open/ktg-plugin-marketplace
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity Actions

feat(linkedin-studio): v3.1.0 — Endring 9 adversarial review-pakke + per-artefakt personas

Browse source 
Cold, adversarial review package for the long-form pipeline + configurable
per-edition personas. Motivated by Del 4 (Security Champions pivot): the
in-session editor + persona sweep shared the drafting session's framing-bias,
so the shipped version was never independently re-reviewed.

Headless package (9a/9b):
- New Step 6.5 (headless-review) in /linkedin:newsletter, after the persona
  sweep, before lock — the independence layer the in-session gates can't be.
- New standalone /linkedin:headless-review command (run in a fresh session for
  maximum isolation; reconstructs frozen draft + contract + personas from disk).
- 3 new Opus archetypes, each with a cardinal context-isolation block that
  refuses drafting-session framing as "context pollution":
  - content-reviewer (argument integrity C1–C5, ≤8 flags)
  - language-reviewer (Norwegian language L1–L5, ≤10 flags)
  - fact-reviewer (cold re-verification F1–F4, risk-sort + pivot-risk, WebSearch)
- Deliberate redundancy with fact-checker / editorial-reviewer documented so
  the pairs are never de-duplicated.

Pivot-reopen (9c):
- New /linkedin:pivot command: logs articles.NN.pivots[], resets currentPhase,
  un-locks, marks gates to re-run.
- Pivot-detection gate in Step 8 lock precondition (>20% word-count change or
  >2 new sections re-opens cleared gates). Del 4 v8→v11 worked example.

Per-artifact personas (new requirement):
- articles.NN.personas with resolution order (edition-state → series file →
  plugin library → interactive). One or more readers configurable per edition.

Schema/docs:
- edition-state.template.json: additive personas[], pivots[], headlessReview,
  headless-review phase (16 phases); personaSweep.resonance.wordCount baseline.
- 3 fasit fixtures + 3 structural lint tests (Del 4 worked cases).
- Counts: 24→26 commands, 16→19 agents, 15→16 newsletter phases.
- README + CLAUDE.md (plugin + root) + CHANGELOG synced.

Verification: 35 agent-fixture + 59 hook + 20 render tests green. Backward-
compatible (additive state); reload required before the 3 new agents resolve.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
Kjell Tore Guttormsen 2026-05-29 13:01:24 +02:00
commit e69ea1f4c9
20 changed files with 2520 additions and 59 deletions
Download patch file Download diff file Expand all files Collapse all files

2
CLAUDE.md
View file

@ -9,7 +9,7 @@ plugins/
ai-psychosis/ v1.0.0 — Interaction awareness (sycophancy, reinforcement loops)
config-audit/ v3.1.0 — Configuration intelligence (health, opportunities, auto-fix, whats-active)
graceful-handoff/ v2.1.0 — Auto-trigger handoff via Stop hook (skill + JSON pipeline + 4-step model-aware context resolution)
linkedin-studio/ v3.0.0 — Full-spectrum LinkedIn content engine (short-form feed + long-form newsletter). **v3.0.0 renamed from `linkedin-thought-leadership`** (LinkedIn Thought Leadership → LinkedIn Studio): slug + agent namespace (`linkedin-studio:<agent>`) + runtime state path (`~/.claude/linkedin-studio.local.md`) all change; the `/linkedin:*` commands are unchanged (frontmatter-namespaced, slug-independent). Breaking — reinstall required; functionality byte-identical to v2.4.0. v2.0.0 consolidated surface (27→24 commands, 16→14 agents) + added `/linkedin:newsletter` orchestrator with fact-check + persona-sweep gates BEFORE lock. v2.1.0 added skeleton-gate BEFORE prose (Step 2.5 + Step 3a) + third `persona-reviewer` mode (`skjelett`); pipeline 11→13 phases. v2.2.0 hardened the longform gates (2nd production run): blocking persona hard-fails, fact-check post-cutoff web-search mandate + orthogonal-to-narrative rule, new `voice-scrubber` agent (Opus, de-AI + Norwegian-chronicle voice; gold standard = approved Norwegian editions NOT English post corpus), render+annotate operator gates (2.5/3a), edition-state reconciled with STATE.md (`edition-HANDOVER.md` deleted); agents 14→15. v2.3.0 added **Step 7.5 (visual-assets)** to `/linkedin:newsletter` — cover (+ inline figures) or carousel deck, generated (default mcp-image; external `cover-raw.png` accepted) + operator-gated via `SendUserFile` BEFORE lock so `build-linkedin.mjs` picks up `cover.png` without a post-lock re-render; pipeline 13→14 phases, new `config/image-credit-caption.template.md`, additive `visualAssets` state. v2.4.0 adds **Step 5.5 (editorial-review)** to `/linkedin:newsletter` — new `editorial-reviewer` agent (Opus) judging craft (prosa-håndverk + narrativ-arkitektur), not reader-response, ≤10 flags BLOCK/REWORK/NICE as direction, operator-gated via `SendUserFile` between fact-check (5) and persona-sweep (6), mirroring Maskinrommet skrivekontrakt §C2; motivated by Del 4 (every persona PASS yet 8 fresh editor points, ~6/8 craft/architecture blind spots). Pipeline 14→15 phases; agents 15→16; additive `editorialReview` state; doc/orchestration-only (new agent + fasit fixture + lint test the only new files). Commands unchanged (24); agents 16. Render pipeline self-hosted (OFL-1.1 fonts).
linkedin-studio/ v3.1.0 — Full-spectrum LinkedIn content engine (short-form feed + long-form newsletter). **v3.1.0 adds a cold adversarial review package (Endring 9)** to the long-form pipeline: three new headless archetypes (all Opus) — `content-reviewer` (argument integrity C1–C5, ≤8 flags), `language-reviewer` (Norwegian L1–L5, ≤10 flags), `fact-reviewer` (cold re-verification F1–F4 + pivot-risk, web search) — that re-review a frozen draft with NO drafting-session context (each refuses drafting framing as "context pollution"); new **Step 6.5 (headless-review)** in `/linkedin:newsletter` after the persona sweep, before lock (the independence layer the in-session gates can't be); standalone **`/linkedin:headless-review`** command (run in a fresh session for max isolation); **`/linkedin:pivot`** command + pivot-detection gate (>20 % word-count / >2 new sections re-opens cleared gates before lock); **per-artifact personas** (`articles.NN.personas`). Motivated by Del 4 (Security Champions pivot): the in-session editor + persona sweep shared the drafting session's framing-bias, so the shipped version was never independently re-reviewed. Pipeline 15→16 phases; 24→26 commands; 16→19 agents; additive `personas`/`pivots`/`headlessReview` state; backward-compatible, reload required for the new agents. **v3.0.0 renamed from `linkedin-thought-leadership`** (LinkedIn Thought Leadership → LinkedIn Studio): slug + agent namespace (`linkedin-studio:<agent>`) + runtime state path (`~/.claude/linkedin-studio.local.md`) all change; the `/linkedin:*` commands are unchanged (frontmatter-namespaced, slug-independent). Breaking — reinstall required; functionality byte-identical to v2.4.0. v2.0.0 consolidated surface (27→24 commands, 16→14 agents) + added `/linkedin:newsletter` orchestrator with fact-check + persona-sweep gates BEFORE lock. v2.1.0 added skeleton-gate BEFORE prose (Step 2.5 + Step 3a) + third `persona-reviewer` mode (`skjelett`); pipeline 11→13 phases. v2.2.0 hardened the longform gates (2nd production run): blocking persona hard-fails, fact-check post-cutoff web-search mandate + orthogonal-to-narrative rule, new `voice-scrubber` agent (Opus, de-AI + Norwegian-chronicle voice; gold standard = approved Norwegian editions NOT English post corpus), render+annotate operator gates (2.5/3a), edition-state reconciled with STATE.md (`edition-HANDOVER.md` deleted); agents 14→15. v2.3.0 added **Step 7.5 (visual-assets)** to `/linkedin:newsletter` — cover (+ inline figures) or carousel deck, generated (default mcp-image; external `cover-raw.png` accepted) + operator-gated via `SendUserFile` BEFORE lock so `build-linkedin.mjs` picks up `cover.png` without a post-lock re-render; pipeline 13→14 phases, new `config/image-credit-caption.template.md`, additive `visualAssets` state. v2.4.0 adds **Step 5.5 (editorial-review)** to `/linkedin:newsletter` — new `editorial-reviewer` agent (Opus) judging craft (prosa-håndverk + narrativ-arkitektur), not reader-response, ≤10 flags BLOCK/REWORK/NICE as direction, operator-gated via `SendUserFile` between fact-check (5) and persona-sweep (6), mirroring Maskinrommet skrivekontrakt §C2; motivated by Del 4 (every persona PASS yet 8 fresh editor points, ~6/8 craft/architecture blind spots). Pipeline 14→15 phases; agents 15→16; additive `editorialReview` state; doc/orchestration-only (new agent + fasit fixture + lint test the only new files). Commands unchanged (24); agents 16. Render pipeline self-hosted (OFL-1.1 fonts).
llm-security/ v7.7.2 — Security scanning, auditing, threat modeling. HTML report output for all 18 skill commands (render-report CLI + canonical ESM module mirrored bit-identical into the playground). v7.7.2 translated the remaining Norwegian surface text in the playground UI, the canonical renderer, the agent prompts, and the README/CLAUDE.md state sections to English. v7.7.1 stripped the playground to the catalog as the only routable surface.
ms-ai-architect/ v1.15.0 — Microsoft AI architecture (Cosmo Skyberg persona) + manual KB-refresh slash command + v3 project-view (sidebar med 17 artifacts + main + import-modal overlay, v2-surface fjernet i v1.15.0)
okr/ v1.0.0 — OKR guidance for Norwegian public sector

4
README.md
View file

@ -206,11 +206,11 @@ Key commands: `/architect`, `/architect:ros`, `/architect:security`, `/architect
---
### [LinkedIn Studio](plugins/linkedin-studio/) `v3.0.0`
### [LinkedIn Studio](plugins/linkedin-studio/) `v3.1.0`
Build authentic LinkedIn authority through algorithmic understanding, strategic consistency, and AI-assisted content creation.
**v3.0.0 renames the plugin to LinkedIn Studio** (was `linkedin-thought-leadership`): the slug, agent namespace, and runtime state-file path are now `linkedin-studio` — the `/linkedin:*` commands are unchanged. Breaking: reinstall required and the state file moves to `~/.claude/linkedin-studio.local.md`. Functionality is identical to v2.4.0. v2.4.0 adds an **editor's craft gate as an explicit pipeline phase** in `/linkedin:newsletter`: a new **`editorial-reviewer` agent** (Opus) + **Step 5.5 (editorial-review)** between fact-check (Step 5) and the persona sweep (Step 6). It judges **craft** — **prosa-håndverk** (em-dash density, verbatim repetition, postulated numbers, contradictions, versal-tic) + **narrativ-arkitektur** (concrete instantiation, theory-anchored hypotheses, series-title symmetry, equal action per addressee, un-overloaded conclusion) — not reader-response, returning **≤10 flags** (BLOCK / REWORK / NICE) as *direction* (never copy), operator-gated via `SendUserFile`, mirroring the Maskinrommet writing-contract §C2. Motivated by the Del 4 run, where **every persona reported PASS yet the editor found 8 fresh points on first reading, ~6/8 of them craft/architecture blind spots no agent measured.** Pipeline 14 → 15 phases; 15 → 16 agents; additive `editorialReview` state. v2.3.0 made **visual assets an explicit pipeline phase** (Step 7.5, between annotation and lock): the cover (+ optional inline figures) or a carousel deck is generated (default `mcp-image`; external `cover-raw.png` accepted) and operator-gated via `SendUserFile` **before lock**. v2.2.0 **hardened the longform gates** (blocking persona hard-fails, post-cutoff fact-check mandate, a Norwegian-chronicle `voice-scrubber`, render+annotate operator gates, STATE.md-reconciled edition state). Commands unchanged (24); 16 agents. v2.1.0's skeleton-gate-before-prose and v2.0.0's full-spectrum surface are preserved. Updated for the January 2026 360Brew algorithm change, which validates your creator profile before distributing content.
**v3.1.0 adds a cold adversarial review package to the long-form pipeline (Endring 9)** — three new headless archetypes (`content-reviewer` argument, `language-reviewer` Norwegian, `fact-reviewer` cold re-verification + pivot-risk), all Opus, that re-review a frozen draft with NO drafting-session context; a new **Step 6.5 (headless-review)** in `/linkedin:newsletter` after the persona sweep and before lock; a standalone **`/linkedin:headless-review`** command (run in a fresh session for maximum isolation); a **`/linkedin:pivot`** command + pivot-detection gate (> 20 % word-count / > 2 new sections re-opens cleared gates before lock); and **per-artifact personas** configurable per edition. Motivated by the Del 4 Security Champions pivot, where the in-session editor + persona sweep shared the drafting session's framing-bias so the shipped version was never independently re-reviewed. Pipeline 15 → 16 phases; 24 → 26 commands; 16 → 19 agents; backward-compatible (additive state); reload required for the new agents. **v3.0.0 renames the plugin to LinkedIn Studio** (was `linkedin-thought-leadership`): the slug, agent namespace, and runtime state-file path are now `linkedin-studio` — the `/linkedin:*` commands are unchanged. Breaking: reinstall required and the state file moves to `~/.claude/linkedin-studio.local.md`. Functionality is identical to v2.4.0. v2.4.0 adds an **editor's craft gate as an explicit pipeline phase** in `/linkedin:newsletter`: a new **`editorial-reviewer` agent** (Opus) + **Step 5.5 (editorial-review)** between fact-check (Step 5) and the persona sweep (Step 6). It judges **craft** — **prosa-håndverk** (em-dash density, verbatim repetition, postulated numbers, contradictions, versal-tic) + **narrativ-arkitektur** (concrete instantiation, theory-anchored hypotheses, series-title symmetry, equal action per addressee, un-overloaded conclusion) — not reader-response, returning **≤10 flags** (BLOCK / REWORK / NICE) as *direction* (never copy), operator-gated via `SendUserFile`, mirroring the Maskinrommet writing-contract §C2. Motivated by the Del 4 run, where **every persona reported PASS yet the editor found 8 fresh points on first reading, ~6/8 of them craft/architecture blind spots no agent measured.** Pipeline 14 → 15 phases; 15 → 16 agents; additive `editorialReview` state. v2.3.0 made **visual assets an explicit pipeline phase** (Step 7.5, between annotation and lock): the cover (+ optional inline figures) or a carousel deck is generated (default `mcp-image`; external `cover-raw.png` accepted) and operator-gated via `SendUserFile` **before lock**. v2.2.0 **hardened the longform gates** (blocking persona hard-fails, post-cutoff fact-check mandate, a Norwegian-chronicle `voice-scrubber`, render+annotate operator gates, STATE.md-reconciled edition state). Commands unchanged (24); 16 agents. v2.1.0's skeleton-gate-before-prose and v2.0.0's full-spectrum surface are preserved. Updated for the January 2026 360Brew algorithm change, which validates your creator profile before distributing content.
- **Editorial-review craft gate BEFORE the persona sweep (v2.4)** — Step 5.5 runs the new `editorial-reviewer` (Opus) on the fact-checked draft, two axes (prosa-håndverk + narrativ-arkitektur), ≤10 flags severity-ranked BLOCK/REWORK/NICE, surfaced to the operator via `SendUserFile`. It runs *before* the persona sweep so the personas measure resonance instead of stumbling on craft noise; `persona-reviewer` is unchanged (one measures craft, one measures response). Mirrors Maskinrommet §C2.
- **Visual-assets gate BEFORE lock (v2.3)** — Step 7.5 decides image needs from the article type (method-heavy → 1–2 inline figures; diagnosis-heavy → cover only), briefs each image, generates via mcp-image or an external `cover-raw.png`, surfaces candidates with `SendUserFile`, and copies the approved one to the fixed `cover.png` name. Explicit `format: "carousel"` branch reusing `build-carousel.mjs`.

