feat(linkedin-studio): S14 — journey layer (create/measure front-doors + 5-journey router), v4.1.0

14a's cold command-rationalization found ZERO redundancy across the 27 commands
(no defensible merge/cut), so the operator reframed S14 from "merge/cut" to
"add a journey layer over the kept atomics".

- Add /linkedin:create + /linkedin:measure — delegate-only guided front-doors
  (Read/Glob/AskUserQuestion only; route to the command that owns the work)
- Re-tier commands/linkedin.md into 5 journeys (Start/Create/Engage/Measure/Grow);
  onboarding/strategy elevated as Start/Grow front-doors; Engage = calendar+firsthour tier
- 14a honesty nits: router now lists firsthour; calendar cross-links to firsthour;
  competitive confirmed UNGATED (the claimed 1K-gating inconsistency was unfounded)
- Lockstep: EXPECT_COMMANDS 27->29, v4.0.0->4.1.0 across plugin.json / README badges /
  plugin+root CLAUDE.md / README / CHANGELOG; new README commands-badge lint guard
- 14a deliverable corrected: multiplatform entry added (26/27 -> 27/27), header counts,
  competitive nit withdrawn
- Independent /trekreview (2 Opus reviewers) raised 3 MAJORs (stale commands badge,
  router competitive self-contradiction, STATE.md count) — ALL remediated in-session;
  verdict ALLOW

Gate: test-runner.sh 74/0/0; node --test 98/98; commands=29; v4.1.0 consistent.
Additive/minor — no command removed/renamed/behavior-changed; reload registers create+measure.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>

