw from real inventory, never invented filler. Consumption is logged back at Step 8 lock (`record-usage` → `usedIn`), so cross-edition re-use of the same material is visible rather than silently encouraged by the dedupe
 - **Series distillate:** `scripts/editions/` (TypeScript, needs `tsx` + `npm install`) — series-level memory: each locked edition's spent narrative units (anecdotes/arguments/hooks) in `<serie>/linkedin/series-distillate.json`; written at `/linkedin:newsletter` Step 8 (`distil-append`), checked against the next skeleton at Step 2.5 (`distil-check`, advisory). Deterministic character-trigram similarity — AI extracts, code compares
+- **Editions register:** `scripts/editions/` (same package) — one row per edition in production in the per-user data dir (`editions/register.json`): series, edition, path, current phase, next action, slot, `startedAt`/`completedAt`. Each of the 16 canonical `/linkedin:newsletter` phase transitions runs `register-upsert --edition-state <path>`, which appends `articles.NN.phaseLog` (lead-time telemetry, additive — schemaVersion stays 1) **and** mirrors the row in one call; Step 10 runs `register-complete`. `register-list` is the work-in-progress view. The register is a **mirror** — deterministic resumption reads `edition-state.json` only, and a lost register is rebuilt by the next transition
 - **Contract-gate:** `scripts/contract-gate/` (TypeScript, needs `tsx` + `npm install`) — deterministic §B/§C1 rule-gate on the full draft (`/linkedin:newsletter` Step 4.5, before the AI sweeps); ratifies `rules.ts` against the edition's §E-manifest, then gates with BLOCK/WARN
 - **Analytics metrics (S16):** parsed CSV columns + an optional, manually-entered `saves` count (count-only in native LinkedIn analytics since ~Sept 2025, no CSV export; the Marketing API exposes `POST_SAVE` on `/memberCreatorPostAnalytics` from v202604, but access is partner-gated — so manual entry remains the right UX). `parseOptionalCount()`: blank / non-numeric / negative → `undefined` (`unknown`, never 0), a genuine `0` is kept; saves surfaced per-post + as `totalSaves`, but **not** folded into `engagementRate`. `dwell` stays **explicitly unmeasurable** (internal to LinkedIn, no export/API). All analytics I/O routes through the `getAnalyticsRoot()` seam (M0 per-user data-dir).