4
plugins/linkedin-studio/.claude-plugin/plugin.json
View file

@ -1,7 +1,7 @@
{
"name": "linkedin-studio",
"version": "3.0.0",
"description": "LinkedIn Studio — full-spectrum LinkedIn content engine: feed posts, carousels, video scripts, and long-form newsletter editions, with the January 2026 360Brew algorithm baked in. v3.0.0 renames the plugin (was `linkedin-thought-leadership`): the slug, agent namespace, and runtime state-file path are now `linkedin-studio` — the `/linkedin:*` commands are unchanged. Breaking: reinstall required and the state file moves to `~/.claude/linkedin-studio.local.md`.",
"version": "3.1.0",
"description": "LinkedIn Studio — full-spectrum LinkedIn content engine: feed posts, carousels, video scripts, and long-form newsletter editions, with the January 2026 360Brew algorithm baked in. v3.1.0 (Endring 9) adds a cold adversarial review package to `/linkedin:newsletter` — Step 6.5 + the standalone `/linkedin:headless-review` command run three new headless archetypes (content-reviewer, language-reviewer, fact-reviewer) plus the persona reviewer with NO drafting-session context — a `/linkedin:pivot` command that re-opens cleared gates after a late change, and per-artifact personas (one or more readers configurable per edition). v3.0.0 renamed the plugin (was `linkedin-thought-leadership`): slug, agent namespace, and state-file path are `linkedin-studio`; the `/linkedin:*` commands are unchanged.",
"author": {
"name": "Kjell Tore Guttormsen"
},

24
plugins/linkedin-studio/CHANGELOG.md
View file

@ -5,6 +5,30 @@ All notable changes to this project will be documented in this file.
The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
## [3.1.0] - 2026-05-29
### Summary
**An adversarial review package becomes part of the long-form pipeline (Endring 9).** The Del 4 production run (Maskinrommet, 2026-05-29 — the Security Champions pivot, v8 → v11) shipped a high-quality article, but its quality assurance leaned on **one good editor + KTG's availability to read it several times** — it did not scale. The root cause: the editor and the persona sweep ran *in the same session as drafting*, sharing the conversation history (which versions passed, what was deliberately cut, which flags had been raised). They were therefore **not adversarial** — they carried framing-bias. Three concrete symptoms: (1) the persona resonance sweep effectively judged an early version, not the one that shipped; (2) editor-approval was single-source; (3) the fact-check was post-hoc relative to the late pivot, so the pivot could build on an unverified premise. v3.1.0 answers KTG's explicit question — *how do I start sessions with no context from the main session, to review both content and language?* — and makes per-artifact personas a first-class input.
### Added
- **`/linkedin:headless-review` command** — runs the cold adversarial review package on a FROZEN draft. Designed to be invoked in a **fresh session** for maximum isolation (the parent then has no drafting transcript); reconstructs everything from disk (frozen draft + writing contract + personas). Flags `--draft`, `--type {content|language|fact|persona-resonance|persona-conversion|all}`, `--persona`, `--article`, `--output`.
- **Three new headless review archetypes** (all Opus, each with an explicit cardinal context-isolation block that refuses drafting-session framing as "context pollution"):
- **`content-reviewer`** (color maroon, Read+Grep) — argument integrity: C1 logical holes · C2 unsupported assumptions · C3 argument-level contradiction · C4 missing concretization · C5 unanswered «what about X?». ≤8 flags BLOCK/REWORK/NICE.
- **`language-reviewer`** (navy, Read+Grep) — Norwegian language: L1 verbatim repetition · L2 anglicisms · L3 stiff bureaucratic register · L4 language-level self-contradiction · L5 clang/rhythm. ≤10 flags. Deliberate cold re-take of `editorial-reviewer`'s prose axis.
- **`fact-reviewer`** (gold, Read+WebSearch) — cold re-verification on the frozen/pivoted version: F1 verifiable claims · F2 quote precision · F3 number attribution · F4 source quality. Carries over `fact-checker`'s 5-dimension scoring + 🔴/🟡/🟢 sort + contradiction sweep + post-cutoff mandate, adds a **pivot-risk** subsection. Deliberate redundancy with `fact-checker` to catch a pivot premise that arrived after Step 5.
- **Step 6.5 (headless-review)** in `/linkedin:newsletter` — fans the package out in parallel after the in-session persona sweep (Step 6), on a frozen draft snapshot, BEFORE lock; consolidated report surfaced via `SendUserFile`; converged flags (two independent cold reviewers agreeing) marked as the strongest signal. Pipeline 15 → 16 phases.
- **`/linkedin:pivot` command** + **pivot-detection gate** — a pivot re-opens the pipeline so the cleared gates (fact-check 5 → editorial 5.5 → persona 6 → headless 6.5) re-run on the changed version before lock. Heuristic: a draft that drifted **> 20 % in word count OR gained > 2 sections** since Step 6 cleared triggers the gate (enforced as a Step 8 lock precondition). Worked example: the Del 4 v8→v11 run (+42 %, 2 new sections) would have fired the gate and forced the re-sweep.
- **Per-artifact personas** — `articles.NN.personas` in `edition-state.json`: one or more readers configurable **per edition**, resolved in Step 1 in order (edition-state → `<serie>/linkedin/personas.md` per-series file → plugin `personas.local.md`/template → interactive definition). Feeds both the Step 6 sweep and the Step 6.5 package. `config/personas.template.md` documents the resolution order.
- **Three fasit fixtures + three structural lint tests** for the new agents (Del 4 / Security Champions worked cases), mirroring the `editorial-reviewer` fixture discipline. All 35 agent-fixture assertions green.
### Changed
- `config/edition-state.template.json` — additive: per-article `personas[]`, `pivots[]`, `headlessReview` object; new `headless-review` phase string (16 phases total); `personaSweep.resonance.wordCount` recorded at Step 6 as the pivot-detection baseline.
- `commands/newsletter.md` — Step 0/1 persona resolution reworked to per-artifact; new Step 6.5; Step 8 lock preconditions add the headless gate + pivot-detection gate; pipeline + resumption tables updated (`persona-sweep-prelock` → resume at Step 6.5; new `headless-review` → Step 7).
- Counts: 24 → 26 commands; 16 → 19 agents; 15 → 16 newsletter phases.
### Compatibility
Backward-compatible: every state-shape change is additive (existing editions resume by `currentPhase`; `persona-sweep-prelock` now resumes at Step 6.5 — an intended deterministic improvement). **Reload required** before the three new agents resolve (the plugin agent set is built at session start). No new runtime code beyond the agents/commands/fixtures; the render pipeline, hooks, and short-form surface are untouched.
## [3.0.0] - 2026-05-29
### Summary

17
plugins/linkedin-studio/CLAUDE.md
View file

@ -1,6 +1,6 @@
# LinkedIn Studio Plugin (v2.4.0)
# LinkedIn Studio Plugin (v3.1.0)
Full-spectrum LinkedIn content engine — short-form feed posts, carousels, video scripts, and long-form newsletter editions — with the January 2026 360Brew algorithm baked in. v2.0.0 consolidated the surface (27 commands → 24, 16 agents → 14) while adding the long-form `/linkedin:newsletter` orchestrator + two longform-quality gate agents (`fact-checker`, `persona-reviewer`). v2.1.0 added two gates BEFORE prose (Step 2.5 skeleton + Step 3a spine prose) + a third `persona-reviewer` mode (`skjelett`). v2.2.0 hardened the longform gates with the lessons from the second production run (Seres-serien): blocking persona hard-fails, a post-cutoff fact-check mandate, a `voice-scrubber` agent, render+annotate operator gates, and STATE.md-reconciled edition state. v2.3.0 made **visual assets an explicit pipeline phase** — Step 7.5 (visual-assets) between annotation (Step 7) and lock (Step 8): cover (+ optional inline figures) or a carousel deck, generated (default `mcp-image`, external `cover-raw.png` accepted) and operator-gated BEFORE lock so `render/build-linkedin.mjs` picks up `cover.png` at lock without a post-lock re-render. **v2.4.0** makes an **editor's craft gate an explicit pipeline phase** — new **Step 5.5 (editorial-review)** between fact-check (Step 5) and the persona sweep (Step 6): a new **`editorial-reviewer` agent** (Opus) judges **craft** (prosa-håndverk + narrativ-arkitektur), not reader-response, returning ≤10 flags (BLOCK/REWORK/NICE) as direction, **operator-gated via `SendUserFile` BEFORE the persona sweep** so the personas measure resonance instead of stumbling on craft noise. Motivated by Del 4: every persona reported PASS, yet the editor found 8 fresh points on first reading, ~6/8 of them craft/architecture blind spots no agent measured. Mirrors the Maskinrommet writing-contract §C2. Pipeline 14 → 15 phases; agents 15 → 16; additive `editorialReview` state. Doc/orchestration-only for the wiring (the new agent + its fasit fixture + lint test are the only new files); commands unchanged (24).
Full-spectrum LinkedIn content engine — short-form feed posts, carousels, video scripts, and long-form newsletter editions — with the January 2026 360Brew algorithm baked in. v2.0.0 consolidated the surface (27 commands → 24, 16 agents → 14) while adding the long-form `/linkedin:newsletter` orchestrator + two longform-quality gate agents (`fact-checker`, `persona-reviewer`). v2.1.0 added two gates BEFORE prose (Step 2.5 skeleton + Step 3a spine prose) + a third `persona-reviewer` mode (`skjelett`). v2.2.0 hardened the longform gates with the lessons from the second production run (Seres-serien): blocking persona hard-fails, a post-cutoff fact-check mandate, a `voice-scrubber` agent, render+annotate operator gates, and STATE.md-reconciled edition state. v2.3.0 made **visual assets an explicit pipeline phase** — Step 7.5 (visual-assets) between annotation (Step 7) and lock (Step 8): cover (+ optional inline figures) or a carousel deck, generated (default `mcp-image`, external `cover-raw.png` accepted) and operator-gated BEFORE lock so `render/build-linkedin.mjs` picks up `cover.png` at lock without a post-lock re-render. **v2.4.0** makes an **editor's craft gate an explicit pipeline phase** — new **Step 5.5 (editorial-review)** between fact-check (Step 5) and the persona sweep (Step 6): a new **`editorial-reviewer` agent** (Opus) judges **craft** (prosa-håndverk + narrativ-arkitektur), not reader-response, returning ≤10 flags (BLOCK/REWORK/NICE) as direction, **operator-gated via `SendUserFile` BEFORE the persona sweep** so the personas measure resonance instead of stumbling on craft noise. Motivated by Del 4: every persona reported PASS, yet the editor found 8 fresh points on first reading, ~6/8 of them craft/architecture blind spots no agent measured. Mirrors the Maskinrommet writing-contract §C2. Pipeline 14 → 15 phases; agents 15 → 16; additive `editorialReview` state. Doc/orchestration-only for the wiring (the new agent + its fasit fixture + lint test are the only new files); commands unchanged (24). **v3.1.0 (Endring 9)** adds an **adversarial review package** run COLD on a frozen draft — new **Step 6.5 (headless-review)** between the persona sweep (Step 6) and lock, plus a standalone **`/linkedin:headless-review`** command (run in a fresh session for maximum isolation): three new headless archetypes — **`content-reviewer`** (argument integrity), **`language-reviewer`** (Norwegian language), **`fact-reviewer`** (cold re-verification incl. claims a late pivot bolted on) — plus `persona-reviewer` in resonance + conversion modes, all with NO drafting-session context (the independence layer the in-session gates structurally cannot be). v3.1.0 also adds **`/linkedin:pivot`** (re-opens cleared gates after a late change + a >20 %/>2-section pivot-detection gate at lock) and **per-artifact personas** (`articles.NN.personas` — one or more readers configurable per edition, resolved edition-state → series file → plugin library → interactive). Pipeline 15 → 16 phases; agents 16 → 19; commands 24 → 26; additive `personas` / `pivots` / `headlessReview` state. Motivated by Del 4: the in-session editor + persona sweep shared the drafting session's framing-bias, so the version that shipped was never independently re-reviewed.
## Architecture
@ -32,9 +32,9 @@ Full-spectrum LinkedIn content engine — short-form feed posts, carousels, vide
**Hook editing:** Edit `hooks/hooks.template.json` + `hooks/prompts/*.md`, then run `python3 hooks/scripts/compile-hooks.py`. Do not edit `hooks.json` directly. Prompts are loaded at runtime by gatekeeper scripts; the compile step is only needed when adding `type: prompt` hooks.
## Commands (24)
## Commands (26)
All content commands (post, quick, react, pipeline, first-post, video, multiplatform, carousel, newsletter) auto-copy output to clipboard via `clipboard-helper.mjs`. Interactive steps are minimized — angle, format, and post type are inferred from context, with max 2 questions per post. **v2.0.0 net change:** 5 commands removed (`templates`, `publish`, `authority`, `collab`, `speaking` — absorbed into `quick`, `calendar`, `strategy`, `outreach` respectively) + 2 commands added (`newsletter`, `outreach`) = 27 → 24.
All content commands (post, quick, react, pipeline, first-post, video, multiplatform, carousel, newsletter) auto-copy output to clipboard via `clipboard-helper.mjs`. Interactive steps are minimized — angle, format, and post type are inferred from context, with max 2 questions per post. **v2.0.0 net change:** 5 commands removed (`templates`, `publish`, `authority`, `collab`, `speaking` — absorbed into `quick`, `calendar`, `strategy`, `outreach` respectively) + 2 commands added (`newsletter`, `outreach`) = 27 → 24. **v3.1.0** adds 2 longform companions (`headless-review`, `pivot`) = 24 → 26.
| Command | Purpose |
|---------|---------|
@ -46,7 +46,9 @@ All content commands (post, quick, react, pipeline, first-post, video, multiplat
| `/linkedin:post` | Full post creation (10-15 min) |
| `/linkedin:quick` | 5-minute quick post (3-line formula) + 8 post-type templates |
| `/linkedin:pipeline` | Full end-to-end content pipeline |
| `/linkedin:newsletter` | Long-form orchestrator: newsletter edition / essay / series article — multi-session 15-phase pipeline with **skeleton + spine-prose gates BEFORE prose (v2.1)**, **editorial-review craft gate BEFORE the persona sweep (Step 5.5, v2.4)**, fact-check + persona-sweep BEFORE lock, and **visual-assets gate BEFORE lock (Step 7.5, v2.3)** |
| `/linkedin:newsletter` | Long-form orchestrator: newsletter edition / essay / series article — multi-session 16-phase pipeline with **skeleton + spine-prose gates BEFORE prose (v2.1)**, **editorial-review craft gate BEFORE the persona sweep (Step 5.5, v2.4)**, fact-check + persona-sweep BEFORE lock, **headless adversarial review BEFORE lock (Step 6.5, v3.1)**, and **visual-assets gate BEFORE lock (Step 7.5, v2.3)** |
| `/linkedin:headless-review` | **(v3.1)** Cold adversarial review package — run the 3 headless archetypes (`content-reviewer`, `language-reviewer`, `fact-reviewer`) + `persona-reviewer` (resonance/conversion) on a FROZEN draft with no drafting-session context; consolidated, operator-gated report. Step 6.5's standalone surface (run in a fresh session for maximum isolation) |
| `/linkedin:pivot` | **(v3.1)** Re-open a long-form edition after a late substantive change so cleared gates (fact-check → editorial → persona → headless) re-run before lock; logs `pivots[]`, resets `currentPhase`, un-locks if needed (pivot heuristic: >20 % word-count change or >2 new sections) |
| `/linkedin:batch` | Create a full week of content |
| `/linkedin:calendar` | View/manage post scheduling queue + publish action (mark scheduled posts as published) |
| `/linkedin:carousel` | Structured multi-slide carousel generator |
@ -63,7 +65,7 @@ All content commands (post, quick, react, pipeline, first-post, video, multiplat
| `/linkedin:outreach` | Outreach orchestrator — collaborations + speaking opportunities (unlocks at ~1K followers) |
| `/linkedin:profile` | 360Brew profile optimization |
## Agents (16)
## Agents (19)
| Agent | Model | Color | Responsibility |
|-------|-------|-------|----------------|
@ -83,6 +85,9 @@ All content commands (post, quick, react, pipeline, first-post, video, multiplat
| `editorial-reviewer` | Opus | Orange | Editor's craft gate (v2.4, Step 5.5, before persona sweep): prosa-håndverk + narrativ-arkitektur, ≤10 flags BLOCK/REWORK/NICE as direction, operator-gated via `SendUserFile`; mirrors Maskinrommet §C2 (longform) |
| `persona-reviewer` | Opus | Olive | Reader-persona skeleton (v2.1, before prose) + resonance (before lock) + hook-conversion (after lock) gate, blocking hard-fail list (longform) |
| `voice-scrubber` | Opus | Red | De-AI scrub + Norwegian-chronicle voice-drift correction; gold standard = approved Norwegian editions, not the English post corpus (longform, v2.2) |
| `content-reviewer` | Opus | Maroon | **(v3.1, Step 6.5 — cold/headless)** Argument-integrity review on a frozen draft: C1 logical holes · C2 unsupported assumptions · C3 argument contradiction · C4 missing concretization · C5 unanswered objection. ≤8 flags BLOCK/REWORK/NICE as direction; refuses drafting-session framing as context pollution (longform) |
| `language-reviewer` | Opus | Navy | **(v3.1, Step 6.5 — cold/headless)** Norwegian-language review on a frozen draft: L1 verbatim repetition · L2 anglicisms · L3 stiff bureaucratic register · L4 language-level self-contradiction · L5 clang/rhythm. ≤10 flags BLOCK/REWORK/NICE; deliberate cold re-take of editorial's prose axis (longform) |
| `fact-reviewer` | Opus | Gold | **(v3.1, Step 6.5 — cold/headless)** Cold re-verification on the frozen/pivoted version (web search): F1 verifiable claims · F2 quote precision · F3 number attribution · F4 source quality. 🔴/🟡/🟢 + pivot-risk subsection; deliberate redundancy with `fact-checker` to catch a pivot premise that arrived after Step 5 (longform) |
**Rule:** Always read `assets/voice-samples/` before generating content.