plugins/linkedin-studio/docs/remediation/command-rationalization.md

@@ -0,0 +1,375 @@
+# Command-Surface Rationalization — Step 14a
+
+_Remediation Voyage, command-surface pass. Independent, COLD read of all 27
+command files in `commands/*.md` (full bodies, not just frontmatter), with the
+router (`commands/linkedin.md`), `CLAUDE.md`, and `README.md` as the map of the
+intended surface. Written 2026-05-30. ANALYSIS ONLY — no command file was edited,
+merged, deleted, or renamed; the single deliverable is this document._
+
+> **Correction (post-review, 2026-05-30).** The first pass omitted `multiplatform`
+> (covered 26/27) and mis-stated two group-header counts; it also raised a
+> `competitive` 1K-gating inconsistency that did **not** survive verification. All
+> three are fixed below: `multiplatform` now has its own entry (Group B), the Group
+> C/F headers read 3/5, and the `competitive` nit is withdrawn (CLAUDE.md `:64`,
+> README `:222`, the router, and the command body all leave it ungated). Net finding
+> is unchanged — **keep 27, 0 merge, 0 cut** — now genuinely covering all 27.
+
+## Method note
+
+The fixtures here are **the command files themselves** — each command's purpose,
+its full workflow body, its frontmatter `description` (the trigger-phrase set that
+governs how it gets invoked), and the sibling surfaces it touches. The discipline
+is borrowed verbatim from `docs/remediation/overlap-measurement.md` (whose SUBJECT
+is the 7 long-form review *agents*, not the commands — kept strictly separate
+here): for every overlap I apply the **subsumption test** — _does either command's
+catch-set / surface fully contain the other's?_ — and the trim rule: **merge or
+cut ONLY where a command catches nothing a sibling doesn't; if the redundancy is
+justified, record it and KEEP; if I can't decide, record "inconclusive — retain
+pending operator view" and do NOT trim.** Default-to-keep under uncertainty. This
+plugin already over-grew once (24→27), so the bias is honest-surface, not
+consolidation-for-its-own-sake and not growth-for-its-own-sake. Every
+recommendation is grounded in a citable trigger set, workflow step, or sibling
+surface — no taste-based cuts.
+
+A leverage note used throughout: a command's **invocation leverage** is the
+product of (a) trigger-phrase breadth in its `description` and (b) how likely a
+user is to actually type the intent when a *broader* command already catches it.
+A command whose entire trigger set is a strict subset of a broader command's, and
+which adds no unique workflow, is the classic merge/cut candidate.
+
+---
+
+## Group A — Entry / router (1)
+
+### `linkedin` (router)
+
+- **Purpose** — Status line (weekly progress, streak, follower phase) + upcoming/overdue queue + a menu/router that disambiguates to the right subcommand.
+- **Overlap with siblings** — Structurally overlaps *every* command (it lists them), but that is its job, not redundancy. No subsumption either way: the router holds no content-production surface, the leaf commands hold no menu/status surface.
+- **Invocation leverage** — Highest-traffic entry point: triggers on the bare `"linkedin"`, `"/linkedin"`, `"linkedin help"`. This is the front door.
+- **Recommendation** — **keep.**
+- **Rationale** — Structural and the single highest-leverage surface. Judged on grouping honesty: the grouping is broadly honest (Getting Started / Content Creation / Strategy / Post-Publish / Growth+Monetization), and it correctly soft-gates the 1K commands. One honest-surface nit for the operator (not a cut): it lists `post-feedback-monitor` as an *agent* under "Post-Publish Monitoring" but does not list the new `/linkedin:firsthour` command in that section — the router's own map is one release behind the surface it routes to.
+
+---
+
+## Group B — Short-form creation (7)
+
+### `post`
+
+- **Purpose** — Full interactive post creation (angle → format → draft → quality/de-AI gate → refine), targeting 1,200–1,800-char substantial posts.
+- **Overlap with siblings** — `quick` (short 150–500-char path), `react` (URL-sourced), `pipeline` (post + schedule + monitor). Subsumption test: `post` does NOT subsume `quick` (different length band + 3-line formula + 8 templates) and `quick` does not subsume `post` (no refinement cycle, no `content-optimizer` delegation). `post` is the body that `pipeline` wraps but `pipeline` adds scheduling/queue/first-hour that `post` lacks; neither subsumes the other.
+- **Invocation leverage** — Very high; broad trigger set ("create linkedin post", "write a post", "turn this into a post"). Core entry point alongside `quick`.
+- **Recommendation** — **keep.**
+- **Rationale** — Distinct length/depth tier with a unique refinement+optimizer surface; one of the two highest-traffic creation commands.
+
+### `quick`
+
+- **Purpose** — 5-minute 3-line-formula post (150–500 chars) AND the home of the 8 post-type templates (reaction/tip/observation/hot-take/failure/question/curation/one-liner).
+- **Overlap with siblings** — `post` (longer tier), `first-post` (which explicitly reuses `quick`'s 3-line formula). Subsumption test: `quick` uniquely owns the template library and the short-length band; `post` owns the long band. The two are explicitly cross-referenced ("When to Upgrade → `/linkedin:post`"). No subsumption.
+- **Invocation leverage** — High; broad trigger set including the absorbed `templates` intent ("post template", "give me a template", "fill in the blank post"). This absorption was the v2.0.0 consolidation — confirmed intact in the body (Step 1 hosts the template library).
+- **Recommendation** — **keep.**
+- **Rationale** — Holds the unique short-form length tier + the entire template surface that `templates` was folded into. Cutting it would orphan the templates.
+
+### `react`
+
+- **Purpose** — URL-to-post pipeline: fetch external content (article/news/research/YouTube), extract, pick a reaction angle, draft in voice; includes a multi-URL comparison path.
+- **Overlap with siblings** — `post` (could in principle ingest a URL via WebFetch — `post` Step 1 even mentions "If they provide a URL, use WebFetch"). Subsumption test: `react` has a unique surface `post` lacks — the multi-source comparison/synthesis path (Steps 1b–8b), the content-type→angle table tuned to *reactions*, and the "react, don't summarize / your take is the hook" discipline. `post` does not subsume that. `react` does not subsume `post` (no general-topic drafting).
+- **Invocation leverage** — High and unambiguous: "react to this", "this article", "this url", "share this news". The router routes any bare URL here. A user with a link will not naturally type `/linkedin:post`.
+- **Recommendation** — **keep.**
+- **Rationale** — Genuinely unique URL-ingestion + multi-source-synthesis surface and a distinct, frequently-typed intent.
+
+### `pipeline`
+
+- **Purpose** — End-to-end lifecycle orchestrator: ideation → draft → optimize → **schedule/queue** → 5x5x5 pre-engagement → publish → first-hour → 48h analysis.
+- **Overlap with siblings** — This is the densest overlap node. It re-does `post`'s drafting (Step 2), `calendar`'s queueing (Step 4 Option 3 calls the same `queue-manager.mjs` `queueAdd`), the 5x5x5/first-hour plan (now also `firsthour`'s job), and points to `analyze` at the end. Subsumption test: no *single* sibling subsumes `pipeline` — it is the only command that strings the whole chain into one guided run with explicit step gates. But `pipeline` is itself **a thinner re-statement of `post` + `calendar` (+ `firsthour`)**: every individual capability it offers exists, more developed, in those leaf commands (its draft step is a condensed `post`; its schedule step is a condensed `calendar`; its first-hour step is a condensed `firsthour`).
+- **Invocation leverage** — Moderate-to-low as a *natural* type. Trigger set ("pipeline", "full workflow", "end to end", "idea to post") is narrow and jargon-y; in practice a user wanting to create+schedule will type `/linkedin:post` then `/linkedin:calendar`. It reads as a workflow-orchestrator showcase more than a daily entry point.
+- **Recommendation** — **keep** (with a flag) — leaning **inconclusive**.
+- **Rationale** — No clean subsumption (it uniquely *sequences* the chain), so the trim rule says retain. But it is the surface most worth the operator's scrutiny: its value is the *orchestration*, and that value erodes as `post`/`calendar`/`firsthour` each grow richer than `pipeline`'s inlined condensations. Flagged in the honest-uncertainty list. If the operator ever wants to shrink the surface, this is the first place a defensible merge could be argued (→ `post`, preserving the explicit schedule+first-hour hand-off chain) — but the evidence does not *compel* it.
+
+### `carousel`
+
+- **Purpose** — Structured slide-by-slide carousel/document generator with template selection, per-slide copy, caption, de-AI gate, and optional mcp-image slide generation.
+- **Overlap with siblings** — `multiplatform` (has a "LinkedIn → Presentation Slides" template) and `post` (notes "could also work as a carousel — run `/linkedin:carousel`"). Subsumption test: `multiplatform`'s slide template is a 10-line generic outline; `carousel` is a full format engine (5 templates, per-slide char rules, mcp-image generation, carousel quality checklist). `multiplatform` does not remotely subsume it. `carousel` owns the highest-engagement organic format end-to-end.
+- **Invocation leverage** — High and specific: "carousel", "slide deck", "pdf post", "swipe post", "document post". A distinct format intent users type directly.
+- **Recommendation** — **keep.**
+- **Rationale** — Deep, unique format surface with image generation; not subsumed by anything.
+
+### `video`
+
+- **Purpose** — Video script generator (talking-head/screen-rec/slideshow, 30s–2min) with pacing/visual/energy cues, captions, thumbnail, first comment, a video quality gate, and delegation to `video-scripter`.
+- **Overlap with siblings** — `multiplatform` ("LinkedIn → YouTube Script" template). Subsumption test: `multiplatform`'s YouTube template is a generic timing outline; `video` is a full LinkedIn-native scripting engine (word-budget math, muted-autoplay test, completeness gate, `video-scripter` agent). No subsumption — and they target different platforms (LinkedIn-native vs. YouTube adaptation).
+- **Invocation leverage** — High and specific: "video script", "talking head script", "record a video". Direct format intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique LinkedIn-native video surface + dedicated agent; the `multiplatform` overlap is a thin adaptation template, not a substitute.
+
+### `multiplatform`
+
+- **Purpose** — Adapt existing LinkedIn content for other platforms: a Twitter/X thread, a generic presentation deck, or a YouTube script. Long-form is explicitly routed out to `/linkedin:newsletter`.
+- **Overlap with siblings** — `carousel` (its "Presentation Slides" template) and `video` (its "YouTube Script" template). Subsumption test: `multiplatform`'s slide/YouTube blocks are ~10-line generic outlines, while `carousel` is a full LinkedIn-native carousel engine and `video` a full LinkedIn-native scripting engine — neither subsumes the other, and they target different outputs (LinkedIn-native vs. Twitter / YouTube / generic deck). Critically, the **Twitter/X-thread** adaptation path exists in **no other command**. No subsumption.
+- **Invocation leverage** — Moderate; "adapt for twitter", "cross-post", "repurpose for", "turn into thread". A distinct cross-platform intent, but lower-traffic than the native creation commands.
+- **Recommendation** — **keep** (flag: **develop-candidate**).
+- **Rationale** — Unique cross-platform / Twitter surface; subsumption fails as a cut. But it is the **thinnest** command in the set — the only one that inlines static templates instead of delegating, where the `content-repurposer` agent already exists to power a richer adaptation. Keep now; a future *develop* (wire `content-repurposer`, deepen the per-platform output) is the defensible improvement, not a merge/cut.
+
+---
+
+## Group C — Long-form (3)
+
+### `newsletter`
+
+- **Purpose** — The single long-form orchestrator: a fixed 16-phase, multi-session pipeline (research → skeleton/spine gates → draft → fact-check → editorial → persona sweep → headless review → visual assets → lock → hook gate → schedule) with maintained edition-state.
+- **Overlap with siblings** — `headless-review` and `pivot` are *phases/companions* of this command (Step 6.5 and the lock-precondition re-open). Subsumption test: `newsletter` invokes the headless package inline (Step 6.5) and runs the pivot heuristic as a lock precondition (Step 8) — so it functionally *contains* both. But containment ≠ the companions are redundant (see their entries: the value is fresh-session isolation and a named re-open ritual, which `newsletter` cannot itself provide). No short-form command overlaps it; `multiplatform` explicitly routes all long-form here.
+- **Invocation leverage** — High for its niche; the sole entry for "newsletter", "long-form", "essay", "series article".
+- **Recommendation** — **keep.**
+- **Rationale** — The entire long-form spine; nothing else does this and everything long-form routes to it.
+
+### `headless-review`
+
+- **Purpose** — Cold/adversarial review package run on a FROZEN draft with starved context — `content-reviewer` + `language-reviewer` + `fact-reviewer` + `persona-reviewer` (resonance/conversion), consolidated into one operator-gated report. Standalone surface for `newsletter` Step 6.5.