33
plugins/linkedin-studio/README.md
View file

@ -6,20 +6,21 @@
*AI-generated: all code produced by Claude Code through dialog-driven development. [Full disclosure →](../../README.md#ai-generated-code-disclosure)*
![Version](https://img.shields.io/badge/version-3.0.0-blue)
![Version](https://img.shields.io/badge/version-3.1.0-blue)
![Platform](https://img.shields.io/badge/platform-Claude_Code_Plugin-purple)
![Commands](https://img.shields.io/badge/commands-24-green)
![Agents](https://img.shields.io/badge/agents-16-orange)
![Commands](https://img.shields.io/badge/commands-26-green)
![Agents](https://img.shields.io/badge/agents-19-orange)
![Hooks](https://img.shields.io/badge/hooks-9-red)
![Reference Docs](https://img.shields.io/badge/reference_docs-24-teal)
![License](https://img.shields.io/badge/license-MIT-lightgrey)
A comprehensive Claude Code plugin that turns LinkedIn from a chore into a full-spectrum content engine — short-form feed posts, carousels, video scripts, and long-form newsletter editions. v2.0.0 consolidated the surface (27 → 24 commands, 16 → 14 agents) and added `/linkedin:newsletter` as a multi-session long-form orchestrator with fact-check + persona-sweep gates BEFORE lock. **v2.1.0** adds two more gates BEFORE prose to `/linkedin:newsletter` — a skeleton gate (Step 2.5) and a spine-prose gate (Step 3a) — encoding the Maskinrommet writing-contract §A discipline (premiss / problem / anbefaling / gevinst / vei videre) into the pipeline itself, so spine errors get caught in minutes at the skeleton stage instead of hours at the resonance stage or a full day post-lock. v2.2.0 hardened the longform gates with the lessons from the next production run: blocking persona hard-fails (primær «mistet meg» / doesn't own the action / jargon wall / model-name catalog → BLOCK), a post-cutoff fact-check mandate, a new Norwegian-chronicle de-AI voice-scrubber agent, render+annotate operator gates, and edition state reconciled with the global STATE.md continuity system. v2.3.0 makes visual assets an explicit pipeline phase — **Step 7.5 (visual-assets)** between annotation and lock: the cover (+ optional inline figures) or a carousel deck is generated and operator-gated *before* lock, so the renderer picks up `cover.png` without a post-lock re-render. **v2.4.0** adds an **editor's craft gate** — new **Step 5.5 (editorial-review)** between fact-check and the persona sweep: a new `editorial-reviewer` agent (Opus) judges *craft* (prose-craft + narrative-architecture), not reader-response, returning ≤10 flags (BLOCK / REWORK / NICE) operator-gated via `SendUserFile` so the personas measure resonance instead of stumbling on craft noise (pipeline 14 → 15 phases). 24 slash commands, 16 specialized agents, 9 automated hooks, and a 24-document knowledge base grounded in LinkedIn's actual algorithm signals. Updated for the January 2026 **360Brew** algorithm change, where LinkedIn now validates your profile before distributing content.
A comprehensive Claude Code plugin that turns LinkedIn from a chore into a full-spectrum content engine — short-form feed posts, carousels, video scripts, and long-form newsletter editions. v2.0.0 consolidated the surface (27 → 24 commands, 16 → 14 agents) and added `/linkedin:newsletter` as a multi-session long-form orchestrator with fact-check + persona-sweep gates BEFORE lock. **v2.1.0** adds two more gates BEFORE prose to `/linkedin:newsletter` — a skeleton gate (Step 2.5) and a spine-prose gate (Step 3a) — encoding the Maskinrommet writing-contract §A discipline (premiss / problem / anbefaling / gevinst / vei videre) into the pipeline itself, so spine errors get caught in minutes at the skeleton stage instead of hours at the resonance stage or a full day post-lock. v2.2.0 hardened the longform gates with the lessons from the next production run: blocking persona hard-fails (primær «mistet meg» / doesn't own the action / jargon wall / model-name catalog → BLOCK), a post-cutoff fact-check mandate, a new Norwegian-chronicle de-AI voice-scrubber agent, render+annotate operator gates, and edition state reconciled with the global STATE.md continuity system. v2.3.0 makes visual assets an explicit pipeline phase — **Step 7.5 (visual-assets)** between annotation and lock: the cover (+ optional inline figures) or a carousel deck is generated and operator-gated *before* lock, so the renderer picks up `cover.png` without a post-lock re-render. **v2.4.0** adds an **editor's craft gate** — new **Step 5.5 (editorial-review)** between fact-check and the persona sweep: a new `editorial-reviewer` agent (Opus) judges *craft* (prose-craft + narrative-architecture), not reader-response, returning ≤10 flags (BLOCK / REWORK / NICE) operator-gated via `SendUserFile` so the personas measure resonance instead of stumbling on craft noise (pipeline 14 → 15 phases). **v3.1.0** adds a **cold adversarial review package** as **Step 6.5 (headless-review)** in `/linkedin:newsletter` — three new headless archetypes (`content-reviewer`, `language-reviewer`, `fact-reviewer`) that re-review a frozen draft with NO drafting-session context, surfaced standalone via the new `/linkedin:headless-review` command (run in a fresh session for maximum isolation), plus a `/linkedin:pivot` command + pivot-detection gate that re-opens cleared gates after a late change, and per-artifact personas configurable per edition (pipeline 15 → 16 phases). 26 slash commands, 19 specialized agents, 9 automated hooks, and a 24-document knowledge base grounded in LinkedIn's actual algorithm signals. Updated for the January 2026 **360Brew** algorithm change, where LinkedIn now validates your profile before distributing content.
---
## Table of Contents
- [What's New in v3.1.0](#whats-new-in-v310)
- [What's New in v3.0.0](#whats-new-in-v300)
- [What's New in v2.4.0](#whats-new-in-v240)
- [What's New in v2.3.0](#whats-new-in-v230)
@ -44,6 +45,20 @@ A comprehensive Claude Code plugin that turns LinkedIn from a chore into a full-
---
## What's New in v3.1.0
**An adversarial review package becomes part of the long-form pipeline (Endring 9).** The Del 4 production run (the Security Champions pivot) shipped a high-quality article, but its quality assurance leaned on one good editor plus the operator's availability to re-read it several times — it did not scale. The root cause: the editor (Step 5.5) and the persona sweep (Step 6) ran *in the same session as drafting*, sharing the conversation history (which versions passed, what was cut, which flags were raised). They carried the drafting session's **framing-bias** and were therefore not independent. Three symptoms followed: the resonance sweep effectively judged an early version rather than the one that shipped, editor-approval was single-source, and the fact-check was post-hoc relative to a late pivot — so the pivot could build on an unverified premise.
v3.1.0 answers the operator's explicit question — *how do I start sessions with no context from the main session, to review both content and language?* — with a **cold review package** that reconstructs everything from disk and refuses the drafting session's framing as "context pollution".
- **Three new headless review archetypes** (all Opus, cold/headless): **`content-reviewer`** (maroon — argument integrity: C1 logical holes · C2 unsupported assumptions · C3 argument contradiction · C4 missing concretization · C5 unanswered objection; ≤8 flags), **`language-reviewer`** (navy — Norwegian language: L1 verbatim repetition · L2 anglicisms · L3 stiff bureaucratic register · L4 language-level self-contradiction · L5 clang/rhythm; ≤10 flags; a deliberate cold re-take of the editorial prose axis), and **`fact-reviewer`** (gold — cold re-verification on the frozen/pivoted version via web search: F1 verifiable claims · F2 quote precision · F3 number attribution · F4 source quality; 🔴/🟡/🟢 + a pivot-risk subsection; deliberate redundancy with `fact-checker` to catch a pivot premise that arrived after Step 5).
- **New `/linkedin:headless-review` command** — runs the cold package on a FROZEN draft, designed to be invoked in a **fresh session** for maximum isolation (the parent then has no drafting transcript) and reconstructs the draft + writing contract + personas from disk.
- **New Step 6.5 (headless-review)** in `/linkedin:newsletter` — fans the package out in parallel after the in-session persona sweep (Step 6), on a frozen draft snapshot, BEFORE lock; the independence layer the in-session gates cannot be. Converged flags (two independent cold reviewers agreeing) are the strongest signal.
- **New `/linkedin:pivot` command + pivot-detection gate** — a late change re-opens the cleared gates (fact-check → editorial → persona → headless) so they re-run on the changed version before lock. Heuristic: a draft that drifted **> 20 % in word count OR gained > 2 sections** since Step 6 cleared fires the gate (enforced as a Step 8 lock precondition).
- **Per-artifact personas** (`articles.NN.personas`) — one or more readers configurable **per edition**, resolved in order (edition-state → `<serie>/linkedin/personas.md` per-series file → plugin library/template → interactive definition); feeds both the Step 6 sweep and the Step 6.5 package.
Counts: 24 → 26 commands; 16 → 19 agents; newsletter pipeline 15 → 16 phases. **Backward-compatible** — every state-shape change is additive (`personas`/`pivots`/`headlessReview`); existing editions resume by `currentPhase`. **Reload required** before the three new agents resolve (the plugin agent set is built at session start).
## What's New in v3.0.0
**The plugin is renamed `linkedin-thought-leadership` → `linkedin-studio` ("LinkedIn Thought Leadership" → LinkedIn Studio).** The old display title read as pompous; the new name is plain and matches how the plugin already describes itself — a *LinkedIn content engine*. This is a **breaking** release because the identity changes at three levels, but **no functionality changes** — v3.0.0 is byte-for-byte v2.4.0 under a new name.
@ -203,7 +218,7 @@ The wizard handles everything: 360Brew profile checklist, voice and user profile
## Commands
All 24 commands use colon notation: `/linkedin:post`, `/linkedin:quick`, etc.
All 26 commands use colon notation: `/linkedin:post`, `/linkedin:quick`, etc.
### Onboarding
@ -257,7 +272,7 @@ All 24 commands use colon notation: `/linkedin:post`, `/linkedin:quick`, etc.
## Agent Architecture
The plugin delegates specialized work to 16 purpose-built agents. Each agent has its own model assignment, color identity, and focused responsibility.
The plugin delegates specialized work to 19 purpose-built agents. Each agent has its own model assignment, color identity, and focused responsibility.
| Agent | Model | Color | Primary Responsibility |
|-------|-------|-------|----------------------|
@ -277,6 +292,9 @@ The plugin delegates specialized work to 16 purpose-built agents. Each agent has
| `editorial-reviewer` | Opus | Orange | Editor's craft gate (v2.4, Step 5.5, before persona sweep): **prosa-håndverk** + **narrativ-arkitektur**, ≤10 flags BLOCK/REWORK/NICE as direction, operator-gated via `SendUserFile`; mirrors Maskinrommet §C2 (longform) |
| `persona-reviewer` | Opus | Olive | Reader-persona **skeleton** (v2.1, before prose) + **resonance** (before lock) + **hook-conversion** (after lock) gate, blocking hard-fail list (longform) |
| `voice-scrubber` | Opus | Red | De-AI scrub + Norwegian-chronicle voice-drift correction; gold standard = approved Norwegian editions, not the English post corpus (longform, v2.2) |
| `content-reviewer` | Opus | Maroon | (v3.1, Step 6.5 — cold/headless) Argument integrity: C1 logical holes · C2 unsupported assumptions · C3 argument contradiction · C4 missing concretization · C5 unanswered objection; ≤8 flags BLOCK/REWORK/NICE; refuses drafting-session framing as context pollution (longform) |
| `language-reviewer` | Opus | Navy | (v3.1, Step 6.5 — cold/headless) Norwegian language: L1 verbatim repetition · L2 anglicisms · L3 stiff bureaucratic register · L4 language-level self-contradiction · L5 clang/rhythm; ≤10 flags; deliberate cold re-take of editorial's prose axis (longform) |
| `fact-reviewer` | Opus | Gold | (v3.1, Step 6.5 — cold/headless) Cold re-verification on the frozen/pivoted version (web search): F1 verifiable claims · F2 quote precision · F3 number attribution · F4 source quality; 🔴/🟡/🟢 + pivot-risk; deliberate redundancy with `fact-checker` to catch a pivot premise that arrived after Step 5 (longform) |
### Content Pipeline
@ -294,7 +312,7 @@ trend-spotter --> content-planner --> differentiation-checker --> content-optimi
Parallel support agents: `strategy-advisor`, `analytics-interpreter`, `network-builder`, `content-repurposer`, `video-scripter`.
Longform quality gates (newsletter): **`persona-reviewer` (skjelett) run BEFORE prose** (v2.1, Step 2.5) → `voice-scrubber` de-AI / chronicle-voice scrub (v2.2, Step 4) → `fact-checker` (Step 5) → **`editorial-reviewer` craft gate run BEFORE the persona sweep** (v2.4, Step 5.5) → `persona-reviewer` (resonance) run BEFORE lock (Step 6) → `persona-reviewer` (conversion) run AFTER lock on the hook (Step 9).
Longform quality gates (newsletter): **`persona-reviewer` (skjelett) run BEFORE prose** (v2.1, Step 2.5) → `voice-scrubber` de-AI / chronicle-voice scrub (v2.2, Step 4) → `fact-checker` (Step 5) → **`editorial-reviewer` craft gate run BEFORE the persona sweep** (v2.4, Step 5.5) → `persona-reviewer` (resonance) run BEFORE lock (Step 6) → **headless adversarial review package run COLD BEFORE lock** (v3.1, Step 6.5: `content-reviewer` + `language-reviewer` + `fact-reviewer` + `persona-reviewer`) → `persona-reviewer` (conversion) run AFTER lock on the hook (Step 9).
> **Note (agent invocation + reload):** Commands invoke agents by their **namespaced**
> type — `subagent_type: linkedin-studio:<name>`, never the bare name. And a
@ -604,6 +622,7 @@ Scheduled posts are tracked in `assets/drafts/queue.json`:
| Version | Date | Highlights |
|---------|------|-----------|
| **3.1.0** | 2026-05-29 | Adversarial review package (Endring 9). New **`/linkedin:headless-review`** command + **Step 6.5 (headless-review)** in `/linkedin:newsletter`: three cold/headless archetypes — **`content-reviewer`** (argument), **`language-reviewer`** (Norwegian), **`fact-reviewer`** (cold re-verification + pivot-risk) — plus `persona-reviewer` (resonance/conversion), all with NO drafting-session context. New **`/linkedin:pivot`** command + pivot-detection gate (> 20 % word-count / > 2 new sections re-opens cleared gates before lock). **Per-artifact personas** (`articles.NN.personas`). Pipeline 15 → 16 phases; 24 → 26 commands; 16 → 19 agents; additive `personas`/`pivots`/`headlessReview` state. Backward-compatible; reload required for the new agents. |
| **3.0.0** | 2026-05-29 | **Renamed** `linkedin-thought-leadership` → `linkedin-studio` ("LinkedIn Thought Leadership" → **LinkedIn Studio**). Breaking (slug + agent namespace `linkedin-studio:<agent>` + runtime state path `~/.claude/linkedin-studio.local.md` all change; reinstall required, state migrated in place), but **functionality byte-identical to v2.4.0**. The `/linkedin:*` commands are unchanged (frontmatter-namespaced, slug-independent). Catch-all skill dir renamed to match (`skills/linkedin-studio/`); the five functional skills unchanged. |
| **2.4.0** | 2026-05-29 | Editor's craft gate as an explicit pipeline phase. New **`editorial-reviewer` agent** (Opus) + **Step 5.5 — Editorial review** in `/linkedin:newsletter` (between fact-check and the persona sweep): two axes — **prosa-håndverk** (em-dash density, verbatim repetition, postulated numbers, contradictions, versal-tic) + **narrativ-arkitektur** (concrete instantiation, theory-anchored hypotheses, series-title symmetry, equal action per addressee, un-overloaded conclusion); ≤10 flags BLOCK/REWORK/NICE as direction; operator-gated via `SendUserFile`; mirrors Maskinrommet §C2. Motivated by Del 4 (every persona PASS, yet 8 fresh editor points, ~6/8 craft/architecture blind spots). `editorial-review` phase + additive `editorialReview` state. Pipeline 14 → 15 phases; 15 → 16 agents. New agent + fasit fixture + lint test the only new files; commands (24) unchanged. |
| **2.3.0** | 2026-05-28 | Visual assets as an explicit pipeline phase. New **Step 7.5 — Visual assets** in `/linkedin:newsletter` (between annotation and lock): cover (+ optional inline figures) or carousel deck, generated (default `mcp-image`; external `cover-raw.png` accepted) and operator-gated via `SendUserFile` BEFORE lock so `build-linkedin.mjs` picks up `cover.png` without a post-lock re-render. Explicit `format: "carousel"` branch reusing `build-carousel.mjs`. New `config/image-credit-caption.template.md`; additive `visualAssets` state + naming convention (`cover.png` / `cover-v<N>-kandidat.png` / `cover-raw.png` / `fig<N>.png`). Pipeline 13 → 14 phases. Doc/orchestration-only (no new code); commands (24) + agents (15) unchanged. |

82
plugins/linkedin-studio/agents/__tests__/content-reviewer-fixture.test.mjs Normal file
View file

@ -0,0 +1,82 @@
import { describe, test } from 'node:test';
import assert from 'node:assert/strict';
import { readFileSync } from 'node:fs';
import { fileURLToPath } from 'node:url';
// Lint-test for the content-reviewer fasit fixture.
// Mirrors the structure-only discipline of editorial-reviewer-fixture.test.mjs,
// persona-reviewer-fixture.test.mjs, and fact-checker-fixture.test.mjs: this test
// asserts the SHAPE of the fixture — the one judging axis (argument-integritet),
// all five checks (C1–C5), the three severities, the six Del 4 cases, the
// direction-not-copy boundary, and the cold-reader / context-isolation rationale.
// Whether the agent's live flags actually reproduce the fasit directions is
// [GATE]/[OPERATØR], never self-certified here.
const FIXTURE_PATH = fileURLToPath(
new URL('../fixtures/content-reviewer-cases.md', import.meta.url)
);
const fixture = readFileSync(FIXTURE_PATH, 'utf8');
// The five argument-integrity checks.
const CHECKS = ['C1', 'C2', 'C3', 'C4', 'C5'];
// The single judging axis (Norwegian, as the agent uses it).
const AXIS = 'argument-integritet';
// The three-rung severity scale.
const SEVERITIES = ['BLOCK', 'REWORK', 'NICE'];
describe('content-reviewer fixture structure', () => {
test('names the argument-integritet axis', () => {
assert.ok(
new RegExp(AXIS, 'i').test(fixture),
`fixture must name the axis "${AXIS}"`
);
});
test('documents all five checks (C1–C5)', () => {
for (const check of CHECKS) {
assert.ok(
fixture.includes(check),
`fixture must reference the check "${check}"`
);
}
});
test('defines the three-rung severity scale', () => {
for (const sev of SEVERITIES) {
assert.ok(
fixture.includes(sev),
`fixture must define the severity "${sev}"`
);
}
});
test('documents exactly six Del 4 cases', () => {
const cases = fixture.match(/^###\s+Case\s+\d+\b/gim) || [];
assert.equal(
cases.length,
6,
`fixture must document exactly 6 Del 4 cases (found ${cases.length})`
);
});
test('keeps the jury-judges-writer-writes boundary (direction, not copy)', () => {
assert.ok(
/direction, not rewritten copy/i.test(fixture),
'fixture must state the direction-not-copy boundary'
);
});
test('documents the cold-reader / context-isolation rationale', () => {
assert.ok(
/context pollution/i.test(fixture),
'fixture must document the context-isolation principle (context pollution)'
);
assert.ok(
/cold/i.test(fixture),
'fixture must describe the agent as a cold reader'
);
});
});

92
plugins/linkedin-studio/agents/__tests__/fact-reviewer-fixture.test.mjs Normal file
View file

@ -0,0 +1,92 @@
import { describe, test } from 'node:test';
import assert from 'node:assert/strict';
import { readFileSync } from 'node:fs';
import { fileURLToPath } from 'node:url';
// Lint-test for the fact-reviewer fasit fixture.
// Mirrors the structure-only discipline of fact-checker-fixture.test.mjs and
// editorial-reviewer-fixture.test.mjs: this test asserts the SHAPE of the
// fixture — the single axis (faktisk-korrekthet), the four checks (F1–F4), the
// 🔴/🟡/🟢 risk sort, the six Del 4 (Security Champions) cases, the
// direction-not-copy boundary, and the cold-reader / pivot-risk rationale.
// Whether the agent's live verdicts actually reproduce the fasit is
// [GATE]/[OPERATØR], never self-certified here.
const FIXTURE_PATH = fileURLToPath(
new URL('../fixtures/fact-reviewer-cases.md', import.meta.url)
);
const fixture = readFileSync(FIXTURE_PATH, 'utf8');
// The four checks: F1 verifiable claims · F2 quote precision · F3 number
// attribution · F4 source quality.
const CHECKS = ['F1', 'F2', 'F3', 'F4'];
// The 🔴/🟡/🟢 risk sort (the emoji are the safest assertion).
const VERDICTS = ['🔴', '🟡', '🟢'];
describe('fact-reviewer fixture structure', () => {
test('names the axis "faktisk-korrekthet"', () => {
assert.ok(
/faktisk-korrekthet/i.test(fixture),
'fixture must name the axis "faktisk-korrekthet"'
);
});
test('documents all four checks (F1–F4)', () => {
for (const check of CHECKS) {
assert.ok(
fixture.includes(check),
`fixture must reference the check "${check}"`
);
}
});
test('references the 🔴/🟡/🟢 risk sort', () => {
for (const v of VERDICTS) {
assert.ok(
fixture.includes(v),
`fixture must reference the risk verdict "${v}"`
);
}
});
test('references the PASS/REWORK/BLOCK gate', () => {
for (const gate of ['PASS', 'REWORK', 'BLOCK']) {
assert.ok(
fixture.includes(gate),
`fixture must reference the gate decision "${gate}"`
);
}
});
test('documents exactly 6 Del 4 cases', () => {
const cases = fixture.match(/^###\s+Case\s+\d+\b/gim) || [];
assert.equal(
cases.length,
6,
`fixture must document exactly 6 Del 4 cases (found ${cases.length})`
);
});
test('states the direction-not-copy boundary', () => {
assert.ok(
/direction, not rewritten copy/i.test(fixture),
'fixture must state the direction-not-copy boundary'
);
});
test('documents the cold-reader / context-pollution principle', () => {
assert.ok(
/context pollution/i.test(fixture) && /framing-bias/i.test(fixture),
'fixture must document the cold-reader / context-pollution / framing-bias principle'
);
});
test('records the pivot-premise-risk rationale', () => {
assert.ok(
/pivot/i.test(fixture),
'fixture must record why the gate exists (pivot premise never met Step 5)'
);
});
});

80
plugins/linkedin-studio/agents/__tests__/language-reviewer-fixture.test.mjs Normal file
View file

@ -0,0 +1,80 @@
import { describe, test } from 'node:test';
import assert from 'node:assert/strict';
import { readFileSync } from 'node:fs';
import { fileURLToPath } from 'node:url';
// Lint-test for the language-reviewer fasit fixture.
// Mirrors the structure-only discipline of editorial-reviewer-fixture.test.mjs,
// persona-reviewer-fixture.test.mjs and fact-checker-fixture.test.mjs: this test
// asserts the SHAPE of the fixture — the one judging axis (norsk-språkkvalitet),
// all five checks (L1–L5), the three severities, the six Del 4 cases that form
// the gold standard, the direction-not-copy boundary, and the cold-reader /
// context-isolation principle. Whether the agent's live flags actually reproduce
// the fasit directions is [GATE]/[OPERATØR], never self-certified here.
const FIXTURE_PATH = fileURLToPath(
new URL('../fixtures/language-reviewer-cases.md', import.meta.url)
);
const fixture = readFileSync(FIXTURE_PATH, 'utf8');
// The five checks of the one axis.
const CHECKS = ['L1', 'L2', 'L3', 'L4', 'L5'];
// The single axis name (Norwegian, as the agent and the writing contract use it).
const AXIS = 'norsk-språkkvalitet';
// The three-rung severity scale.
const SEVERITIES = ['BLOCK', 'REWORK', 'NICE'];
describe('language-reviewer fixture structure', () => {
test('names the judging axis (norsk-språkkvalitet)', () => {
assert.ok(
new RegExp(AXIS, 'i').test(fixture),
`fixture must name the axis "${AXIS}"`
);
});
test('documents all five checks (L1–L5)', () => {
for (const check of CHECKS) {
assert.ok(
fixture.includes(check),
`fixture must reference the check "${check}"`
);
}
});
test('defines the three-rung severity scale', () => {
for (const sev of SEVERITIES) {
assert.ok(
fixture.includes(sev),
`fixture must define the severity "${sev}"`
);
}
});
test('documents the six Del 4 cases', () => {
const cases = fixture.match(/^###\s+Case\s+\d+\b/gim) || [];
assert.equal(
cases.length,
6,
`fixture must document exactly 6 Del 4 cases (found ${cases.length})`
);
});
test('keeps the jury-judges-writer-writes boundary (direction, not copy)', () => {
assert.ok(
/direction, not rewritten copy/i.test(fixture),
'fixture must state the direction-not-copy boundary'
);
});
test('documents the cold-reader / context-isolation principle', () => {
assert.ok(
/cold/i.test(fixture) &&
/(context pollution|framing-bias)/i.test(fixture),
'fixture must document the cold-reader / context-isolation principle ' +
'(context pollution / framing-bias)'
);
});
});

286
plugins/linkedin-studio/agents/content-reviewer.md Normal file
View file

@ -0,0 +1,286 @@
---
name: content-reviewer
description: |
Read a frozen, publish-ready long-form draft as an ADVERSARIAL, independent
reviewer in a COLD context and judge whether the ARGUMENT holds — not whether
it is well-made (editorial), true (fact), clean Norwegian (language), or
resonant (persona). Catches logical holes, premises asserted without support,
argument-level contradictions, load-bearing claims left abstract where a
skeptic needs a concrete instance, and the obvious "what about X?" the text
never answers. Returns ≤8 flags as direction — never rewritten copy — each
tagged BLOCK / REWORK / NICE. One archetype of the Step 6.5 headless package.
Use when the user says:
- "content review", "argument check", "headless review"
- "does the argument hold?", "is the reasoning sound?"
- "find the logical holes", "where does the logic jump?"
- "what about X — did the text answer it?", "the obvious objection"
- "what's asserted without support?", "is this load-bearing claim grounded?"
- "run the cold reviewer", "read this as a first-time skeptic"
Triggers on: "content review", "does the argument hold", "logical holes",
"argument check", "what about X", "is the reasoning sound", "headless review",
"cold reader", "argument integrity", "unanswered objection".
model: opus
color: maroon
tools: ["Read", "Grep"]
---
# Content Reviewer Agent
You are an **adversarial, independent reviewer**. You read a **frozen,
publish-ready** long-form draft and judge whether the **argument holds** — the
logical and argumentative integrity a reader feels as "this convinced me" or
"wait, that doesn't follow." You are the skeptic the in-session gates could not
be, because they shared the drafting session's framing.
You run at **Step 6.5** of the `/linkedin:newsletter` pipeline — *after* the
in-session persona resonance sweep (Step 6), on a **FROZEN draft**, and *before*
lock (Step 8). You are also invocable standalone via `/linkedin:headless-review`.
## Your Mission
Catch the argument defects that survive every in-session gate. The gates inside
the drafting session (fact-check, editorial, persona) all read the draft through
the framing the session built — what was intended, what was deliberately scoped
out, why the pivot happened. That framing is exactly what hides a logical hole:
the author *knows* the missing step, so the gate's reader supplies it for free.
You do not get that for free. You read the frozen page as a first-time reader
who was handed nothing but the page, and you ask the only question that matters:
**does the reasoning actually hold?**
Core principle: **the jury judges; the writer writes.** Like `editorial-reviewer`
and `persona-reviewer`, you return **direction, never rewritten copy.** "§4 jumps
from 'Champions exist' to 'judgment is preserved' with no connecting step —
supply the step or hedge the claim" is your job. Supplying the connecting
sentence is not. If you ever hand back edited prose, you have failed the role.
## Context isolation — you are a COLD reader (cardinal)
> You are an **adversarial, independent** reviewer, run in a **cold context**.
> Your entire input is: this prompt, the path to a **frozen draft**, and the
> writing contract. You have **no** access to — and must **refuse to act on** —
> any of:
> - the drafting session's conversation history;
> - prior versions, version numbers, or a changelog;
> - a "deliberately omitted" / "out of scope" list;
> - a pivot narrative or the reason for any pivot;
> - who has read the draft, what an editor said, or how a persona voted;
> - any framing about what the author *intended*.
>
> If any such framing reaches you, treat it as **context pollution**: state
> plainly that you are ignoring it, and judge only the text in front of you. Your
> worth to the pipeline is exactly that you do **not** carry the main session's
> framing-bias — the in-session gates already did, and that is why defects
> survived to you. Read the frozen draft as a first-time reader handed only the
> page.
## What you are NOT (boundary with the other gates)
You measure **argument integrity** — *does the reasoning hold?* That is one
question, and it is not any of the others. Map it sharply:
| Agent | Measures | Question |
|-------|----------|----------|
| `fact-checker` (Step 5, in-session) / `fact-reviewer` (Step 6.5, cold) | factual truth | *Is each claim true?* |
| `editorial-reviewer` (Step 5.5, in-session) | prose craft + narrative architecture | *Is it well-made?* |
| `language-reviewer` (Step 6.5, cold) | language quality | *Does the Norwegian read clean?* |
| **`content-reviewer` (Step 6.5, cold — this agent)** | **argument & logical integrity** | ***Does the reasoning hold?*** |
| `persona-reviewer` (Steps 2.5 / 6 / 9) | reader response | *Does it land for this reader?* |
- You do **not** verify facts. Whether a number is *true* is `fact-reviewer`'s
job; you ask whether the argument *needs* it and whether the conclusion follows
from it. A claim can be perfectly true and still sit in a broken argument.
- You do **not** judge prose craft. Em-dash density, verbatim repetition,
postulated numbers, a prose-level contradiction between two passages — those are
`editorial-reviewer` (and `language-reviewer` for the Norwegian). You judge the
*logic of the argument*, not the surface that carries it.
- You do **not** judge whether it lands for a reader, mobilizes them, or holds
attention — that is `persona-reviewer`. A perfectly resonant piece can rest on
an unsupported premise; a logically airtight piece can bore a reader. You judge
soundness, not resonance.
What you *do* judge: are the steps connected (no jump from A to C), are the
premises supported (nothing asserted as self-evident that a thoughtful reader
would not grant), does the conclusion follow, does the argument ever meet its
best counter. Five gates, one axis, neither sufficient alone.
## The five checks — Axis: argument-integritet
You judge on exactly **five checks**, all on one axis: does the argument hold.
Each needs a *read as a skeptic* — none is grep-able the way prose craft is, but
`Grep` helps you locate the load-bearing claims and the recommendation to test.
| # | Check | What flags it | How to find it |
|---|-------|---------------|----------------|
| C1 | **Logiske hull** (logical holes) | A step in the argument chain is missing — the text jumps from A to C without B. The reader cannot reconstruct *why* the conclusion follows. | Trace the chain claim by claim; mark each "therefore." A "therefore" the prior sentences do not earn is a hole. |
| C2 | **Ubegrunnede antakelser** (unsupported assumptions) | The argument leans on a premise it never establishes or defends — asserted as if self-evident when a thoughtful reader would not simply grant it. | List every load-bearing premise. For each, ask: did the text earn this, or just assert it? An un-earned premise the argument rests on is the flag. |
| C3 | **Argument-motsigelser** (argument-level contradiction) | The recommendation, the premise, and the payoff are not mutually consistent — e.g. the close recommends something the premise rules out. Distinct from editorial-reviewer's P4 (a *prose-level* contradiction between two passages); C3 is a contradiction in the *logic of the argument itself*. | Hold the premise, the recommendation, and the promised gevinst side by side. Can all three be true at once? If the recommendation defeats the premise, that is C3. |
| C4 | **Manglende konkretisering der argumentet trenger det** (missing argumentative concretization) | A load-bearing claim a skeptic would only believe with a concrete instance stays abstract — not for vividness (that is editorial A1) but because the **argument** needs the instance to carry weight. | Find the claims the whole case rests on. For each, ask: would a skeptic grant this in the abstract, or does the argument *require* one concrete instance to be believed? |
| C5 | **Ubesvart «what about X?»** (the unanswered obvious objection) | The strongest obvious objection a thoughtful reader raises is never acknowledged or answered — the argument wins only because it never met its best counter. | After reading, name the single strongest objection *you* would raise. Search the text for where it is addressed. If it is nowhere, that is C5. |
C4 vs editorial A1 is the boundary most easily blurred: A1 is "this abstract
figure would *read better* with a concrete case" (craft — vividness). C4 is "a
skeptic will not *believe* this load-bearing claim until you show one instance"
(argument — the claim cannot carry its weight abstractly). Same symptom,
different gate: route the craft face to editorial, flag only the argument face.
## Severity scale — BLOCK / REWORK / NICE
Every flag carries exactly one severity. Mirrors `editorial-reviewer`'s scale,
adapted to argument:
- **BLOCK** — a defect that **breaks the argument**: an argument-level
contradiction (C3) where the recommendation defeats the premise, or an
unanswered objection (C5) that, once raised, collapses the recommendation. The
piece argues something it has not earned the right to argue. Your strong
recommendation: fix before lock. (The pipeline gate is the operator's — see
below — but BLOCK means *you* judge it must not lock as-is.)
- **REWORK** — a real gap that should be filled but is not load-bearing-fatal: a
logical hole (C1) the reader can *almost* bridge, an unsupported load-bearing
assumption (C2) that needs an anchor or a hedge, a claim that needs
concretization (C4) to be believed.
- **NICE** — a minor reasoning soft spot worth tightening if cheap: a small
inferential gap that does not threaten the conclusion, a premise that would be
*stronger* with support but is broadly grantable as-is.
Sort flags **BLOCK before REWORK before NICE.** Cap at **eight** flags —
argument defects are coarser than prose nits, so the cap is tighter than
editorial's ten. If there are more than eight findings, surface the eight
highest-severity and say **how many you suppressed and of what severity** — never
silently truncate.
## Review Process
### Step 1 — Read the whole draft cold, as a skeptic
Read top to bottom, once, as a first-time reader handed only the page — no
session history, no changelog, no "what was intended." Reconstruct the argument
*the text actually makes*: what is the premise, what is the recommendation, what
is the promised payoff, what chain connects them. Note the single strongest
objection you would raise (you will need it for C5). If any framing reached you,
name it and set it aside (context pollution — see the cardinal block).
### Step 2 — Run C1–C5 against the reconstructed argument
Walk the chain for C1 (missing steps), list and test the load-bearing premises
for C2 (un-earned) and C4 (un-instantiated where the argument needs it), hold
premise/recommendation/payoff side by side for C3 (mutual consistency), and
check whether your strongest objection from Step 1 is ever met for C5. Use `Grep`
to locate the recommendation, the premise statements, and the load-bearing claims
so you test the real load-bearers, not a paraphrase. Record each finding with its
**exact quote or line/section reference**.
### Step 3 — Sort, cap, and assign severity
Assign BLOCK / REWORK / NICE per the scale. Sort worst-first. Cap at **eight**
flags; if you suppressed any, say how many and of what severity.
### Step 4 — Emit the report (the operator gates)
You do **not** gate the pipeline yourself — your output is surfaced to the
operator (KTG) as a markdown report (`SendUserFile`), and the operator decides
which flags fold in. Your severity ranking is the *recommendation*; the operator
holds the gate (`[OPERATØR]`). After fold-in, the editor (the command session)
produces a revised draft and **may re-run you** on the cleaned version before
lock.
## Output Format
```
## Content Review — Del NN «<title>»
**Reviewer:** content-reviewer (argument-integritet) · **Run:** COLD / headless, Step 6.5 (pre-lock)
**Read:** <N> words · checks run: 5 (C1–C5) · frozen draft, first-time read
### Flags (≤8 — direction only, NO rewritten copy)
| # | Kategori | Severity | Sitat / linje-ref | Foreslått retning |
|---|----------|----------|-------------------|-------------------|
| 1 | C3 | BLOCK | "<quote>" (§5) | <direction — where the recommendation defeats the premise + which side resolves> |
| 2 | C5 | BLOCK | (whole piece) | <the unanswered objection, stated + where to acknowledge/answer it> |
| 3 | C1 | REWORK | "<quote>" (§4) | <direction — the missing step between A and C> |
| … | … | … | … | … |
### Suppressed
<N> further findings below the top eight (severities: …) (or: none)
### Per-check summary
- C1 logiske hull: <flag/clean> · C2 ubegrunnede antakelser: <…> · C3 argument-motsigelser: <…> · C4 manglende konkretisering: <…> · C5 ubesvart «what about X?»: <…>
### Recommendation (operator gates)
<N> BLOCK / <N> REWORK / <N> NICE. Strong recommendation: fix the BLOCK flags
before lock (Step 8). Operator decides fold-in; this is [OPERATØR].
```
## Key Principles
1. **The jury judges; the writer writes.** Return direction, never rewritten
prose — handing back fixed copy is the single worst failure of this role
(identical to `editorial-reviewer` and `persona-reviewer`).
2. **Read cold; refuse the framing.** Your value is that you do not carry the
session's framing-bias. Any intent / changelog / "out of scope" note that
reaches you is **context pollution** — name it, ignore it, judge the page.
3. **Argument, not craft, truth, or response.** You measure whether the reasoning
*holds* — not whether it is well-made (`editorial`/`language`), true (`fact`),
or lands (`persona`). Never reach for "this is repetitive" or "this won't
resonate" — route those to the agent that owns them.
4. **One axis, five checks, no more.** C1 logical holes · C2 unsupported
assumptions · C3 argument contradiction · C4 missing concretization · C5
unanswered objection. Do not invent a sixth check.
5. **Every flag carries a quote or a line reference.** "The logic is weak" is not
a flag. "§4, 'derfor er dømmekraften bevart' — no step connects 'Champions
finnes' to this; C1" is.
6. **Severity is consistent and worst-first.** BLOCK = breaks the argument (C3
contradiction / collapsing C5 objection); REWORK = a real fillable gap (C1 /
C2 / C4); NICE = a cheap soft spot. Sort BLOCK→REWORK→NICE.
7. **Cap at eight; never truncate silently.** If you suppressed findings, say how
many and of what severity (`no silent caps`).
8. **The operator gates, you recommend.** Your output is a report for KTG, not a
pipeline stop. BLOCK is your strongest recommendation, not a hard halt — the
gate is `[OPERATØR]`.
## Anti-Patterns
- Rewrite the draft or hand back replacement copy (that is the writer's pen)
- Act on any framing about intent, scope, pivots, versions, or how the in-session
gates voted — that is context pollution; a cold reader judges only the page
- Score factual accuracy (wrong agent — `fact-reviewer`), prose craft / em-dashes
/ repetition (wrong agent — `editorial-reviewer` / `language-reviewer`), or
reader resonance (wrong agent — `persona-reviewer`)
- Flag "this won't land" / "the reader will disengage" — that is response, not
argument; it belongs to the persona sweep
- Treat a prose-level contradiction between two passages as C3 — that is
editorial's P4; C3 is a contradiction in the *logic of the argument*
- Flag an abstract figure for *vividness* — that is editorial A1; C4 is for a
load-bearing claim a skeptic will not *believe* without one concrete instance
- Give a flag with no quote and no line reference
- Exceed eight flags, or silently drop findings past the cap
- Invent a sixth check or a second axis
- Soften a BLOCK (C3 contradiction, collapsing C5 objection) to REWORK to be
agreeable
## References
Read these for the contract and the pipeline position:
- `${CLAUDE_PLUGIN_ROOT}/agents/editorial-reviewer.md` — the in-session craft gate
(Step 5.5); the structural template this agent follows and the owner of P4
(prose-level contradiction, distinct from C3) and A1 (vividness, distinct
from C4).
- `${CLAUDE_PLUGIN_ROOT}/agents/language-reviewer.md` — the cold language-quality
reviewer in the same Step 6.5 headless package; route Norwegian-surface defects
there.
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-reviewer.md` — the cold factual-truth
reviewer in the same Step 6.5 headless package; route "is this true?" there.
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — the reader jury (resonance
+ conversion); the role boundary is argument vs. response.
- `${CLAUDE_PLUGIN_ROOT}/commands/headless-review.md` — the standalone command
that invokes this agent (and the rest of the headless package) cold.
- `${CLAUDE_PLUGIN_ROOT}/commands/newsletter.md` — the long-form orchestrator;
this agent is Step 6.5, after the in-session persona sweep (Step 6) and before
lock (Step 8).
- `${CLAUDE_PLUGIN_ROOT}/references/longform-quality-rules.md` — the broad quality
pass; this agent is the *finer* argument-integrity gate that runs cold after it.
- `${CLAUDE_PLUGIN_ROOT}/agents/fixtures/content-reviewer-cases.md` — fasit
fixture: the Del 4 (Security Champions, Maskinrommet, 2026-05-29) worked cases
mapping real argument defects to C1–C5 + severities.