+- **Overlap with siblings** — `newsletter` (Step 6.5 fans the same package inline). Subsumption test: this is the calibration case (c) the brief flags. Does `newsletter` Step 6.5 subsume `headless-review`? **No** — and the command's own body states *why*: the cardinal value is **Layer 1 fresh-session isolation** ("the parent itself then has no drafting transcript"). When the package runs inline inside the drafting `newsletter` session it carries exactly the framing-bias the package exists to eliminate. The standalone command is the *only* way to get a genuinely cold parent context. So the standalone surface has a unique catch (true independence) the inline phase structurally cannot.
+- **Invocation leverage** — Moderate; specific trigger set ("headless review", "cold review", "adversarial review", "review the frozen draft"). Niche but real, and the recommended path is to type it *in a fresh session* — which is precisely an action a phase cannot perform.
+- **Recommendation** — **keep** — justified standalone, not merely-as-a-phase.
+- **Rationale** — Subsumption fails in the independence dimension: the fresh-session parent is a capability the `newsletter` phase cannot replicate. This is the "redundancy is justified — record and keep" case, mirroring the agent-overlap finding for the cold trio in `overlap-measurement.md`.
+
+### `pivot`
+
+- **Purpose** — A named ritual to re-open an already-cleared long-form edition after a substantive late change: logs `pivots[]`, resets `currentPhase`, un-locks, invalidates downstream verdicts, marks which gates must re-pass. Includes the >20%/>2-section heuristic.
+- **Overlap with siblings** — `newsletter` (which runs the same heuristic as a Step 8 lock precondition and owns the gate re-runs). Subsumption test (calibration case c): does `newsletter` subsume `pivot`? **Partially but not fully.** `newsletter` Step 8 *detects* drift and *stops* the lock, but it explicitly **points the operator to `/linkedin:pivot`** to perform the state surgery (log the pivot entry, reset phase, un-lock, invalidate verdicts). `pivot` does NOT run the gates ("Do not run the gates yourself — `/linkedin:newsletter` owns the pipeline"); `newsletter` does NOT perform the re-open bookkeeping. The two are complementary halves of one ritual — clean separation, no subsumption.
+- **Invocation leverage** — Low-to-moderate; narrow trigger set ("pivot", "re-open edition", "added a section", "changed the angle"). Rarely hit, but when hit it does deterministic state work no other command does.
+- **Recommendation** — **keep.**
+- **Rationale** — Holds a unique state-mutation surface (the re-open bookkeeping) that `newsletter` deliberately delegates out; not subsumed. Lower-traffic, but the trim rule retains a non-redundant specialist.
+
+---
+
+## Group D — Onboarding / setup (3)
+
+### `onboarding`
+
+- **Purpose** — Multi-step wizard chaining profile → personalization → first-post as one guided flow for brand-new installs, with an "already onboarded" short-circuit.
+- **Overlap with siblings** — `setup` (Phase 2 is a condensed `setup`) and `first-post`/`profile` (Phases 1 and 3 route to them). Subsumption test (calibration case f): does `setup` + `first-post` subsume `onboarding`? **No.** `onboarding` uniquely provides the *cohesive single path* ("a single guided path instead of navigating 27 commands") and the cross-phase state logic (already-onboarded detection, score-gated branching, "what's next — your first week"). Notably it does NOT invoke the sub-commands directly — it tells the user to run them — so it is an orchestration/triage layer, not a duplicate of their bodies.
+- **Invocation leverage** — Moderate; trigger set heavily overlaps `first-post` and `setup` ("get started", "just installed", "walk me through"). The router lists it first. Real risk: a "get started" user could land on `first-post` instead — but the intent ("walk me through *everything*") is distinct from "help me publish my first post".
+- **Recommendation** — **keep.**
+- **Rationale** — Unique cohesive-wizard surface + cross-phase triage state; it orchestrates rather than duplicates `setup`/`first-post`. No subsumption.
+
+### `setup`
+
+- **Purpose** — Guided personalization: computes the 8-category personalization score and runs 6 sub-workflows (voice samples → `voice-trainer`, case study, framework, post analysis, demographics, user profile) to populate asset templates.
+- **Overlap with siblings** — `onboarding` Phase 2 (a 2-category condensation of this). Subsumption test: `onboarding` only ever touches the top-2 weighted categories and explicitly defers ("I'll run `/linkedin:setup` for the full setup"); `setup` owns all 8 categories, the full score dashboard, and the `voice-trainer` delegation. `onboarding` does not subsume it.
+- **Invocation leverage** — High; broad trigger set ("setup", "personalize", "personalization score", "configure plugin", "my score", "fill in assets"). Also the router's destination for any score/asset-completeness query.
+- **Recommendation** — **keep.**
+- **Rationale** — The canonical, full-depth personalization surface; `onboarding`'s version is a deliberate condensation that defers back here.
+
+### `first-post`
+
+- **Purpose** — First-post accelerator: maximum hand-holding, voice quick-check, simple topic pick, 3-line draft (reuses `quick`'s formula), 4-item quality check, sets `first_post_date`, includes a guard that redirects returning users to `post`/`quick`.
+- **Overlap with siblings** — `quick` (whose 3-line formula it reuses) and `onboarding` Phase 3 (which routes here). Subsumption test: does `quick` subsume `first-post`? **No.** `first-post` adds first-timer-specific scaffolding `quick` lacks: the welcome/expectation framing, the voice-from-scratch setup branch (samples vs. 5 questions), the "exists > perfect" philosophy, the new-creator 90-day-boost messaging, and the returning-user guard. Its whole value is the zero-to-one onboarding hand-holding, not the drafting mechanics.
+- **Invocation leverage** — Moderate; "first post", "never posted", "new to linkedin". A genuinely distinct beginner intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique first-timer scaffolding + the `first_post_date` side-effect; not subsumed by `quick`'s formula it borrows.
+
+---
+
+## Group E — Scheduling / publishing / post-publish (3)
+
+### `batch`
+
+- **Purpose** — Create a full week (3–5 posts) in one session from a theme/pillar: trend-spotter angles, `content-planner` plan, scheduling, draft files, queue entries, and a generated `.ics`.
+- **Overlap with siblings** — `pipeline` (single post + schedule), `calendar` (queue management). Subsumption test: `batch` uniquely produces a *multi-post balanced week* with format/pillar rotation and `.ics` export; `pipeline` is single-post; `calendar` only manages an existing queue. No subsumption.
+- **Invocation leverage** — High and specific: "batch content", "week of posts", "sunday prep". Distinct planning intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique multi-post-week surface + iCal generation; the queue-write overlap with `calendar`/`pipeline` is shared *plumbing* (`queue-manager.mjs`), not a shared *surface*.
+
+### `calendar`
+
+- **Purpose** — View/manage the 14-day queue (reschedule/cancel/view-draft) AND host the **publish action** (mark-as-published → update queue+state → first-hour plan). Absorbed the former `publish` command (v2.0.0).
+- **Overlap with siblings** — `firsthour` (its publish flow ends with a first-hour plan; `firsthour` is a richer standalone version), `pipeline`/`batch` (which write to the queue it reads). Subsumption test (calibration case d): does `calendar`'s publish action subsume `firsthour`? **No** — `calendar`'s first-hour block is a static 6-line checklist + an optional `post-feedback-monitor` hand-off; `firsthour` builds a *worked, timestamped plan with named targets and draft comments* via `engagement-coach`. Conversely `firsthour` does not manage the queue. No subsumption.
+- **Invocation leverage** — Very high; very broad trigger set spanning both calendar ("schedule", "queue", "upcoming") and publish ("mark as published", "just published", "post is live"). High-traffic.
+- **Recommendation** — **keep.**
+- **Rationale** — Canonical queue+publish surface that already absorbed `publish`; its first-hour checklist is a lightweight in-flow nudge, distinct from `firsthour`'s full sprint planner.
+
+### `firsthour`
+
+- **Purpose** — Post-publish first-hour / reply-loop sprint: delegates to `engagement-coach` for a timestamped target list (whales/inner-circle/ICP), draft self-comments + CEA replies in voice, a minute-by-minute timeline; persists via `recordFirstHourPlan`; hands off to `post-feedback-monitor`.
+- **Overlap with siblings** — `calendar`'s publish-action first-hour block, and the `engagement-coach` / `post-feedback-monitor` agents it delegates to. Subsumption test (calibration case d): `firsthour` is the orphan-wiring command for `engagement-coach` (its raison d'être per CLAUDE.md/README). It uniquely produces a *worked* plan (named targets + draft comments + clipboard copy + state persistence) that neither `calendar` (static checklist) nor `post-feedback-monitor` (48h monitoring, different time window) provides. The three form a relay (firsthour → coach for the plan → feedback-monitor for the marathon), not a redundancy.
+- **Invocation leverage** — Moderate; "first hour", "I just posted", "reply loop", "work my post". A real, frequently-recurring moment (every publish), but its trigger phrases partly collide with `calendar`'s "just published". Some users will reach the lighter `calendar` first-hour block instead of `firsthour`.
+- **Recommendation** — **keep** (with a note).
+- **Rationale** — Distinct worked-plan surface + it is the wiring point for an otherwise-orphaned agent (`engagement-coach`); cutting it would re-orphan that agent and undo a core v4.0.0 remediation goal. Note for the operator: the trigger overlap with `calendar`'s publish action means the two should cross-link (calendar's first-hour block could point to `/linkedin:firsthour` for the full plan) — a wiring nit, not a merge.
+
+---
+
+## Group F — Analytics (5)
+
+### `analyze`
+
+- **Purpose** — Performance *troubleshooting/diagnosis*: 6-symptom intake → diagnostic patterns → penalty checklist → severity assessment → 14-day recovery protocol. Optionally grounds in `analytics-interpreter` (interpret mode).
+- **Overlap with siblings** — `audit`, `report`, `import` (calibration case h). Subsumption test: `analyze` is the only *diagnostic/recovery* surface (why is reach down, how do I recover). `report` is *reporting* (weekly numbers from imported data); `audit` is *strategy review* (quarterly top/bottom posts, topic mix); `import` is *ingestion*. Four different jobs on the analytics axis — none subsumes another; `analyze` even works with no imported data (self-report intake).
+- **Invocation leverage** — High and distinct: "why isn't my content performing", "low reach", "reach dropped". A panic-moment intent users type directly.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique troubleshooting/recovery surface; orthogonal to the other three analytics commands.
+
+### `audit`
+
+- **Purpose** — Periodic (quarterly) content-strategy audit: top/bottom performers, topic distribution vs. pillars, format mix, engagement trend, milestone progress, profile alignment — then routes the *fix* to `strategy` and the *profile deep-audit* to `profile`.
+- **Overlap with siblings** — `analyze` (diagnosis), `report` (metrics), `strategy`/`profile` (which it explicitly defers to). Subsumption test: `audit` is the only *holistic strategy-review* surface (90-day lookback across topic/format/trajectory). It deliberately does NOT own trajectory prescription (delegates to `strategy`) or profile checklist (delegates to `profile`) — "audit names the gap; strategy prescribes the fix." That self-limiting design means it does not duplicate them; and `report`/`analyze` don't do the strategy lookback. No subsumption.
+- **Invocation leverage** — Moderate; "content audit", "quarterly review", "review my content strategy". A periodic, distinct intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique holistic-review surface with clean delegation boundaries to `strategy`/`profile`; not a duplicate of `analyze`/`report`.
+
+### `import`
+
+- **Purpose** — Ingest a LinkedIn analytics CSV (auto-detect from ~/Downloads, quick-import helper, the analytics-CLI npm-install fix), parse → JSON → anomaly detection → baseline update, then **delegates the analysis fan-out to `/linkedin:report`** (one analysis pipeline, not two).
+- **Overlap with siblings** — `report` (which it now calls rather than re-implements). Subsumption test: `import` uniquely owns *data ingestion* (CSV detection, copy-to-exports, CLI invocation, baseline creation); `report` owns *presentation* of already-imported data. The Step 6 hand-off is the explicit de-duplication ("keeping a second analysis pipeline here drifted out of sync"). Clean separation — no subsumption, and a redundancy already removed.
+- **Invocation leverage** — Moderate; "import analytics", "import CSV", "parse LinkedIn data". A distinct mechanical intent (you must import before you can report).
+- **Recommendation** — **keep.**
+- **Rationale** — Unique ingestion surface; the analysis overlap with `report` was already collapsed (delegation), which is exactly the discipline this audit endorses.
+
+### `report`
+
+- **Purpose** — Generate weekly/monthly/heatmap performance reports from imported data via the `trends` CLI: metrics table, top performers, 4-week trend, alert detection, `analytics-interpreter` (report mode) recommendations, markdown export.