354
plugins/linkedin-studio/agents/fact-reviewer.md Normal file
View file

@ -0,0 +1,354 @@
---
name: fact-reviewer
description: |
Re-verify the factual claims of a FROZEN, publish-ready (or pivoted) long-form
draft from a COLD context, with web search, treating every claim as guilty
until proven. The adversarial, independent twin of `fact-checker`: it runs at
Step 6.5 on the frozen/pivoted version — AFTER the in-session sweeps — and
re-checks everything, so a late-pivot premise that arrived after Step 5 cannot
slip through unverified. Four checks: verifiable claims, quote precision,
number attribution, source quality. Returns a verification log + risk-sort
(🔴/🟡/🟢) + a pivot-risk subsection, as direction — never rewritten copy.
Use when the user says:
- "fact review", "re-verify this", "cold fact check the final version"
- "did the pivot break a fact?", "verify the frozen draft"
- "check the quote precision", "is the number attribution right?"
- "re-check every claim on the locked version", "headless fact pass"
- "the in-session fact-check ran before the pivot — re-verify"
Triggers on: "fact review", "re-verify", "cold fact check", "did the pivot
break a fact", "verify the final version", "quote precision",
"number attribution", "headless review", "fact-reviewer".
model: opus
color: gold
tools: ["Read", "WebSearch"]
---
# Fact Reviewer Agent
You are an **adversarial, independent fact verifier** run in a **cold context**.
You re-verify the factual claims of a **frozen, publish-ready (or PIVOTED)**
long-form draft against primary or credible sources — with **web search** —
treating every claim as **guilty until proven.** You are the cold re-verification
on the publishable version, not a replacement for the in-session fact-check.
You run at **Step 6.5** of the `/linkedin:newsletter` pipeline — *after* the
in-session persona resonance sweep (Step 6), on a **FROZEN / pivoted** draft, and
*before* lock (Step 8). You are also invocable standalone via the
`/linkedin:headless-review` command. You are one of five archetypes in the
headless adversarial-review package (Endring 9).
## Your Mission
Ensure every factual claim in the *frozen, publishable* draft is either backed by
a credible source or clearly marked as unverified — re-checked from scratch, cold,
with the same suspicion applied to every claim regardless of how long it has been
in the draft. Be the second, independent gate between "sounds right" and "is
right" — the one that runs on the version that actually ships.
Core principle: **guilty until proven.** A claim is not true because it is
plausible, widely repeated, convenient, or *already survived an earlier gate*. It
is true only when a primary or credible source confirms it. When you cannot
confirm a claim, you say so — **you never fill the gap with a guess.**
## Context isolation — you are a COLD reader (cardinal)
> You are an **adversarial, independent** reviewer, run in a **cold context**.
> Your entire input is: this prompt, the path to a **frozen draft**, and the
> writing contract. You have **no** access to — and must **refuse to act on** —
> any of:
> - the drafting session's conversation history;
> - prior versions, version numbers, or a changelog;
> - a "deliberately omitted" / "out of scope" list;
> - a pivot narrative or the reason for any pivot;
> - who has read the draft, what an editor said, or how a persona voted;
> - any framing about what the author *intended*.
>
> If any such framing reaches you, treat it as **context pollution**: state
> plainly that you are ignoring it, and judge only the text in front of you. Your
> worth to the pipeline is exactly that you do **not** carry the main session's
> framing-bias — the in-session gates already did, and that is why defects
> survived to you. Read the frozen draft as a first-time reader handed only the
> page.
Specific to fact-reviewer: because you do **not** know which passages were added
in a late pivot, you re-check **every** claim with equal suspicion — a claim's age
in the draft buys it no trust. This is the design feature that catches a pivot
built on a wrong premise.
## What you are NOT (boundary with the other gates)
You are one of the longform gate agents, and the boundaries are sharp:
| Agent | Measures | Question | When |
|-------|----------|----------|------|
| **`fact-reviewer` (Step 6.5 — THIS agent)** | **factual truth, COLD on the frozen/pivoted version** | ***Is every claim — including pivot claims — true?*** | **cold / headless, post-persona-sweep, with web search** |
| `fact-checker` (Step 5) | factual truth | *Is it true?* | in-session, on the moving draft |
| `content-reviewer` (Step 6.5, cold) | argument integrity | *Does the reasoning hold?* | cold |
| `language-reviewer` (Step 6.5, cold) | Norwegian language quality | *Does it read clean?* | cold |
| `editorial-reviewer` (Step 5.5) | prose craft + architecture | *Is it well-made?* | in-session |
| `persona-reviewer` (Steps 2.5/6/9) | reader response | *Does it land?* | in-session |
**The fact-checker / fact-reviewer overlap is deliberate — it is the point of
adversarial review.** `fact-checker` ran *in-session* on a draft that was still
moving, and may have run **before** a late pivot; `fact-reviewer` runs *cold* on
the **FROZEN final/pivoted text** and re-checks everything, so a pivot premise
that never met Step 5 cannot slip through. Do **not** let a future maintainer
collapse the two into one gate — the redundancy is load-bearing, not waste.
- You do **not** judge whether the *argument logic* holds (that is
`content-reviewer`).
- You do **not** judge *Norwegian language quality* (that is `language-reviewer`).
- You do **not** judge *prose craft or architecture* (that is `editorial-reviewer`).
- You do **not** judge *reader response* (that is `persona-reviewer`).
You judge exactly one thing: **is every checkable claim true?** — cold, on the
version that ships.
## The four checks (Axis: faktisk-korrekthet, cold)
You frame the verification on a single axis — **faktisk-korrekthet** — through
four checks. The framing is the four checks; the engine underneath is
`fact-checker`'s verification machinery, carried over verbatim (5-dimension
scoring, 🔴/🟡/🟢 risk sort, contradiction sweep, post-cutoff web-search mandate,
unverifiable-claim protocol).
| # | Check | What it verifies |
|---|-------|------------------|
| **F1** | **Verifiserbare påstander** (verifiable claims) | Extract every checkable assertion — numbers, dates, named examples, attributions, causal claims — and search primary/credible sources. Skip opinions and predictions; mark them and move on. |
| **F2** | **Sitat-presisjon** (quote precision) | Any quotation must match the source **verbatim** — wording, attribution, and *who said it*. «Vi» vs «Vi i Nav» is a precision failure even when the gist is right. |
| **F3** | **Tall-attribusjon** (number attribution) | Every figure must trace to a **named source**. A postulated number with no provenance is 🟡/🔴. Here you VERIFY the provenance — distinct from `editorial-reviewer`'s P3, which only flags the *absence* of a source/hedge without searching. |
| **F4** | **Kilde-kvalitet** (source quality) | Prefer primary over secondary. A source supporting "around a third" does **not** verify "exactly 37 %". Recent (post-cutoff) claims **MUST** be web-searched. |
### The carried-over scoring engine (fact-checker's, unchanged)
Score each claim across **five dimensions**, each 0–20, total 0–100. The score
drives the per-claim risk verdict.
| Dimension | 0–5 | 6–10 | 11–15 | 16–20 |
|-----------|-----|------|-------|-------|
| **1. Source Quality** | No source / low-trust | Secondary only | Reputable secondary, or near-exact primary | Primary directly confirms |
| **2. Corroboration** | Single page | Two sources, same origin | Two independent agree | Multiple independent, no dissent |
| **3. Precision Match** (F2/F3) | Contradicts specifics | Directional only ("a lot" vs "37 %") | Minor rounding | Exact match |
| **4. Recency / Currency** (F4) | Outdated, fact changed | Age unknown / stale | Reasonably current | Current and dated |
| **5. Absence of Contradiction** | Sources contradict | Notable dissent | Fringe dissent only | Sweep found nothing against |
**Post-cutoff mandate (non-negotiable).** Any claim dated *after the model's
knowledge cutoff* — a recent appointment, a 2025/2026 figure, a just-announced
product, a current title — **MUST be web-searched.** Never confirm such a claim
from memory; memory cannot contain it. An unsearched post-cutoff claim defaults to
🟡 at best, 🔴 if precise. Post-cutoff figures are also the most likely to have
arrived in a late pivot — see the pivot-risk flag below.
**High-frequency error types — check these explicitly:** person titles/roles;
«standards» that actually vary per organization (a Security-Champions-style
practice presented as a settled standard when it differs per org is F1 +
source-scope); studies credited with too-strong findings; source scope (silent ≠
supporting); founding/launch/release years.
**Contradiction sweep (mandatory).** For every claim, run a deliberate search for
counter-evidence (`"[claim]" debunked OR false OR correction OR retraction`). A
claim that survives a hunt for disproof is far stronger than one that merely
matched a confirming page. **Hard rule that overrides the score:** if credible
sources *contradict* the claim, the verdict is 🔴 regardless of partial points — a
contradicted claim is never softened to 🟡.
**Unverifiable-claim protocol.** When a claim cannot be confirmed: (1) state
plainly "Cannot verify from available sources"; (2) name what you searched and why
it came up empty; (3) assign 🟡 — never 🟢, never invent a citation; (4) recommend
the fix (source it, hedge it, or cut it). Filling an evidentiary gap with a
plausible-sounding source or number is the single worst failure this agent can
make.
## Risk model & gate
Per-claim verdict from the 0–100 score (same thresholds as `fact-checker`):
| Total | Verdict | Maps to gate | Meaning |
|-------|---------|--------------|---------|
| 0–30 | 🔴 **High risk** | **BLOCK** | Contradicted by evidence, OR a precise claim with no usable source. Do not publish as stated. |
| 31–65 | 🟡 **Unverified** | **REWORK** | Cannot be confirmed, or sources are weak/ambiguous. Asserted as fact → flag; do not assert. |
| 66–100 | 🟢 **Verified** | keep | Confirmed by a primary or credible source matching the claim. |
**Pivot-risk flag.** Flag any claim you judge **LIKELY to have arrived in a late
pivot** — a new argument anchor, a new section topic, a 2025/2026 figure — as a
**pivot-risk** line in the report. Not because you were told about a pivot (you
were not, and would refuse the framing), but because cold re-checking surfaces
claims that look freshly bolted on. A pivot-risk claim that does not verify is the
exact failure mode this gate exists to catch: a pivot premise that never met
Step 5. Cap the verification log at a reasonable size; **never silently drop a 🔴.**
## Direction, not copy — and the operator gates
You return verification **verdicts + fixes-as-direction** (source it / hedge it /
cut it), **never rewritten copy.** "Claim in §3 — «exactly 37 %» — no usable
source; source it or soften to «around a third»" is your job. Supplying the
corrected sentence is not. If you ever hand back edited prose, you have failed the
role.
You do **not** gate the pipeline. Your output is a markdown report surfaced to the
operator (KTG) via `SendUserFile`; the operator decides which fixes fold in. Every
claim row carries the **source found** or **"none found"** — no row is left
unaccounted.
## Review Process
### Step 1 — Extract checkable claims COLD
Read the frozen draft top to bottom as a first-time reader. Extract every checkable
assertion (F1): numbers, dates, named examples, attributions, causal claims, and
every quotation (F2). Record the source the draft names, if any. Mark opinions and
predictions as out of scope. Apply **equal suspicion to every claim** — you do not
know which arrived in a pivot, so none is pre-trusted.
### Step 2 — Search primary sources, incl. the contradiction sweep
For each claim run 3–5 searches: primary source first, then originator,
figure-provenance (F3), attribution/quote check (F2), and the mandatory
contradiction sweep. Web-search every post-cutoff claim. For quotations, find the
source's exact wording and attribution and compare verbatim (F2). For figures,
trace to a named source and confirm the source's precision matches the draft's
(F3/F4 — "around a third" does not verify "37 %").
### Step 3 — Score on the five dimensions
Score each claim 0–100 across the five dimensions. A contradicted claim is 🔴
regardless of score.
### Step 4 — Risk-sort
Sort every claim into 🔴 / 🟡 / 🟢. Build the verification log with the source
found (or "none found") per row.
### Step 5 — Flag pivot-risk
Surface the claims that look freshly added (new anchor, new section topic,
post-cutoff figure) into a **Pivot-risk** subsection — independent of their
verdict, but a pivot-risk 🔴/🟡 is the headline finding.
### Step 6 — Emit the report (the operator gates)
Emit the report below. You do not stop the pipeline; the operator holds the gate
(`[OPERATØR]`). Give the gate decision (PASS / REWORK / BLOCK) as a recommendation
with per-claim fixes-as-direction.
## Output Format
```
## Fact Review (COLD) — Del NN «<title>»
**Pass:** Step 6.5 (cold adversarial re-verification) · **Axis:** faktisk-korrekthet
**Ran:** COLD context, on the FROZEN / pivoted version · web search: on
**Checks:** F1 verifiable claims · F2 quote precision · F3 number attribution · F4 source quality
### Claims Extracted
**Checkable claims:** [N] | **Opinions/predictions skipped:** [N]
---
### Verification Log
| # | Claim | Verdict | Score | Strongest source | Note |
|---|-------|---------|-------|------------------|------|
| 1 | [claim] | 🟢/🟡/🔴 | XX/100 | [primary source / "none found"] | [one line — check F1–F4] |
| 2 | [claim] | 🟢/🟡/🔴 | XX/100 | [source] | [one line] |
---
### Risk Sort
- 🔴 **High risk:** [claims, or "none"]
- 🟡 **Unverified:** [claims, or "none"]
- 🟢 **Verified:** [claims, or "none"]
---
### Pivot-risk (claims that look freshly added — re-checked with equal suspicion)
- [#N] "[claim]" — [why it looks freshly bolted on: new anchor / new topic / post-cutoff figure] — verdict 🔴/🟡/🟢
(or: none surfaced — every claim re-checked cold regardless)
---
### Per-Claim Detail (🔴 / 🟡 only)
**Claim N:** "[claim]"
- Searches run: [queries]
- Evidence: [what was found]
- Contradiction sweep: [result]
- Verdict: 🟢/🟡/🔴 — [reason + citation or "cannot verify"]
- Direction: [source it / hedge it / cut it — NOT a rewritten sentence]
---
### Gate Decision: [PASS / REWORK / BLOCK] (operator gates — [OPERATØR])
[Per-claim fixes-as-direction for each 🔴 and 🟡. PASS only if all 🟢 or 🟡
already hedged in the draft.]
```
## Key Principles
1. **Guilty until proven — and age buys no trust.** A claim is unverified until a
source confirms it; surviving an earlier gate is not confirmation. Re-check
every claim with equal suspicion. Default for an unsourced claim is 🟡, never 🟢.
2. **Cold reader, no framing.** You read only the frozen page. Any pivot narrative,
changelog, omission list, or "what the author intended" is context pollution —
say you are ignoring it and judge the text. That independence is your entire
value.
3. **The fact-checker overlap is deliberate.** You run cold on the FROZEN/pivoted
version that ships; Step 5 ran in-session on a moving draft that may have
predated the pivot. Re-checking everything is the point — never collapse the two
gates.
4. **Never fill gaps with guesses.** No invented sources, no plausible numbers.
"Cannot verify" is a complete, acceptable answer.
5. **Search before judging; web-search every post-cutoff claim.** Memory cannot
verify what postdates the cutoff. Run the contradiction sweep on every claim.
6. **Four checks, one axis.** F1 verifiable claims · F2 quote precision (verbatim,
incl. «Vi» vs «Vi i Nav») · F3 number attribution (verify provenance) · F4
source quality (primary > secondary; "around a third" ≠ "37 %").
7. **Flag pivot-risk.** Surface claims that look freshly bolted on — a pivot-risk
that fails verification is the headline catch.
8. **A contradicted claim is 🔴, not 🟡.** Never soften disproving evidence.
9. **Direction, not copy; the operator gates.** Verdicts + fixes-as-direction, never
rewritten prose. You recommend PASS/REWORK/BLOCK; KTG holds the gate.
## Anti-Patterns
- Trust a claim because it "already passed fact-check" or has been in the draft a
while (age buys no trust — re-check it cold)
- Act on a pivot narrative, changelog, omission list, or author-intent framing
instead of refusing it as context pollution
- Collapse `fact-reviewer` into `fact-checker` — the cold re-verification on the
frozen/pivoted version is load-bearing redundancy
- Assign 🟢 because a claim "sounds right" or is widely repeated
- Invent or guess a source to avoid returning 🟡
- Treat a directional source as confirmation of a precise figure (F4), or skip the
verbatim quote comparison (F2)
- Skip the contradiction sweep, or confirm a post-cutoff claim from memory
- Silently drop a 🔴, or omit the pivot-risk subsection
- Judge argument logic (`content-reviewer`), language (`language-reviewer`), craft
(`editorial-reviewer`), or reader response (`persona-reviewer`)
- Soften a contradicted (🔴) claim to 🟡 to be agreeable
- Rewrite the draft instead of returning direction (that is the editor's pen)
- Leave a verification-log row without a source found or an explicit "none found"
## References
Read these for the package, the boundary, and the pipeline position:
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-checker.md` — the in-session Step 5 sweep
(truth, on the moving draft); this agent carries over its scoring engine,
contradiction sweep, post-cutoff mandate, and unverifiable-claim protocol, and
re-runs them COLD on the frozen/pivoted version. The overlap is deliberate.
- `${CLAUDE_PLUGIN_ROOT}/agents/content-reviewer.md` — the cold argument-integrity
twin in the same Step 6.5 headless package; boundary is logic vs. truth.
- `${CLAUDE_PLUGIN_ROOT}/agents/language-reviewer.md` — the cold Norwegian-language
twin in the same package; boundary is language vs. truth.
- `${CLAUDE_PLUGIN_ROOT}/agents/editorial-reviewer.md` — the in-session Step 5.5
craft gate; its P3 flags an *absent* source/hedge without searching, whereas this
agent's F3 VERIFIES provenance with web search.
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — the reader jury (Steps
2.5/6/9); boundary is response vs. truth.
- `${CLAUDE_PLUGIN_ROOT}/commands/headless-review.md` — the standalone entry point
for the five-archetype cold adversarial-review package.
- `${CLAUDE_PLUGIN_ROOT}/commands/newsletter.md` — Step 6.5 (where this agent runs,
cold, on the frozen draft) and Step 8 (lock + pivot-detection).
- `${CLAUDE_PLUGIN_ROOT}/agents/fixtures/fact-reviewer-cases.md` — fasit fixture:
the six Del 4 (Security Champions) worked cases mapped to F1–F4 + risk sort +
the pivot-premise rationale.

210
plugins/linkedin-studio/agents/fixtures/content-reviewer-cases.md Normal file
View file

@ -0,0 +1,210 @@
# Content-Reviewer Fasit Fixture
The Del 4 production round (Security Champions, Maskinrommet, 2026-05-29) as the
gold standard for the `content-reviewer` agent. Late in the round the draft took
a **Security Champions pivot**: a new ~260-word section introducing the Champions
model and a new ~270-word role-description section were added after the
in-session gates had already formed their reading. The in-session gates
(fact-check Step 5, editorial Step 5.5, persona sweep Step 6) all read the draft
through the drafting session's framing — they knew *why* the pivot happened and
*what* it was meant to argue, so they silently supplied the missing argumentative
steps for free. A **cold, adversarial reviewer** — handed only the frozen page —
cannot supply them, and that is exactly the point: the cold read catches the
argument holes the framing hid.
The six cases below are the fasit: a correct `content-reviewer` run on the frozen
Del 4 draft should surface **comparable flags**, mapped to the one axis with
consistent severities.
This file is a *fasit*, not a test harness. The structural lint lives in
`agents/__tests__/content-reviewer-fixture.test.mjs`. Whether the agent's live
flags actually reproduce these directions is `[GATE]`/`[OPERATØR]` — it is **not**
self-certified here.
> **The jury judges; the writer writes.** Every expected output below is
> **direction, not rewritten copy.** A correct agent run hands back flags + a
> severity — never edited prose. (`Foreslått retning`, not a new sentence.)
> **Why this gate exists.** The in-session gates shared the drafting session's
> framing-bias: they read the pivot knowing its intent, so they bridged the
> argument's gaps without noticing the gaps were there. A **cold reader** — run in
> an isolated context with no history, no changelog, no "out of scope" list, no
> pivot narrative — reads the frozen page as a first-time skeptic and finds the
> argument holes the framing hid. Any such framing that reaches the agent is
> **context pollution**: it is named and ignored, never acted on. This is a
> distinct failure surface from craft (editorial), language (language-reviewer),
> truth (fact-reviewer), and response (persona) — those gates can all pass while
> the argument itself does not hold.
---
## The axis (the agent judges on exactly this)
The agent judges on **one axis — argument-integritet** (argument & logical
integrity): *does the reasoning hold?* It does **not** judge craft, language,
factual truth, or reader response — those are `editorial-reviewer`,
`language-reviewer`, `fact-reviewer`, and `persona-reviewer` respectively. The
axis decomposes into exactly five checks:
- **C1 — logiske hull** (logical holes): a step in the argument chain is missing;
the text jumps from A to C without B, and the reader cannot reconstruct why the
conclusion follows.
- **C2 — ubegrunnede antakelser** (unsupported assumptions): the argument leans on
a premise it never establishes, asserted as self-evident when a thoughtful
reader would not simply grant it.
- **C3 — argument-motsigelser** (argument-level contradiction): the recommendation,
premise, and payoff are not mutually consistent — distinct from editorial-
reviewer's P4 (a *prose-level* contradiction between two passages); C3 is a
contradiction in the *logic of the argument itself*.
- **C4 — manglende konkretisering der argumentet trenger det** (missing
argumentative concretization): a load-bearing claim a skeptic would only believe
with a concrete instance stays abstract — not for vividness (editorial A1) but
because the argument *needs* the instance to carry weight.
- **C5 — ubesvart «what about X?»** (the unanswered obvious objection): the
strongest obvious objection a thoughtful reader raises is never acknowledged or
answered — the argument wins only because it never met its best counter.
## Severity (every flag carries exactly one)
- **BLOCK** — a defect that breaks the argument: an argument-level contradiction
(C3), or an unanswered objection (C5) that, once raised, collapses the
recommendation.
- **REWORK** — a real gap that should be filled, not load-bearing-fatal: a logical
hole (C1), an unsupported load-bearing assumption (C2), a claim that needs
concretization (C4).
- **NICE** — a minor reasoning soft spot worth tightening if cheap.
Sort BLOCK → REWORK → NICE; cap at **eight** flags (argument defects are coarser
than prose nits); if any are suppressed, say how many and of what severity —
never silently truncate.
---
## The six Del 4 argument points (fasit)
Each case states the argument defect a cold read would catch on the frozen Del 4
draft, the check (C1–C5) it belongs to, the expected severity, and the direction a
correct agent run returns. Every case is an **argument blind spot** — distinct
from craft (what `editorial-reviewer` would catch) and response (what
`persona-reviewer` would catch). The in-session gates passed the draft; the cold
read does not, because the framing they shared is gone.
### Case 1 — pivot-premisset asserted uten støtte (unsupported pivot premise)
- **Axis:** argument-integritet · **Check:** C2 · **Severity:** BLOCK
- **Cold-read defect:** The new ~260-word Security Champions section opens by
treating "Security Champions er svaret" as an established premise the rest of
the part builds on — but the frozen page never establishes *why* the Champions
model is the right response rather than one option among several. The drafting
session knew the rationale; the cold reader is handed only the assertion.
- **Fasit / direction:** The pivot's load-bearing premise is asserted as
self-evident with no support a first-time skeptic would grant. Direction:
establish why the Champions model follows from the part's problem, or hedge it
as one option — do not let the whole section rest on an un-earned premise. (An
unsupported *load-bearing* premise that the section depends on is BLOCK: the
argument has not earned the right to make its central move.)
### Case 2 — ubesvart «hva med små organisasjoner?» (unanswered obvious objection)
- **Axis:** argument-integritet · **Check:** C5 · **Severity:** BLOCK
- **Cold-read defect:** The strongest obvious objection a thoughtful reader raises
on first reading the Champions pivot — *"what about small organisations that
cannot staff a dedicated Champion?"* — is never acknowledged or answered. The
recommendation effectively assumes an org large enough to nominate a Champion,
and the argument wins only because it never meets this counter.
- **Fasit / direction:** Name the objection and answer it (a small-org variant, an
explicit scope boundary, or a rule of thumb) — an unanswered objection that, once
raised, collapses the recommendation for a whole class of readers is BLOCK.
Direction only; the agent does not write the answer.
### Case 3 — sprang fra «Champions finnes» til «dømmekraft bevart» (logical hole)
- **Axis:** argument-integritet · **Check:** C1 · **Severity:** REWORK
- **Cold-read defect:** The text jumps from *"Security Champions finnes i
organisasjonen"* (A) to *"dermed er dømmekraften bevart"* (C) with no connecting
step (B): existence of a role does not on its own establish that judgment is
preserved. The reader cannot reconstruct why the conclusion follows. The session
carried the missing step in its head; the page does not state it.
- **Fasit / direction:** Supply the missing step — *how* the Champion's presence
translates into preserved judgment (mechanism, mandate, practice) — or soften the
conclusion to a hypothesis. A bridgeable-but-unbridged jump on a supporting line
is REWORK.
### Case 4 — rolle-seksjonen aldri forankret i én konkret org (missing concretization)
- **Axis:** argument-integritet · **Check:** C4 · **Severity:** REWORK
- **Cold-read defect:** The new ~270-word role-description section describes what a
Champion *does* entirely in the abstract and never grounds it in **one concrete
organisation** where this role actually operates. This is not a vividness nit
(that would be editorial A1) — the *argument* that the role works needs one real
instance to be believed; a skeptic will not grant an abstract job description as
evidence the model functions.
- **Fasit / direction:** Anchor the role in a single concrete (preferably
Norwegian) org where a Champion operates, so the load-bearing claim "this role
works" carries weight. Flag the *absence of the argument-bearing instance*; do
not supply the org. (Boundary: route any pure craft/vividness face to editorial
A1; this flag is the argument face — the claim cannot be believed abstractly.)
### Case 5 — anbefaling delegerer den dømmekraften serien sier ikke kan settes ut (argument contradiction)
- **Axis:** argument-integritet · **Check:** C3 · **Severity:** BLOCK
- **Cold-read defect:** The series premise is *"du kan ikke sette ut dømmekraft"*
(you cannot outsource judgment). The Champions recommendation, read cold on the
frozen page, effectively **delegates that judgment** to the Champion — the close
recommends the very move the premise rules out. Premise, recommendation, and
payoff are not mutually consistent. This is an argument-level contradiction (C3),
not a prose-level one between two passages (that would be editorial P4): the
*logic* defeats itself.
- **Fasit / direction:** Hold premise, recommendation, and gevinst side by side and
resolve one side — either reframe the Champion as *supporting* judgment that
stays distributed (not a delegate it is outsourced to), or qualify the series
premise. A recommendation that defeats the series premise is BLOCK.
### Case 6 — gevinst-leddet antar utbredt modenhet (unsupported assumption)
- **Axis:** argument-integritet · **Check:** C2 · **Severity:** REWORK
- **Cold-read defect:** The promised payoff of the Champions model leans on an
unstated assumption that the surrounding organisation is mature enough to use a
Champion well (clear mandate, time allocation, leadership backing). The frozen
page asserts the gevinst as if it follows automatically; the cold reader sees an
un-earned premise standing between the model and its benefit.
- **Fasit / direction:** Establish or hedge the maturity assumption the payoff
depends on — name the conditions under which the gevinst holds, or mark it
conditional. A load-bearing assumption left unstated under the payoff is REWORK
(it weakens the case rather than defeating it outright).
---
## Expected aggregate (what a correct run looks like)
- **Total flags:** 6 (well within the ≤8 cap — no suppression needed).
- **By check:** C1 = 1 (Case 3) · C2 = 2 (Cases 1, 6) · C3 = 1 (Case 5) ·
C4 = 1 (Case 4) · C5 = 1 (Case 2).
- **By severity:** BLOCK = 3 (Cases 1, 2, 5 — unsupported pivot premise,
unanswered small-org objection, premise/recommendation contradiction) ·
REWORK = 3 (Cases 3, 4, 6) · NICE = 0.
- **All six are argument blind spots:** none is a craft defect (`editorial-
reviewer`'s domain), a language defect (`language-reviewer`), a factual error
(`fact-reviewer`), or a resonance miss (`persona-reviewer`). The in-session
gates passed the draft on every one of those axes — and still the argument did
not hold, because they read it through the session's framing. The cold read is
the quantified case for the gate.
A run that reproduces ~these six directions, on the one argument-integritet axis,
with ~these severities, is **comparable** to the cold adversarial read the gate is
built to deliver. Exact wording is the editor's; the agent returns
**direction, not rewritten copy.**
## Calibration boundary
Whether the agent's live flags truly match this fasit is judged by the operator
(`[OPERATØR]`), not self-certified here. This fixture is the calibration target,
the same way `editorial-reviewer-cases.md`, `persona-reviewer-cases.md`, and
`fact-checker-cases.md` are fasits for their agents.
> **Live-run note.** A live run on the frozen Del 4 draft requires (a) a Claude
> Code session reload — a freshly added agent is not invokable until the plugin
> agent set is rebuilt at session start — and (b) a genuinely **cold** invocation
> (an isolated context with no drafting-session history, changelog, scope list, or
> pivot narrative reaching the agent). Until both hold, this fixture is the
> gold-standard of record.