+- **Overlap with siblings** — `import` (which delegates *to* it), `analyze` (troubleshooting), `audit` (strategy review). Subsumption test: `report` is the only *recurring-numbers reporting* surface and the shared analysis engine `import` reuses. It does not diagnose recovery (`analyze`) or run the quarterly strategy lookback (`audit`). No subsumption.
+- **Invocation leverage** — High and distinct: "weekly report", "performance report", "how did I do", "show my stats". Recurring intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Canonical reporting engine, now the single analysis pipeline both it and `import` use; distinct from `analyze`/`audit`.
+
+### `ab-test`
+
+- **Purpose** — Full A/B testing lifecycle (design → log → analyze → history → suggest) with a directional-not-significant statistical honesty layer; delegates optimized-variant rewrites to `content-optimizer`.
+- **Overlap with siblings** — `report`/`analyze` (it cross-references analytics data; offers "View weekly performance report"). Subsumption test: `ab-test` uniquely owns the *experiment* surface (hypothesis, variant design, running comparison, verdict heuristic). No other command designs or tracks experiments; it merely *reads* analytics for cross-reference. No subsumption.
+- **Invocation leverage** — Moderate-to-high; specific trigger set ("A/B test", "test my hooks", "compare formats", "split test", "which hook works"). Distinct, directly-typed intent.
+- **Recommendation** — **keep.**
+- **Rationale** — Entirely unique experimentation surface; the analytics commands it links to don't touch experiments.
+
+---
+
+## Group G — Growth / strategy (2)
+
+### `strategy`
+
+- **Purpose** — Phase-based growth plan (Phase 0–4 by follower count) + trajectory overlay + authority-building/signature-content compounding (Phase 2+). The canonical trajectory source `audit` defers to. Absorbed the former `authority` command (v2.0.0).
+- **Overlap with siblings** — `audit` (names the gap, routes here for the fix), `profile` (which `strategy` defers to for profile-alignment), `monetize` (Phase 3+ monetization setup is mentioned but not owned here). Subsumption test: `strategy` uniquely owns the phase roadmap + trajectory prescription + authority compounding. It explicitly delegates profile to `profile`. No subsumption.
+- **Invocation leverage** — High; very broad trigger set ("growth plan", "build authority", "signature content", "linkedin roadmap", "my best content"). Also the router's destination for milestone/follower-progress and the below-1K nudge target.
+- **Recommendation** — **keep.**
+- **Rationale** — Deep, canonical growth+authority surface that absorbed `authority`; the hub other commands (`audit`, the 1K-gated ones) route toward.
+
+### `profile`
+
+- **Purpose** — The canonical profile/topic-relevance optimization checklist (7 sections: Headline/About/Experience/Featured/Skills/Network/Engagement) for the 2026 relevance model + the profile-SEO/search-surface layer. The single source `analyze`, `audit`, and `strategy` all defer to.
+- **Overlap with siblings** — `analyze` (Step 6 "If Profile-Content Mismatch" routes here), `audit` (Step 6 routes here), `strategy` (authority audit defers profile signals here). Subsumption test: `profile` is explicitly the *canonical* profile audit — three siblings delegate to it rather than duplicate it. None subsumes it; it subsumes none of them (no content/analytics/growth surface). This is textbook clean delegation.
+- **Invocation leverage** — High and specific: "optimize profile", "profile audit", "fix my profile", "why is my reach low" (profile angle). Distinct, directly-typed intent.
+- **Recommendation** — **keep.**
+- **Rationale** — The single profile source-of-truth that three other commands point to; cutting/merging it would scatter the checklist they deliberately centralized.
+
+---
+
+## Group H — Gated at ~1K followers (3)
+
+### `monetize`
+
+- **Purpose** — Monetization strategy: readiness scorecard, 4 stage-specific plans, lead-magnet blueprint, 4-week funnel, DM conversion workflow, CTA optimization, featured-section + revenue-model worksheets. Soft-gated at ~1K.
+- **Overlap with siblings** — `strategy` (growth phases overlap the monetization stages; `strategy` Phase 3 mentions "lead magnets and monetization setup"), `outreach` (both 1K-gated revenue/opportunity surfaces). Subsumption test: `monetize` uniquely owns the *revenue mechanics* (pricing, offers, funnels, lead magnets, DM conversion). `strategy` only *mentions* monetization as a Phase 3+ focus area; it does not build a funnel or price an offer. `outreach` is collaborations/speaking, not products/funnels. No subsumption.
+- **Invocation leverage** — Moderate; specific ("monetize", "lead generation", "consulting pipeline", "pricing strategy", "lead magnet"). Soft-gated, so naturally lower-frequency until the audience exists.
+- **Recommendation** — **keep.**
+- **Rationale** — Deep, unique revenue-mechanics surface; `strategy`'s monetization mention is a pointer, not a duplicate. The soft-gating is honest (works at any count, value compounds at 1K).
+
+### `outreach`
+
+- **Purpose** — Two-track orchestrator (collaborations + speaking) under one pitch paradigm: readiness, partner/event search + scoring, 12 collab formats + 4 talk templates, outreach messages, co-creation workflow + speaker portfolio, pipeline trackers (now state-persisted via `recordOutreachContact`), progression ladders. Absorbed the former `collab` + `speaking` commands (v2.0.0). Delegates to `network-builder`.
+- **Overlap with siblings** — `monetize` (both 1K-gated), `strategy` (Phase 2+ collaborations). Subsumption test: `outreach` uniquely owns the partner/event pitch machinery. `strategy` only *names* collaborations as a focus area; `monetize` is products not partnerships. No subsumption. Internally it consolidated two commands and has an explicit capability checklist proving nothing was lost in that merge — a model of disciplined consolidation.
+- **Invocation leverage** — Moderate; very broad trigger set spanning both tracks ("collaboration", "co-author", "speaking", "CFP", "pitch a talk", "outreach"). Soft-gated.
+- **Recommendation** — **keep.**
+- **Rationale** — Unique, deep two-track outreach surface that already consolidated `collab`+`speaking` cleanly; not subsumed.
+
+### `competitive`
+
+- **Purpose** — Competitive analysis of niche thought leaders: identify competitors (WebSearch), content/hook/engagement analysis, landscape map, gap analysis, differentiation strategy, inspired-not-copied takeaways.
+- **Overlap with siblings** — `strategy` (differentiation is a growth lever there), `audit` (analyzes *your* content; this analyzes *others'*). Subsumption test: `competitive` is the only *external/competitor-facing* analysis surface. `audit` is inward-facing; `strategy` prescribes *your* plan, not a competitor scan. No subsumption.
+- **Invocation leverage** — Low-to-moderate; specific ("competitive analysis", "analyze competitor", "what are others doing", "learn from others"). A periodic, distinct intent. CLAUDE.md/README describe it as 1K-gated in the same breath as monetize/outreach, though the file itself carries no gating logic.
+- **Recommendation** — **keep** (with a minor honesty note).
+- **Rationale** — Unique outward-facing competitive surface; nothing else scans competitors. Note: the router does not actually list `competitive` under the 1K-gated group nor apply the below-1K nudge to it (unlike `monetize`/`outreach`), and the command body has no follower check — a small documentation-vs-behavior inconsistency for the operator to reconcile, not a reason to merge or cut.
+
+---
+
+## Summary table
+
+| Command | Recommendation | Merge target | One-line reason |
+|---------|---------------|--------------|-----------------|
+| `linkedin` | keep | — | Structural router; highest-leverage front door (minor map-staleness nit). |
+| `post` | keep | — | Canonical long-band creation + refine/optimizer surface; core entry. |
+| `quick` | keep | — | Unique short-band + the 8-template library `templates` folded into. |
+| `react` | keep | — | Unique URL-ingestion + multi-source synthesis; directly-typed intent. |
+| `pipeline` | keep (flag) | — | No clean subsumption, but a thin re-statement of `post`+`calendar`+`firsthour`; watch for erosion. |
+| `carousel` | keep | — | Deep unique carousel/document engine + image generation. |
+| `video` | keep | — | Unique LinkedIn-native video scripting + `video-scripter`. |
+| `multiplatform` | keep (flag) | — | Unique Twitter/X + cross-platform surface; thinnest command, develop-candidate (could wire `content-repurposer`). |
+| `newsletter` | keep | — | The long-form spine; everything long-form routes here. |
+| `headless-review` | keep | — | Unique fresh-session cold-isolation a `newsletter` phase can't provide. |
+| `pivot` | keep | — | Unique re-open state surgery `newsletter` delegates out. |
+| `onboarding` | keep | — | Unique cohesive wizard + cross-phase triage; orchestrates, not duplicates. |
+| `setup` | keep | — | Canonical full 8-category personalization; `onboarding` defers here. |
+| `first-post` | keep | — | Unique first-timer scaffolding + `first_post_date` side-effect. |
+| `batch` | keep | — | Unique multi-post-week + iCal; queue-write is shared plumbing. |
+| `calendar` | keep | — | Canonical queue + publish action (absorbed `publish`). |
+| `firsthour` | keep (note) | — | Unique worked first-hour plan; wiring point for orphaned `engagement-coach`. |
+| `analyze` | keep | — | Unique troubleshooting/recovery surface; works without imported data. |
+| `audit` | keep | — | Unique holistic strategy review; delegates fix to `strategy`/`profile`. |
+| `import` | keep | — | Unique ingestion surface; analysis already delegated to `report`. |
+| `report` | keep | — | Canonical reporting engine + shared analysis pipeline. |
+| `ab-test` | keep | — | Entirely unique experimentation lifecycle. |
+| `strategy` | keep | — | Canonical phase+trajectory+authority hub (absorbed `authority`). |
+| `profile` | keep | — | Single profile source-of-truth three commands delegate to. |
+| `monetize` | keep | — | Unique revenue-mechanics surface; honest soft-gating. |
+| `outreach` | keep | — | Unique two-track pitch surface (absorbed `collab`+`speaking`). |
+| `competitive` | keep (note) | — | Unique outward competitor scan; minor gating doc-vs-behavior nit. |
+
+**Tally: keep 27 · merge 0 · cut 0.** Two keeps carry a flag — `pipeline` (inconclusive, retained) and `multiplatform` (develop-candidate).
+
+---
+
+## Count-impact summary (if every recommendation were applied as-is)
+
+Every recommendation is **keep**. Therefore, if all were applied:
+
+- **Commands removed:** 0 (0 cut + 0 merged-away).
+- **Resulting count:** **27** — unchanged.
+- **Lint / roster surfaces needing update:** **none.** `EXPECT_COMMANDS=27` in
+  `scripts/test-runner.sh` stays as-is; the CLAUDE.md command table (27), the
+  README command table, the `commands` header the lint checks, the router
+  `commands/linkedin.md`, and any SKILL roster all remain at 27. No roster
+  touched, no count contract changed.
+
+This mirrors the agent-overlap outcome in `overlap-measurement.md` (no agent
+trimmed → 19/27 baseline unchanged): the command surface, audited under the same
+subsumption discipline, also holds — every command has a unique surface or a
+justified redundancy, and the two prior consolidations (`templates`/`publish`/
+`authority`/`collab`/`speaking` → `quick`/`calendar`/`strategy`/`outreach`, and
+the `import`→`report` analysis delegation) have already removed the genuine
+duplications. The surface is over-grown only in *count*, not in *redundancy*.
+
+The recommendations are informational; the operator decides per command. Two
+non-blocking honesty nits surfaced along the way (each a wiring/doc fix, not a
+merge/cut; a claimed third was withdrawn on verification):
+
+1. The router lists `post-feedback-monitor` (agent) under Post-Publish but not the
+   new `/linkedin:firsthour` command.
+2. `calendar`'s publish-action first-hour block and `firsthour` share trigger
+   phrases ("just published") and should cross-link.
+3. ~~`competitive` 1K-gating doc-vs-behavior.~~ **Withdrawn on verification:**
+   CLAUDE.md `:64`, README `:222`, the router, and the command body all leave
+   `competitive` ungated — only `monetize`/`outreach` carry "(unlocks at ~1K)".
+   No inconsistency exists; the claimed nit was unfounded.
+
+---
+
+## Honest-uncertainty list (inconclusive — retain pending operator view)
+
+- **`pipeline`** — *inconclusive, retained.* It has no clean subsumption (it
+  uniquely *sequences* draft → schedule → first-hour → analysis into one guided
+  run), which under the trim rule mandates retention. But its constituent steps
+  are each a thinner inline condensation of richer leaf commands
+  (`post` / `calendar` / `firsthour`), and its trigger set is narrow/jargon-y, so
+  its real-world leverage as a *natural type* is the weakest in Group B. I could
+  not decide whether the orchestration value justifies the standalone command
+  versus folding its unique sequencing into `post` (with an explicit
+  schedule + first-hour hand-off). Per the methodology: **inconclusive — retain
+  pending operator view.** This is the one command where a future, operator-blessed
+  merge could be defensible; the evidence informs but does not compel it.
+
+All other 26 commands are clear **keep** decisions with no residual uncertainty.

plugins/linkedin-studio/docs/remediation/finish-plan.md

@@ -47,6 +47,21 @@ S17 (C13–C46 triage)            → ALLOW  [process complete]
 ## S14 — Command rationalization (re-opens the original command-surface Non-Goal)
 *Analysis → operator decision → execute. Nothing deleted without explicit per-command yes.*
 
+> **AMENDMENT — S14 reframed by operator decision (2026-05-30): merge/cut → journey layer.**
+> 14a's cold per-command review (`command-rationalization.md`) found **zero redundancy** —
+> no command is a defensible merge/cut candidate (the two prior consolidations already
+> removed the genuine overlaps). So instead of cutting, the operator chose to **add a
+> journey layer over the kept atomics**. The build contract is
+> **`journey-layer-design.md`** (this supersedes the 14b/14c "execute merges/cuts" bullets
+> below — there are none to execute). **Delivered:** two new guided front-doors
+> (`/linkedin:create`, `/linkedin:measure`) + the router re-tiered into five journeys
+> (Start · Create · Engage · Measure · Grow) + `onboarding`/`strategy` elevated as the
+> Start/Grow front-doors; the 27 atomic commands kept; **27 → 29 commands; v4.0.0 → v4.1.0
+> (minor/additive)**. The two 14a honesty nits real on verification were fixed (router lists
+> `firsthour`; `calendar` cross-links to it); a third (a `competitive` 1K-gating claim) was
+> withdrawn as unfounded. The **gate is unchanged**: `test-runner.sh` + `node --test` green
+> → `/trekreview` **ALLOW** (no WARN-override) → commit own files → push.
+
 - **14a Analysis (no edits):** cold per-command review of all 27 → `docs/remediation/command-rationalization.md`.
   Per command: purpose · overlap with siblings · invocation leverage (algorithmic + likely use) ·
   recommendation **keep / develop / merge→X / cut** + rationale. Delegate the cold read to an