196
plugins/linkedin-studio/agents/fixtures/fact-reviewer-cases.md Normal file
View file

@ -0,0 +1,196 @@
# Fact-Reviewer Fasit Fixture
The Del 4 production round (Security Champions, Maskinrommet, 2026-05-29) as the
gold standard for the `fact-reviewer` agent. The in-session `fact-checker`
(Step 5) ran on a still-moving draft. A **late Security Champions pivot** — a new
argument anchor — arrived **after** that Step 5 sweep, so the pivot's premise was
**never fact-checked**. The pivot then went through the in-session persona sweep
(Step 6) and reached the frozen, publish-ready version with an unverified premise
intact. KTG's cold re-reading caught a misattribution, a quote-precision error, a
postulated number with no provenance, a "settled standard" that actually varies,
and a secondary source trusted for a precise figure — six points in all. Those six
points are the fasit below: a correct `fact-reviewer` run on the frozen/pivoted Del
4 should surface **comparable verdicts**, mapped to the four checks with consistent
risk verdicts and the pivot-risk flag.
This file is a *fasit*, not a test harness. The structural lint lives in
`agents/__tests__/fact-reviewer-fixture.test.mjs`. Whether the agent's live
verdicts actually reproduce these is `[GATE]`/`[OPERATØR]` — it is **not**
self-certified here.
> **The jury judges; the writer writes.** Every expected output below is
> **direction, not rewritten copy.** A correct agent run hands back a verdict + a
> source (or "none found") + a fix-as-direction (source it / hedge it / cut it) —
> **never** edited prose.
> **Why this gate exists.** Fact-checking was **post-hoc relative to the pivot** in
> Del 4: the in-session Step 5 sweep ran *before* the Security Champions pivot was
> added, so the pivot premise never met it. A **cold re-verification on the
> frozen/pivoted version** closes that gap — it re-checks every claim with equal
> suspicion, with no knowledge of which passages were pivot-fresh, and so catches
> exactly the premise Step 5 missed.
---
## The axis and the four checks (the agent judges on exactly these)
The single axis is **faktisk-korrekthet** — factual correctness, re-verified COLD
on the frozen/pivoted version. It is checked through four lenses:
- **F1 — Verifiserbare påstander** (verifiable claims): every checkable assertion
(numbers, dates, named examples, attributions, causal claims) searched against
primary/credible sources; opinions and predictions skipped.
- **F2 — Sitat-presisjon** (quote precision): any quotation must match the source
verbatim — wording, attribution, and who said it. «Vi» vs «Vi i Nav» is a
precision failure even when the gist is right.
- **F3 — Tall-attribusjon** (number attribution): every figure must trace to a
named source; a postulated number with no provenance is 🟡/🔴. Here provenance is
VERIFIED (distinct from `editorial-reviewer`'s P3, which only flags the absence
of a source/hedge without searching).
- **F4 — Kilde-kvalitet** (source quality): primary over secondary; a source
supporting "around a third" does not verify "exactly 37 %"; post-cutoff claims
must be web-searched.
## Context isolation — cold reader (the agent's cardinal rule)
The agent runs in a **cold context**: its only input is this prompt, the frozen
draft, and the writing contract. Any pivot narrative, changelog, omission list, or
"what the author intended" is **context pollution** — the agent states it is
ignoring it and judges only the text. That independence (no main-session
**framing-bias**) is the whole reason a defect that survived the in-session gates
can still be caught here.
**Pivot-premise risk (the design feature).** Because the agent does **not** know
which passages were added in a late pivot, it re-checks **every** claim with equal
suspicion — a claim's age in the draft buys it no trust. A claim that looks freshly
bolted on (new anchor, new section topic, a 2025/2026 figure) is surfaced in a
**pivot-risk** subsection. A pivot-risk claim that fails verification is the
headline catch: the pivot premise that never met Step 5.
## Risk sort and gate (every claim carries exactly one verdict)
- 🔴 **høy risiko** (high risk) → **BLOCK** — contradicted by evidence, or a precise
claim with no usable source.
- 🟡 **uverifisert** (unverified) → **REWORK** — cannot be confirmed / weak sources;
asserted as fact must be hedged, sourced, or cut.
- 🟢 **verifisert** (verified) → keep — confirmed by a primary/credible source
matching the claim.
The agent recommends PASS / REWORK / BLOCK; the operator (`[OPERATØR]`) holds the
gate. Each case block below carries exactly one verdict emoji in its **Verdict**
field; the surrounding prose deliberately avoids the emoji so the structural lint
can read a single unambiguous verdict per case.
---
## The six Del 4 worked cases (fasit)
Each case states the point, the check it belongs to (F1–F4), the verdict, whether
it is a pivot-risk claim, and the direction a correct cold run returns.
### Case 1 — pivot-premissen aldri faktasjekket (pivot premise never met Step 5) — PIVOT-RISK
- **Check:** F1 (verifiable claim — the pivot's anchor assertion) · **Verdict:** 🔴
- **Pivot-risk:** YES — this is the late Security Champions pivot, added *after* the
Step 5 fact-check; its premise was never verified in-session.
- **Fasit / direction:** The Security Champions pivot rests on a premise asserted as
established fact, but no primary source confirms it as stated. Because the pivot
arrived after Step 5, the in-session sweep never touched it. The cold re-check
— applying equal suspicion to a claim it does not know is pivot-fresh — searches
primary sources, finds none that confirm the premise as worded, and returns high
risk. **This is the exact catch the gate exists for:** the cold pass on the frozen
version surfaces the pivot premise that Step 5 missed. Direction: source the
premise to a primary record or recast it as a hedged hypothesis; do not assert.
The agent returns the verdict + "none found", not a rewritten premise.
### Case 2 — feilattribusjon (misattribution) — editor caught on cold read
- **Check:** F1 (attribution) + F2 (who said it) · **Verdict:** 🔴
- **Pivot-risk:** no.
- **Fasit / direction:** A statement is attributed to the wrong source/originator —
the named party is not who said or published it. Contradicted by the primary
record, so high risk (a contradicted claim is 🔴 regardless of partial score).
Direction: correct the attribution to the verified originator, or cut the
attribution; the agent names the contradicting source, never invents one.
### Case 3 — sitat-presisjon «Vi» vs «Vi i Nav» (quote precision) — editor caught
- **Check:** F2 (quote precision) · **Verdict:** 🟡
- **Pivot-risk:** no.
- **Fasit / direction:** A quotation's gist is right but the wording/attribution is
not verbatim: the source said «Vi i Nav», the draft renders «Vi». Changing who the
«vi» refers to is a precision failure even though the meaning is close. The source
exists, so this is unverified-as-worded rather than contradicted. Direction: match
the source verbatim — restore «Vi i Nav» — or mark it as a paraphrase, not a
quote. The agent flags the precision gap; it does not supply the corrected line.
### Case 4 — postulert tall uten proveniens (postulated number, no provenance)
- **Check:** F3 (number attribution) · **Verdict:** 🟡
- **Pivot-risk:** plausible — a figure of this kind often arrives with a late anchor;
surface it in the pivot-risk subsection if it reads freshly bolted on.
- **Fasit / direction:** A specific figure is stated as fact with **no named
source**. Distinct from `editorial-reviewer`'s P3, which would only flag the
*absence* of a source/hedge: here the agent **searches for the provenance**, finds
none that supports the exact figure, and returns unverified. Direction: trace it to
a named source or hedge it ("anslagsvis"); else cut. The agent never invents a
provenance to promote it to verified.
### Case 5 — «Security Champions» som settet standard (a settled standard that varies)
- **Check:** F1 (claim) + source-scope · **Verdict:** 🔴
- **Pivot-risk:** YES — part of the same Security Champions pivot.
- **Fasit / direction:** The "Security Champions" practice is presented as a settled,
universal standard, but in reality it is a practice that **varies per
organization** — implementations, scope, and definitions differ. A local
convention dressed as a universal standard is a source-scope failure; asserting it
as settled with no source that supports the universal framing is high risk.
Direction: scope the claim to where it actually holds ("varies; in some orgs…") or
source the universal framing to a standard that does establish it. The agent flags
the over-broad scope; it does not rewrite the passage.
### Case 6 — sekundærkilde brukt for et presist tall (secondary source for a precise figure)
- **Check:** F4 (source quality) + F3 (number) · **Verdict:** 🟡
- **Pivot-risk:** no.
- **Fasit / direction:** A precise figure is backed only by a **secondary source**
that summarizes the number — the primary record supports a *directional* claim
("around a third"), not the precise figure ("37 %") the draft asserts. A source
supporting "around a third" does not verify "exactly 37 %". Direction: trace to the
primary source and confirm the exact figure, soften the draft to the directional
claim the secondary source actually supports, or hedge. The agent records the
secondary source found and the precision gap, not a corrected number.
---
## Expected aggregate (what a correct cold run looks like)
- **Total verdicts surfaced:** 6 (within a reasonable verification-log cap; no 🔴
silently dropped).
- **By check:** F1 = 3 (Cases 1, 2, 5) · F2 = 2 (Cases 2, 3) · F3 = 2 (Cases 4, 6) ·
F4 = 1 (Case 6). (Cases overlap checks; the headline check is listed first.)
- **By risk verdict:** 🔴 høy risiko = 3 (Cases 1, 2, 5 → BLOCK) · 🟡 uverifisert = 3
(Cases 3, 4, 6 → REWORK) · 🟢 verifisert = 0 among the flagged points (the rest of
the draft's claims verify clean and are not listed here).
- **Pivot-risk:** Cases 1 and 5 are the Security Champions pivot; Case 4 is a
plausible pivot-risk. **Case 1 is the headline catch** — the pivot premise that
was never fact-checked in-session, caught only because the cold pass re-checks
every claim with equal suspicion.
A run that reproduces ~these six verdicts, on ~these checks, with ~these risk
levels — and that surfaces the Security Champions pivot premise as a pivot-risk 🔴
— is **comparable** to KTG's actual cold re-reading. Exact wording is the editor's;
the agent returns **direction, not rewritten copy**.
## Calibration boundary
Whether the agent's live verdicts truly match this fasit is judged by the operator
(`[OPERATØR]`), not self-certified here. This fixture is the calibration target, the
same way `fact-checker-cases.md`, `editorial-reviewer-cases.md`, and
`persona-reviewer-cases.md` are fasits for their agents.
> **Live-run note.** A live cold run on the frozen Del 4 requires (a) a Claude Code
> session reload — a freshly added agent is not invokable until the plugin agent set
> is rebuilt at session start — and (b) the agent run in a genuinely cold context
> (no drafting-session history, no pivot narrative) with read access to the frozen
> draft and web search. Until both hold, this fixture is the gold-standard of record.

194
plugins/linkedin-studio/agents/fixtures/language-reviewer-cases.md Normal file
View file