plugins/linkedin-studio/docs/remediation/journey-layer-design.md

@@ -0,0 +1,154 @@
+# LinkedIn Studio — Journey-Layer Design (S14)
+
+> **Supersedes the S14 "merge/cut" framing.** 14a's cold review found 27 commands
+> with **zero redundancy** (subsumption fails pairwise) — the surface is over-grown
+> in *count*, not in *redundancy*. The operator therefore reframed S14 from
+> "rationalize by cutting" to "**add a journey layer over the kept atomics**." This
+> doc is the build contract. It also **absorbs S15's router-tiering** (finish-plan
+> B2) — the router is re-tiered here, journey-style, so S15 collapses into S14.
+
+## Decisions locked (operator, 2026-05-30)
+
+1. **Architecture = journey layer.** Keep all 27 atomic commands as the *execution
+   tier*; add journey *front-door* orchestrators as the *primary surface*; subagents
+   remain the engine (no Workflow-tool / deep-research adoption — both are wrong-fit
+   for an everyday product surface, see the session reasoning).
+2. **Realization = 2 new + 2 elevated.** Add `create` + `measure` as new front-doors;
+   elevate existing `onboarding` (Start) and `strategy` (Grow); **Engage** stays a
+   router-tier relay (`calendar` + `firsthour`), no front-door. Atomics → "advanced"
+   tier under each journey.
+3. **Version = v4.1.0 (minor), NOT v5.0.0.** Honest SemVer: this is additive (2 new
+   commands) + presentation (router retier) + one tiny doc-vs-behavior reconcile
+   (`competitive`). No removal, rename, or behavior break on existing commands.
+   Precedent: v3.1.0 added 3 agents as a MINOR with a "reload required" note — same
+   shape. Reload note applies (the 2 new commands register at session reload).
+
+## Journey taxonomy (5 journeys + router)
+
+`Improve` is **not** a separate journey: improving is the action *after* measuring
+(audit→fix, ab-test, `content-optimizer`), not a distinct entry. The operator's
+"Improve + Analyze" is therefore folded into **Measure** (the measure→optimize loop).
+
+| Journey | Front-door | Atomics (advanced tier) |
+|---------|-----------|--------------------------|
+| **Start** | `onboarding` *(elevate — exists)* | setup · first-post |
+| **Create** | **`create`** *(NEW)* | post · quick · react · carousel · video · multiplatform · batch · newsletter · pivot · headless-review · pipeline |
+| **Engage** | *(router-tier — relay, no front-door)* | calendar · firsthour |
+| **Measure** | **`measure`** *(NEW — `analyze` name is taken, so no collision)* | import · report · analyze · audit · ab-test |
+| **Grow** | `strategy` *(elevate — exists)* | profile · competitive · monetize ⚿ · outreach ⚿ |
+| *(map)* | `linkedin` router | — |
+
+⚿ = unlocks/soft-gates at ~1K followers. Count: Start 3 + Create 11 + Engage 2 +
+Measure 5 + Grow 5 + router 1 = **27 atomics** + **2 new front-doors** = **29**.
+
+## The two new front-door commands
+
+Both are **thin guided routers** — they add *navigation* (a single "what do you
+want?" entry), not duplicated logic. They **delegate** to the atomic that owns the
+work. No drafting/analysis logic is copied into them.
+
+### `commands/create.md` — Content-creation front-door
+- **Purpose:** one guided entry for "make something" when the user hasn't already
+  named the format.
+- **Flow:** `AskUserQuestion` "What do you want to create?" →
+  - Short substantial post → `post`
+  - Quick 5-min post → `quick`
+  - React to a URL / article / news → `react`
+  - Carousel / document → `carousel`
+  - Video script → `video`
+  - Adapt for another platform → `multiplatform`
+  - A full week (batch) → `batch`
+  - Long-form / newsletter / essay → `newsletter`
+  - Then hand off to the chosen command (it owns the work).
+- **Triggers (undirected create intent only — must NOT steal `post`/`quick` direct
+  intent):** "create", "make something", "create content", "what should I make",
+  "new content", "help me create", "linkedin create".
+- **allowed-tools:** Read, Glob, AskUserQuestion. **No Write** (it delegates).
+
+### `commands/measure.md` — Performance/insight front-door
+- **Purpose:** one guided entry for "understand how I'm doing."
+- **Flow:** `AskUserQuestion` "What do you want to do?" →
+  - Import a new analytics CSV → `import`
+  - Weekly / monthly report → `report`
+  - Diagnose a problem (reach dropped, low engagement) → `analyze`
+  - Quarterly strategy audit → `audit`
+  - Design / review an A/B test → `ab-test`
+  - Then hand off to the chosen command.
+- **Triggers (undirected measure intent):** "measure", "how am I doing", "my
+  performance", "show my analytics", "performance overview", "linkedin measure".
+- **allowed-tools:** Read, Glob, AskUserQuestion. **No Write** (it delegates).
+
+## Elevation (2) — presentation only, no behavior change
+- **`onboarding` = Start front-door.** Already the cohesive wizard. Router presents
+  it as the Start journey entry; body unchanged (optional one-line "Start journey"
+  framing only).
+- **`strategy` = Grow front-door.** Already the phase/trajectory/authority hub.
+  Router presents it as the Grow journey entry; body unchanged.
+
+## Router re-tiering (`commands/linkedin.md`) — absorbs S15 B2
+Restructure the menu into the 5 journeys, each headed by its front-door with the
+atomics listed beneath as "also / advanced":
+- **Start** → `onboarding` · setup · first-post
+- **Create** → `create` · post · quick · react · carousel · video · multiplatform · batch · newsletter · pivot · headless-review · pipeline
+- **Engage** → calendar · firsthour  *(tier, no front-door)*
+- **Measure** → `measure` · import · report · analyze · audit · ab-test
+- **Grow** → `strategy` · profile · competitive · monetize ⚿ · outreach ⚿
+- Keep the status line (weekly progress / streak / follower phase) + queue summary.
+- **Fixes nit #1:** `/linkedin:firsthour` is now listed (under Engage) — the router
+  previously listed only the `post-feedback-monitor` agent there.
+
+## Honesty nits fixed in lockstep (from 14a)
+1. **Router omits `firsthour`** → fixed by the retier (Engage tier lists it).
+2. **`calendar` ↔ `firsthour` trigger overlap** → `calendar.md`'s publish-action
+   first-hour block gets a one-line cross-link: "for the full worked sprint plan,
+   run `/linkedin:firsthour`."
+3. **`competitive` gating doc-vs-behavior** → **14a's nit was unfounded** (verified
+   against the files: CLAUDE.md `:64`, README `:222`, the router, and the command body
+   all leave `competitive` **ungated** — only `monetize`/`outreach` carry "(unlocks at
+   ~1K)"). There is no inconsistency to fix. Competitive analysis helps at any stage
+   (especially early positioning), so it stays ungated. Action: place it in the Grow
+   tier marked **"Any phase"** (no ⚿), and add a one-line note to the router's gating
+   rule that `competitive` is explicitly not gated. (An interim router edit that
+   mistakenly added a ~1K marker was reverted in the same pass.)
+
+## 14a doc correction (`command-rationalization.md`)
+- Add the missing **`multiplatform`** entry: **keep** (unique Twitter/X-thread +
+  cross-platform surface; subsumption fails as a cut), flagged **thinnest /
+  develop-candidate** (only command that inlines static templates instead of
+  delegating — could use `content-repurposer`). Belongs in **Create**.
+- Fix the two header-count typos (Group C "(4)"→3; Group F "(4)"→5).
+- Corrected tally: **keep 27 (27/27 covered)** · 0 merge · 0 cut · 1 develop-candidate.
+
+## Count / version / roster lockstep (the release mechanics)
+- `scripts/test-runner.sh`: `EXPECT_COMMANDS` **27 → 29**.
+- Version **4.0.0 → 4.1.0** everywhere it appears (version-sync rule: grep both
+  `4.0.0` and the `27`-count claims).
+- Three-doc rule (same commit / immediately after):
+  - `CLAUDE.md` (plugin): command table +2 rows, count 27→29, version, v4.1.0 para.
+  - `README.md` (plugin): command table, count, version, CHANGELOG entry.
+  - Root `README.md` (marketplace): the linkedin-studio line.
+- Also: `CHANGELOG.md` if present; any SKILL roster enumerating commands; the router
+  roster (done above); `STATE.md` count block (27→29, version 4.1.0).
+
+## Verification (testable — gate before push)
+- `ls commands/*.md | wc -l` == **29**.
+- `bash scripts/test-runner.sh` exit **0** (EXPECT_COMMANDS=29; stat/version/model
+  consistency + render-chain + $-safety guards all green).
+- `node --test hooks/scripts/__tests__/*.test.mjs` — all pass (no hook logic touched
+  → expect the existing **98**).
+- `grep -rn "4\.0\.0\|27 command" commands README.md CLAUDE.md` (user-facing) → 0
+  stale hits (historical CHANGELOG mentions allowed).
+- Router renders 5 journeys; primary tier = the 5 front-doors (`create`, `measure`,
+  `onboarding`, `strategy` + Engage tier); atomics nested; 1K-gated flagged;
+  `firsthour` present.
+- `create` / `measure`: valid frontmatter, `AskUserQuestion` routing, **delegate
+  only** (grep shows no inlined drafting/analysis logic, no `Write` in create/measure).
+- `/trekreview --project docs/remediation/` → **ALLOW** (no WARN-override) → commit
+  (own files only) → push origin.
+
+## Out of scope (this session / S14)
+- Aggressive consolidation / folding atomics (operator chose keep-all).
+- Building the `multiplatform` "develop" (content-repurposer wiring) — recorded as a
+  keep + develop-candidate, not built unless requested.
+- S16 (saves) / S17 (C13–C46) — later finish-plan steps.
+- UI-brief M0 (move mutable data out of tree) — pending; conflicts with S16, not S14.

plugins/linkedin-studio/docs/remediation/review.md

@@ -1,13 +1,13 @@
 ---
 type: trekreview
 review_version: "1.0"
-task: "Remediate linkedin-studio from the baseline audit — correctness, honesty, generalization, and the highest-leverage 2026 coverage gaps (full Phase 0–3 roadmap, phased)"
+task: "S14 — journey layer over the LinkedIn Studio command surface (operator-reframed from merge/cut after 14a found zero redundancy)"
 slug: remediation
 project_dir: docs/remediation/
 brief_path: docs/remediation/brief.md
-scope_sha_start: c5b4c58f4f390aca83c8937880c5fd0bcc983e44
-scope_sha_end: 36f79dd702b9315a0cd9100c3a8dd6dd81b3797f
-reviewed_files_count: 50
+scope_sha_start: 431a893
+scope_sha_end: 431a893
+reviewed_files_count: 15
 verdict: ALLOW
 mode: default
 effort: high
@@ -15,152 +15,132 @@ profile: premium
 findings: []
 ---
 
-# Review — linkedin-studio audit-remediation (S13 re-review: `$`-class closure + scalar-test fix)
+# Review — linkedin-studio S14 (journey layer; merge/cut → add a layer)
 
 ## Executive Summary
 
-**Verdict: ALLOW** — 0 BLOCKER, 0 MAJOR, 0 MINOR, 0 SUGGESTION.
+**Verdict: ALLOW** (post-remediation) — 0 BLOCKER, 0 MAJOR, 0 MINOR, 0 SUGGESTION open.
+Two independent reviewers (brief-conformance, code-correctness) ran COLD, high-effort,
+without cross-feeding, on the as-delivered uncommitted working tree (HEAD `431a893` +
+the S14 delta). **Each returned 2 MAJOR findings; deduped to 3 distinct issues.** All
+three were **remediated in-session** with deterministic re-verification; the final
+state that pushes is clean.
 
-This is the **S13 re-review** — the **seventh** full-brief sweep
-(`c5b4c58..36f79dd` + the uncommitted S13 working-tree delta), run COLD and
-high-effort. S13 was commissioned to close the two findings the S12 re-review
-left open (verdict WARN, 0/1/1/0): the MAJOR `MISSING_TEST` (the S12 `$`-bearing
-test asserted the Recent Posts section but never the `last_post_topic` scalar, so
-the corruption shipped green) and the MINOR `MISSING_ERROR_HANDLING` (the
-`last_post_topic` scalar was still `$`-unsafe because `replaceField` used a
-replacement *string* for untrusted content). Two independent reviewers
-(brief-conformance, code-correctness) ran without cross-feeding; the coordinator
-applied bounded dedup + the HubSpot Judge filters + verdict (high-effort →
-Cloudflare reasonableness filter skipped; the operator weighs borderline findings).
-**Both reviewers returned empty finding sets.**
+S14 was **reframed by operator decision (2026-05-30)**: 14a's cold per-command review
+(`command-rationalization.md`) found **zero redundancy** across the 27 commands (no
+defensible merge/cut), so instead of cutting, a **journey layer** was added over the
+kept atomics. Build contract: `journey-layer-design.md`. Delivered: two new
+delegate-only front-doors (`/linkedin:create`, `/linkedin:measure`), the router
+re-tiered into five journeys (Start · Create · Engage · Measure · Grow), `onboarding`/
+`strategy` elevated as the Start/Grow front-doors; 27 → 29 commands; v4.0.0 → v4.1.0
+(minor/additive). Two of 14a's honesty nits (router lists `firsthour`; `calendar`
+cross-links to it) were real and fixed; a third (a `competitive` 1K-gating claim) was
+withdrawn as unfounded on verification.
 
-**Both S12 findings are CLOSED at the reviewed (uncommitted) state:**
-- `replaceField` (`hooks/scripts/state-updater.mjs:14-24`) now uses a replacement
-  **function** (`() => \`${field}: ${value}\``), so the untrusted `last_post_topic`
-  at the `:64`-equivalent call site is inserted verbatim — no `$&`/`$1`/`` $` ``
-  expansion. The MINOR is closed.
-- The existing `$`-bearing test (`hooks/scripts/__tests__/state-updater.test.mjs`)
-  now carries `assert.match(result.content, /^last_post_topic: "\$100 budget — \$&
-  and \$1 rule"$/m, …)`, distinct from the section-entry `includes()` it already
-  had. This assertion **fails on the old string-`replaceField` and passes on the
-  function form** (orchestrator-verified by reverting the fix: the test went
-  40 pass / 1 fail). The false-green MAJOR is closed.
-
-**The class — not just the line — is closed.** The recurring S9→S12 lesson is
-"close the class, not the line"; the class here is "untrusted user content reaching
-ANY `String.replace` replacement *string*". Beyond the `replaceField` scalar, S13
-also converted the three remaining additive-insert sites (`recordFirstHourPlan`
-`:246`-equiv; `recordOutreachContact` `:305/:308`-equiv) from a string replacement
-carrying an intentional `$1` backref + interpolated date to a replacement function
-(`(m) => \`${m}\n…\``). The code-correctness reviewer verified rigorously that this
-is **behavior-preserving**: each regex's capture group spans the *entire* match (the
-only chars outside the group are the zero-width `^`/`$` anchors), so the full match
-`m` is character-identical to the old `$1`. After S13, **every `.replace()` in
-`state-updater.mjs` uses a replacement function or a `$`-free literal** — the class
-is closed by construction, not by per-line patch.
-
-**A structural guard replaces the per-line proof.** New
-`scripts/check-replace-safety.mjs` (wired as `test-runner.sh` Section 12) proves the
-property behaviorally: it drives every exported mutator with an adversarial payload
-of every special replacement token (`$&`, `` $` ``, `$'`, `$$`, `$n`) in every
-free-text *and* date field and asserts the payload survives verbatim. Two structural
-backstops run on every invocation — **coverage-completeness** (a newly-exported
-mutator without `$`-coverage fails the guard) and a **non-vacuity self-test** (a
-naive string-replace MUST corrupt the payload and a function MUST preserve it, else a
-PASS is meaningless), mirroring the Section 8/10/11 self-tests. The orchestrator
-mutation-proved it end-to-end: reverting `replaceField` to a string makes the guard
-exit 1 with two findings; restoring it returns exit 0.
-
-**No Phase-0–3 Success Criterion regressed.** The brief-conformance reviewer traced
-each S13 clause to delivered code and confirmed the counts (19 agents / 27 commands /
-25 references / 6 skills), the version (4.0.0), the single-source algorithm-signal,
-the model-consistency guard, and the render-chain-propagation guard all still hold;
-S13 touched no command/agent/reference file. The two Non-Goals the brief amendment
-re-opens (the command invocation surface for S14; saves manual-entry for S16) trace
-to **explicit operator decisions** in the brief amendment, and S13 itself did not
-touch either surface — no `SCOPE_CREEP_BUILT`.
-
-**Push decision: ALLOW.** The two S12 findings are closed, the class is closed
-structurally, the lint is non-vacuous and mutation-proven, all suites are green
-(`scripts/test-runner.sh` → 71/0/0; `node --test` → 98/98), and no SC regressed. The
-ORIGINAL remediation brief now closes clean. Per `feedback_trekreview_always_last` +
-Handover 6, this review is the gate; with ALLOW, S13 may push.
+**What both reviewers confirmed conformant + correct** (the bulk of the delivery):
+the two front-doors are delegate-only (no `Write`; route to the contracted targets);
+the router re-tier preserves **reachability for all 29 commands** (no atomic dropped;
+`post-feedback-monitor` still mentioned in prose, not orphaned); `EXPECT_COMMANDS`
+27→29 + `## Commands (29)` + `ls`=29 agree; version 4.1.0 is consistent across
+plugin.json / README version badge / CLAUDE.md header / CHANGELOG / root docs; SemVer
+framing is honest (additive, nothing falsely "breaking", nothing removed/renamed);
+out-of-scope discipline held (no atomic folded; `multiplatform`-develop only recorded,
+not built; S16/S17/UI-brief-M0 untouched).
 
 ## Coverage
 
-Scope SHA range: `c5b4c58` (= `origin/main`, parent of remediation Steg 1) →
-`36f79dd` (HEAD, the S12 commit) **plus the uncommitted S13 working-tree delta**
+Scope: HEAD `431a893` (S13's commit) + the **uncommitted S14 working-tree delta**
 (annotated `[uncommitted]` — a brief-level contract; the brief's Assumptions allow
-uncommitted review). The committed range (47 files) was already deep-reviewed and
-cleared at S12 except the 2 WARN findings; the active S13 delta is the 9 working-tree
-files below. **No silent skips.**
+uncommitted review). 15 files = the operator's own changes; the 3 untracked
+not-mine files (`docs/linkedin-studio-persona-brief.md`, `…-ui-brief.md`,
+`docs/voyage-build/progress.json`) are explicitly excluded from scope and from the
+commit. **No silent skips.**
 
 | Treatment | Count | Notes |
 |-----------|-------|-------|
-| `deep-review` (hooks/** + the new guard) | 4 | `state-updater.mjs`, `state-updater.test.mjs` `[uncommitted]`; `scripts/check-replace-safety.mjs` `[uncommitted, new]`; `scripts/test-runner.sh` `[uncommitted]` |
-| `summary-only` | 46 | the committed `c5b4c58..36f79dd` range (already cleared at S12) + the S13 doc edits `CLAUDE.md`/`README.md`/`docs/integration-test-guide.md` `[uncommitted]` + `docs/remediation/{brief.md (amendment), finish-plan.md}` `[uncommitted]` |
+| `deep-review` | 6 | `commands/create.md` `[new]`, `commands/measure.md` `[new]`, `commands/linkedin.md` (router re-tier), `commands/calendar.md`, `scripts/test-runner.sh`, `README.md` |
+| `summary-only` | 9 | plugin `CLAUDE.md`/`CHANGELOG.md`/`plugin.json`/`commands/onboarding.md`; root `CLAUDE.md`/`README.md`; `docs/remediation/{finish-plan.md, command-rationalization.md, journey-layer-design.md}` |
 | `skip` | 0 | no lockfiles / svg / generated / dist |
 
 **Cross-cutting execution criteria (run by orchestrator):**
-`scripts/test-runner.sh` → 71 passed / 0 failed / 0 warnings, exit 0 (was 70;
-+1 — Section 12 `$`-safety guard).
-`node --test hooks/scripts/__tests__/*.test.mjs` → 98 tests, 98 pass, 0 fail (the
-S13 scalar assertion was added to an existing test, not a new test).
-`node scripts/check-replace-safety.mjs` → exit 0 (8 adversarial cases / 5 mutators,
-coverage-complete, self-test non-vacuous); mutation-proven (reverting the fix → exit 1).
-
-**S12 findings — both confirmed CLOSED:**
-`replaceField` (`state-updater.mjs:14-24`) → replacement function; the `$`-bearing
-test now pins the `last_post_topic` scalar (`state-updater.test.mjs`); the three
-remaining additive-insert string-replacements (`:246/:305/:308`) → functions; the
-class is closed and guarded structurally (Section 12).
+`scripts/test-runner.sh` → **74 passed / 0 failed / 0 warnings**, exit 0 (was 71;
++2 the two new commands' frontmatter checks, +1 a new README commands-badge guard,
+−1 net rounding of the count-loop — all green). `node --test …/*.test.mjs` → **98/98**
+(no hook logic changed). `ls commands/*.md` → **29**.
 
 ## Findings
 
-None. Both independent reviewers returned empty finding sets; the coordinator's
-bounded passes (dedup, HubSpot Judge, verdict) on an empty set yield ALLOW.
+**3 distinct MAJOR findings were raised by the independent reviewers and ALL closed
+in-session.** No finding remains open; the trailing JSON `findings` is empty because
+the pushed state carries none.
+
+### [MAJOR — RESOLVED] README commands badge stale at `commands-27`
+*Raised by both reviewers (`COUNT_DRIFT` / `SUCCESS_CRITERION_UNMET`), `README.md:11`.*
+The release reconciled the version badge, intro prose, CLAUDE.md header, `EXPECT_COMMANDS`,
+and rosters to 29, but the shields **commands** badge still read `commands-27-green`. It
+slipped both the human pass and the gate (the version-consistency grep checks only the
+*version* badge; the design-doc verification grep searched `"27 command"` with a space,
+which cannot match `commands-27-green`).
+**Resolution:** `README.md:11` → `commands-29-green`; **and** a new
+`test-runner.sh` Section-2 guard now asserts the `badge/commands-${EXPECT_COMMANDS}-`
+shields badge matches the contract (closing the class, not just the line — the
+count-badge analogue of the existing version-badge check). Re-verified: lint 74/0/0;
+`grep commands-27` → 0 hits.
+
+### [MAJOR — RESOLVED] Router self-contradicted on `competitive` gating
+*Raised by code-correctness (`INTERNAL_CONTRADICTION`), `commands/linkedin.md:140`.*
+The Grow journey **table** marked `competitive` "Any phase" and the gating-rule note
+said it is **not** gated, but the **"Ask the User"** menu line still rendered
+`competitive ⚿` (the file's own ~1K soft-gate marker). An interim erroneous gating
+(added then reverted in the same pass) had been reverted in the table + the gating note
++ the design doc, but the menu line was missed — the exact inconsistency the release
+claims to fix.
+**Resolution:** `commands/linkedin.md` Ask-the-User Grow line → `competitive` (no ⚿);
+re-verified `grep "competitive ⚿"` → 0 hits across commands + the design doc. The router
+now states competitive ungated consistently in the table, the note, and the menu.
+
+### [MAJOR — RESOLVED] STATE.md binding count block stale at 27 / v4.0.0
+*Raised by brief-conformance (`SUCCESS_CRITERION_UNMET`), `STATE.md:46`.*
+The design-doc lockstep enumerates the STATE.md count block as a surface to update
+(27→29, v4.1.0). `STATE.md` is **gitignored** (not part of the pushed commit, so not a
+push-blocking three-doc violation), but it is a contracted lockstep item and would
+mislead the next session.
+**Resolution:** updated to 29 / v4.1.0 as part of the session-close STATE.md overwrite
+(which also advances the pointer to S15). Not in the pushed artifact (gitignored).
 
 ## Remediation Summary
 
-**Gate: ALLOW.** The S12 WARN is fully resolved: the `replaceField` scalar
-`$`-corruption (MINOR) and the false-green test (MAJOR) are both closed; the
-`$`-injection class is closed across the whole `state-updater.mjs` mutation surface;
-and a behavioral, coverage-complete, self-testing Section-12 lint guards it
-structurally against future regressions. All suites green; no Phase-0–3 SC regressed;
-the two re-opened Non-Goals trace to explicit operator decisions and S13 stayed in its
-lane.
+**Gate: ALLOW (post-remediation).** Two independent high-effort reviewers found 3
+distinct MAJORs — all stale-count / incomplete-revert completion-misses of THIS
+session's own committed lockstep, none a design flaw. All three were closed in-session
+with deterministic re-verification (lint 74/0/0; targeted greps → 0 residual hits); a
+new lint guard prevents the badge regression class from recurring. No BLOCKER; no
+design or correctness defect in the journey layer itself; reachability for all 29
+commands preserved; SemVer honest. The pushed state is clean.
 
-**Two non-blocking observations, recorded by both reviewers, neither rising to a
-catalogue finding:**
-1. The S13 dead binding both reviewers named (`check-replace-safety.mjs` `HERE` +
-   its now-unused `node:url`/`node:path` imports) was removed during this review
-   pass; the guard remains green (exit 0) after removal.
-2. The behavioral guard catches a new *unguarded exported mutator* (coverage
-   backstop) but not a new unsafe `String.replace` added *inside* an existing
-   battery-covered mutator on a field the battery does not fuzz. This is a documented,
-   deliberate limit of the behavioral proxy (vs a per-`.replace()` AST enumeration)
-   and is **moot today** — every `.replace()` in `state-updater.mjs` is already a
-   function. Recorded as a known boundary, not a defect; closing it further is out of
-   S13 scope (no real `$`-unsafe site exists to catch).
+Decision rationale for in-session fix (vs `feedback_trekreview_always_last`'s
+fix-in-next-session): the three findings are mechanical completions of the current
+task's own lockstep — #1 and #2 would have failed the design doc's own Verification
+grep, #3 is an incomplete revert — i.e. *unfinished scope of S14*, not deliberation-
+needing review findings. Shipping a README badge that reads "27" and a router that
+self-contradicts on gating would be dishonest to the release's own claims. Closing them
+now reaches a genuine ALLOW (not a WARN-override).
 
-The two adjacent machine-value `.replace()` sites the correctness reviewer probed —
-`session-start.mjs:396` (`${actualWeek}`, a computed ISO week) and `week-rollover.mjs`
-(computed week / literal int) — carry no untrusted content and are therefore not
-members of the defect class, consistent with how the S12 review itself classified the
-date/integer `replaceField` call sites. No finding.
-
-Per Handover 6, this `review.md` is consumable by
-`/trekplan --brief docs/remediation/review.md`; the trailing JSON block is the machine
-contract for that handover. With an ALLOW verdict and no BLOCKER/MAJOR findings, no
-follow-up remediation plan is required — the ORIGINAL brief is closed clean and the
-finish-plan continues at S14.
+Per Handover 6, this `review.md` is consumable by `/trekplan --brief …`; with an ALLOW
+verdict and no open findings, no follow-up plan is required — S14 may commit + push, and
+the finish-plan continues at S15.
 
 ```json
 {
   "verdict": "ALLOW",
-  "scope": { "sha_start": "c5b4c58f4f390aca83c8937880c5fd0bcc983e44", "sha_end": "36f79dd702b9315a0cd9100c3a8dd6dd81b3797f", "reviewed_files_count": 50, "uncommitted_delta": true },
+  "scope": { "sha_start": "431a893", "sha_end": "431a893", "reviewed_files_count": 15, "uncommitted_delta": true },
   "counts": { "BLOCKER": 0, "MAJOR": 0, "MINOR": 0, "SUGGESTION": 0 },
   "findings": [],
+  "resolved_in_session": [
+    { "severity": "MAJOR", "title": "README commands badge stale at 27", "file": "README.md", "line": 11, "rule_key": "COUNT_DRIFT", "resolution": "badge -> commands-29-green + new test-runner.sh commands-badge guard" },
+    { "severity": "MAJOR", "title": "Router self-contradicts on competitive gating", "file": "commands/linkedin.md", "line": 140, "rule_key": "INTERNAL_CONTRADICTION", "resolution": "Ask-the-User Grow line competitive ⚿ -> competitive; 0 residual hits" },
+    { "severity": "MAJOR", "title": "STATE.md binding count block stale at 27 / v4.0.0", "file": "STATE.md", "line": 46, "rule_key": "SUCCESS_CRITERION_UNMET", "resolution": "updated to 29 / v4.1.0 in session-close overwrite (gitignored, not pushed)" }
+  ],
   "dropped_findings": []
 }
 ```