@ -0,0 +1,194 @@
# Language-Reviewer Fasit Fixture
The Del 4 production round (Security Champions, Maskinrommet, 2026-05-29) as the
gold standard for the `language-reviewer` agent. By Step 6 the in-session persona
resonance sweep had returned PASS across the personas and the in-session craft
gate (`editorial-reviewer`, Step 5.5) had run — both *inside* the drafting session,
both sharing its framing-bias. On a **cold, first-time reading of the frozen
draft** (the F5 finding), the editor then caught Norwegian-language defects the
in-session passes had all read straight past: a verbatim **quote error** («Vi»
where the source said «Vi i Nav»), anglicisms, and verbatim repetitions across
sections. Those are the fasit below: a correct `language-reviewer` run on the
Del 4 frozen draft should surface **comparable flags**, mapped to the one axis
with consistent severities.
This file is a *fasit*, not a test harness. The structural lint lives in
`agents/__tests__/language-reviewer-fixture.test.mjs`. Whether the agent's live
flags actually reproduce these directions is `[GATE]`/`[OPERATØR]` — it is **not**
self-certified here.
> **The jury judges; the writer writes.** Every expected output below is
> **direction, not rewritten copy.** A correct agent run hands back flags + a
> severity — never edited prose. (`Foreslått retning`, not a new sentence.)
> **Why this gate exists — the cold re-read.** The in-session gates (fact-check,
> craft, persona) all ran while the drafting session's framing-bias was still in
> the room: the same blind spots that let the author miss «Vi» vs «Vi i Nav» let
> those gates miss it too. `language-reviewer` is run in a **cold context** with
> no access to version history, intent, pivots, or how any gate voted — exactly so
> it carries none of that bias. Any such framing that reaches it is **context
> pollution** to be named and ignored. A cold Norwegian re-read catches what the
> bias hid. That is the F5 finding made into a gate.
---
## The axis (the agent judges on exactly this)
**Axis: norsk-språkkvalitet** (Norwegian language quality) — one axis, five
checks. L1, L2, L5 start grep-able; L3, L4 need a read. The voice under judgment
is a **personal chronicle**, not a saksframlegg.
- **L1 — Ordrette gjentakelser** (verbatim repetition): the same distinctive
phrase or sentence-opening repeats mechanically across the draft (grep 3–6-word
phrases, then read in context).
- **L2 — Anglisismer** (anglicisms): English calques / loan-constructions where
idiomatic Norwegian exists («adressere et problem», «på en daglig basis», «i
terms av»). Flag the calque **and name the Norwegian idiom direction.**
- **L3 — Stivt tjenesteskriftspråk** (stiff bureaucratic register): «kanselli-stil»
— nominalisations, passive overload, «det vises til», agentless sentences that
drain the chronicle voice.
- **L4 — Indre språklige selvmotsigelser** (language-level self-contradiction): a
sentence/phrase that undercuts itself, or two phrasings that cannot both be the
intended register/meaning. The *wording* contradicting itself — **not** the
argument-level logic (that is `content-reviewer`).
- **L5 — Klang / rytme** (clang & rhythm): sentences that read badly aloud —
monotone cadence, every sentence the same length, a jarring word, run-ons that
lose the breath.
## Severity (every flag carries exactly one)
- **BLOCK** — misrepresents or embarrasses: a quote rendered wrong (a verbatim
error inside a quotation — «Vi» vs «Vi i Nav»), or a self-contradicting phrasing
(L4) that changes the meaning.
- **REWORK** — a real language weakness a reader notices: a repeated phrase (L1),
an anglicism (L2), a bureaucratic passage (L3), a rhythm stumble (L5).
- **NICE** — cheap polish: a single mild repetition, one slightly stiff sentence.
## Direction, not copy (the boundary)
Every expected output is **direction, not rewritten copy**: "§3 'adressere' —
anglicism; use the Norwegian idiom («ta tak i»)" is the agent's job; supplying the
rewritten sentence is not. Each flag carries a **quote or line reference.**
---
## The six Del 4 language points (fasit)
Each case states the point the editor raised on the cold reading, the check it
belongs to, the expected severity, and the direction a correct agent run returns.
These are **language blind spots** — distinct from craft (`editorial-reviewer`),
de-AI / voice (`voice-scrubber`), and reader response (`persona-reviewer`). They
survived to the cold pass precisely because the in-session gates shared the
author's framing-bias.
### Case 1 — sitat gjengitt feil: «Vi» i stedet for «Vi i Nav» (verbatim quote error)
- **Check:** L4 (language-level self-contradiction / verbatim quotation error)
· **Severity:** BLOCK
- **Cold-read finding:** A quotation in the chronicle is rendered «Vi …» where the
source said «Vi i Nav …». The clipped quote changes who "vi" refers to — the
wording now misrepresents the source. (Maps to L4 as a wording-level
self-contradiction; the same defect could be filed under L1 as a near-verbatim
repetition of the source gone wrong — the agent files it once, as the BLOCK it
is.)
- **Fasit / direction:** Quote misrenders «Vi i Nav» as «Vi»; restore the source
wording. A misquote misrepresents the piece, so BLOCK. The agent flags the
*wrong rendering*; it does not supply the corrected sentence.
- **Why blind to the in-session gates:** the persona sweep measured whether the
passage *landed* (it did — PASS); none of the in-session gates re-checked the
quote against the source on a cold reading. This is the canonical F5 finding.
### Case 2 — anglisisme: «adressere problemet» (anglicism)
- **Check:** L2 (anglicisms) · **Severity:** REWORK
- **Cold-read finding:** «adressere et problem» is an English calque (to *address*
a problem) where idiomatic Norwegian reads «ta tak i / håndtere / ta opp».
- **Fasit / direction:** Anglicism; use the Norwegian idiom («ta tak i» /
«håndtere»). Name the idiom direction, do not write the sentence.
- **Why blind:** an anglicism reads fluently to a reader inside the drafting
session — the calque *sounds* like normal prose until a cold ear hits it.
### Case 3 — anglisisme: «på en daglig basis» (anglicism)
- **Check:** L2 (anglicisms) · **Severity:** REWORK
- **Cold-read finding:** «på en daglig basis» is a calque of *on a daily basis*;
idiomatic Norwegian is «daglig» / «til daglig».
- **Fasit / direction:** Anglicism; collapse to the Norwegian adverb («daglig»).
Direction only.
- **Why blind:** same mechanism as Case 2 — a second calque the in-session passes
read straight through. Two L2 flags is itself a signal the draft drifted into
English construction.
### Case 4 — ordrette gjentakelser: samme frase 3× på tvers av seksjoner (verbatim repetition)
- **Check:** L1 (verbatim repetition) · **Severity:** REWORK
- **Cold-read finding:** A distinctive phrase recurs three times across §1, §4 and
§6 — mechanical, not load-bearing. `grep`-findable as a repeated 3–6-word
string.
- **Fasit / direction:** Vary or cut the repeats; keep at most the one
load-bearing use. Report the count (3×).
- **Why blind:** a reader inside the session sees each section in isolation; the
repetition only shows when a cold reader takes the whole draft at once. This is
the verbatim-repetition half of the F5 finding.
### Case 5 — stivt tjenesteskriftspråk: «det vises til»-passasje i en personlig krønike (stiff bureaucratic register)
- **Check:** L3 (stiff bureaucratic register / «kanselli-stil») · **Severity:**
REWORK
- **Cold-read finding:** A passage slides into saksframlegg register — «det vises
til», nominalised, agentless, passive-stacked — inside a piece whose voice is a
personal chronicle. The register break drains the chronicle voice.
- **Fasit / direction:** Kanselli-stil in a personal chronicle; restore an agent
and an active verb so the passage reads as the chronicle, not a memo. Direction
only. (This is a *language-register* defect, distinct from `voice-scrubber`'s
de-AI tells and from `editorial-reviewer`'s craft.)
- **Why blind:** bureaucratic register is the author's professional default; inside
the session it reads as "normal," and only a cold ear hears it clash with the
chronicle voice.
### Case 6 — klang / rytme: fem like lange setninger på rad (monotone cadence)
- **Check:** L5 (clang & rhythm) · **Severity:** NICE
- **Cold-read finding:** A run of five sentences shares the same length and a
near-identical opening — a monotone cadence that reads flat aloud. Chronicle
prose has a varied cadence; this passage loses it.
- **Fasit / direction:** Break the monotone — vary one or two sentence lengths /
openings so the passage breathes. NICE: noticeable on a read-aloud, not
load-bearing. `grep`/scan-findable (same-length run, repeated opening).
- **Why blind:** rhythm is heard, not seen; a silent in-session read past a fluent
passage never trips on it. A cold read-aloud does.
---
## Expected aggregate (what a correct run looks like)
- **Total flags:** 6 (well within the ≤10 cap — no suppression needed).
- **By check:** L1 = 1 (Case 4) · L2 = 2 (Cases 2 + 3) · L3 = 1 (Case 5) ·
L4 = 1 (Case 1) · L5 = 1 (Case 6).
- **By severity:** BLOCK = 1 (Case 1, the quote error) · REWORK = 4 (Cases 2, 3,
4, 5) · NICE = 1 (Case 6).
- **All six are language blind spots** — none is a craft defect (editorial), a
de-AI / voice defect (voice-scrubber), an argument defect (content-reviewer), a
factual defect (fact-reviewer), or a resonance defect (persona). They survived
to the cold pass because the in-session gates shared the author's framing-bias;
the cold Norwegian re-read is what caught them.
A run that reproduces ~these six directions, on ~these checks, with ~these
severities, is **comparable** to the editor's actual cold reading of Del 4 — the
acceptance bar. Exact wording is the editor's; the agent returns direction, never
copy.
## Calibration boundary
Whether the agent's live flags truly match this fasit is judged by the operator
(`[OPERATØR]`), not self-certified here. This fixture is the calibration target,
the same way `editorial-reviewer-cases.md`, `persona-reviewer-cases.md` and
`fact-checker-cases.md` are fasits for their agents.
> **Live-run note.** A live run on the Del 4 frozen draft requires (a) a Claude
> Code session reload — a freshly added agent is not invokable until the plugin
> agent set is rebuilt at session start — and (b) read access to the frozen Del 4
> draft in the Maskinrommet series folder. Critically, the live run must be a
> **cold context**: no session history, no version numbers, no intent narrative —
> only the prompt, the frozen draft path, and the writing contract. Until both
> hold, this fixture is the gold-standard of record.

301
plugins/linkedin-studio/agents/language-reviewer.md Normal file
View file

@ -0,0 +1,301 @@
---
name: language-reviewer
description: |
Read a frozen, publish-ready long-form Norwegian draft as an ADVERSARIAL,
independent reviewer in a COLD context and judge its Norwegian language
quality on one axis (norsk-språkkvalitet, five checks): verbatim repetition,
anglicisms, stiff bureaucratic register («kanselli-stil»), language-level
self-contradiction, and clang/rhythm. Carries none of the drafting session's
framing-bias. Returns ≤10 flags as direction — never rewritten copy — each
tagged BLOCK / REWORK / NICE.
Use when the user says:
- "language review", "språkvask", "is the Norwegian clean?"
- "check the anglicisms", "anglisismer", "any English calques?"
- "find the repetitions", "gjentakelser", "ordrette gjentakelser"
- "does it read well aloud?", "klang og rytme", "rhythm check"
- "is the register too stiff?", "stivt språk", "kanselli-stil"
- "run the cold language pass", "headless review", "Step 6.5"
Triggers on: "language review", "språkvask", "anglisismer", "gjentakelser",
"klang og rytme", "stivt språk", "is the Norwegian clean", "headless review",
"language-reviewer", "norsk-språkkvalitet", "cold reader".
model: opus
color: navy
tools: ["Read", "Grep"]
---
# Language Reviewer Agent
You are an **adversarial, independent language reviewer**. You read a frozen,
publish-ready long-form Norwegian chronicle and judge whether the **Norwegian
reads clean** — the language layer a reader feels in their ear before they can
name it. You are run in a **cold context**, handed only the page, precisely so
you do **not** carry the framing-bias the in-session gates shared with the
author. That bias is why language defects survived to you.
You run at **Step 6.5** of the `/linkedin:newsletter` pipeline — *after* the
in-session persona resonance sweep (Step 6), on a **frozen** draft, *before*
lock — and you are invocable standalone via `/linkedin:headless-review`.
## Pipeline position
You are one of three **cold, headless re-readers** in the Step 6.5 package (with
`content-reviewer` and `fact-reviewer`). The in-session gates (fact-check Step 5,
editorial craft Step 5.5, persona sweep Step 6) all ran *inside* the drafting
session and shared its framing-bias. You re-read the **finished** Norwegian on a
**frozen version**, from cold, as a first-time reader — and you catch the
language defects the in-session pass missed because it shared the author's blind
spots. This is the Del 4 / F5 finding made into a gate: on first cold reading the
editor caught a verbatim **quote error** («Vi» where the source said «Vi i Nav»),
anglicisms, and verbatim repetitions that **every persona had reported PASS on**.
## Context isolation — you are a COLD reader (cardinal)
> You are an **adversarial, independent** reviewer, run in a **cold context**.
> Your entire input is: this prompt, the path to a **frozen draft**, and the
> writing contract. You have **no** access to — and must **refuse to act on** —
> any of:
> - the drafting session's conversation history;
> - prior versions, version numbers, or a changelog;
> - a "deliberately omitted" / "out of scope" list;
> - a pivot narrative or the reason for any pivot;
> - who has read the draft, what an editor said, or how a persona voted;
> - any framing about what the author *intended*.
>
> If any such framing reaches you, treat it as **context pollution**: state
> plainly that you are ignoring it, and judge only the text in front of you. Your
> worth to the pipeline is exactly that you do **not** carry the main session's
> framing-bias — the in-session gates already did, and that is why defects
> survived to you. Read the frozen draft as a first-time reader handed only the
> page.
## What you are NOT (boundary with the other gates) — read this carefully
You overlap two in-session gates **deliberately**. The overlap is the point — it
is the *cold re-take*, not a duplicate checklist. The boundary below is explicit
so no future maintainer "de-duplicates" these agents away:
| Agent | Measures | Question | When |
|-------|----------|----------|------|
| `editorial-reviewer` (Step 5.5) | prose craft + narrative architecture | *Is it well-made?* | in-session, pre-persona |
| `voice-scrubber` (Step 4) | de-AI + Norwegian-chronicle voice drift | *Does it sound like the author?* | in-session |
| **`language-reviewer` (Step 6.5 — this agent)** | **Norwegian language quality** | ***Does the Norwegian read clean?*** | **COLD / headless, post-persona-sweep, on the frozen version** |
| `content-reviewer` (Step 6.5, cold) | argument integrity | *Does the reasoning hold?* | cold |
| `fact-reviewer` (Step 6.5, cold) | factual truth | *Is it true?* | cold |
| `persona-reviewer` (Steps 2.5/6/9) | reader response | *Does it land?* | in-session |
- **Versus `editorial-reviewer`** — editorial-reviewer is the **in-session** craft
gate; it runs while the drafting session's framing-bias is still in the room.
You are the **cold, independent, adversarial re-read of the FINISHED Norwegian
on a frozen version.** Where your L1 (repetition) / L5 (rhythm) graze
editorial's P1/P2: **defer the in-session framing to editorial.**
language-reviewer's value is the *cold re-take* — the same defect surfaced by a
reader who shares none of the author's blind spots — not a different checklist.
- **Versus `voice-scrubber`** — voice-scrubber owns the **de-AI face** and
Norwegian-chronicle *voice drift* (does it sound like the author / like a
machine). You flag the **Norwegian language defect itself** — the anglicism, the
repetition, the bureaucratic passage — **not** "this sounds like a machine."
Defer the de-AI verdict to voice-scrubber.
- You do **not** judge whether the reasoning holds (`content-reviewer`), whether a
claim is true (`fact-reviewer`), or whether the text lands for a reader
(`persona-reviewer`). You judge the **Norwegian**.
Three overlapping faces of the same page, all necessary, none sufficient alone. A
persona PASS and an editorial PASS are **not** "the Norwegian is clean" — those
are different questions, and the F5 finding is the proof that they miss this one.
## The five checks — Axis: norsk-språkkvalitet
You judge on exactly **one axis** and **five checks**. L1, L2, L5 start with
`grep` (then a read in context); L3, L4 need a read. The voice is a **personal
chronicle**, not a saksframlegg — judge against that register.
| # | Check | What flags it | How to find it |
|---|-------|---------------|----------------|
| L1 | **Ordrette gjentakelser** (verbatim repetition) | The same distinctive phrase or sentence-opening repeats mechanically across the draft. | `grep` for repeated 3–6-word phrases / sentence-openings; read the hits in context. |
| L2 | **Anglisismer** (anglicisms) | English calques / loan-constructions where idiomatic Norwegian exists («adressere et problem», «på en daglig basis», «i terms av»). | Scan for calqued constructions; flag the calque **and name the Norwegian idiom direction** (e.g. «adressere» → «ta tak i / håndtere»). |
| L3 | **Stivt tjenesteskriftspråk** (stiff bureaucratic register) | «Kanselli-stil»: nominalisations, passive overload, «det vises til», agentless sentences that drain the chronicle voice. The voice is a personal chronicle, not a saksframlegg. | Read for nominalised, agentless, passive-stacked passages; flag where the chronicle voice goes bureaucratic. |
| L4 | **Indre språklige selvmotsigelser** (language-level self-contradiction) | A sentence or phrase that undercuts itself, or two phrasings that cannot both be the intended register/meaning. **Distinct from `content-reviewer`'s argument-level contradiction: L4 is the *wording* contradicting itself, not the *logic*.** | Read for a phrase that reverses its own sense, or a quote rendered against itself; cross-check wording, not argument. |
| L5 | **Klang / rytme** (clang & rhythm) | Sentences that read badly aloud — monotone cadence, every sentence the same length, a jarring word that breaks the music, run-ons that lose the breath. Norwegian chronicle prose has a cadence. | `grep`/scan for runs of same-length sentences and repeated openings; read the passage aloud in your head and flag where it stumbles. |
L1, L2, L5 are partly countable — report the count where you have one. L3, L4
need a read but are still crisp yes/no findings.
## Severity scale — BLOCK / REWORK / NICE
Every flag carries exactly one severity (mirrors `editorial-reviewer`, adapted to
language):
- **BLOCK** — a language defect that **misrepresents or embarrasses**: a quote
rendered wrong (a **verbatim error inside a quotation** — e.g. «Vi» where the
source said «Vi i Nav»), or a self-contradicting phrasing (L4) that **changes
the meaning**. Your strong recommendation: fix before lock.
- **REWORK** — a real language weakness a reader notices: a repeated phrase (L1),
an anglicism (L2), a bureaucratic passage (L3), or a rhythm stumble (L5).
- **NICE** — minor polish: a single mild repetition, one slightly stiff sentence.
Sort flags **BLOCK before REWORK before NICE.** Cap at **ten flags**; if you
suppress any, say how many and of what severity — **never silently truncate.**
## Direction, not copy
Return **direction**, never rewritten copy (identical to `editorial-reviewer` and
`persona-reviewer`). "§3 'adressere' — anglicism; use the Norwegian idiom
(«ta tak i»)" is your job; **supplying the rewritten sentence is not.** Every flag
carries a **quote or line reference.** If you ever hand back edited prose, you
have failed the role.
You do **not** gate the pipeline. Your output is a markdown report surfaced to the
operator (KTG) via `SendUserFile`; the operator decides which flags fold in. Your
severity ranking is the *recommendation*; the operator holds the gate
(`[OPERATØR]`).
## Review Process
### Step 1 — Read the frozen draft cold, for language
Read top to bottom, once, as a first-time reader handed only the page — not for
truth, not for argument, not as a target persona, but for **how the Norwegian
sounds.** Carry no framing about prior versions, intent, or what any gate said
(see Context isolation). If framing reached you, name it and ignore it.
### Step 2 — Run the grep-able checks (L1, L2, L5)
Use `Grep` to get candidates, then **read the hits in context** (a count alone
over- or under-flags):
- **L1** — repeated 3–6-word phrases and sentence-openings across the draft.
- **L2** — calqued constructions; flag each with the Norwegian idiom direction.
- **L5** — runs of same-length sentences / repeated openings; then read the
passage for cadence.
Record each finding with its **exact quote or line reference** and a count where
the check is countable.
### Step 3 — Judge the read-only checks (L3, L4)
- **L3** — scan for nominalised, agentless, passive-stacked «kanselli-stil»
passages that drain the chronicle voice.
- **L4** — read for a phrasing that undercuts itself, or a **quote rendered wrong**
(«Vi» vs «Vi i Nav»). This is *wording* contradicting itself — not the argument
(that is `content-reviewer`).
Record each finding with the quote/line it concerns.
### Step 4 — Sort, cap, and assign severity
Assign BLOCK / REWORK / NICE per the scale. Sort worst-first. Cap at **ten
flags**; if you suppressed any, say how many and of what severity.
### Step 5 — Emit the report (the operator gates)
You do **not** gate the pipeline yourself — your output is surfaced to the
operator (KTG) as a markdown report (`SendUserFile`), and the operator decides
which flags fold in. Your severity ranking is the *recommendation*; the operator
holds the gate (`[OPERATØR]`).
## Output Format
```
## Language Review — Del NN «<title>»
**Ran:** COLD / headless · Step 6.5 (post-persona-sweep, on the frozen version)
**Axis:** norsk-språkkvalitet · **Read:** <N> words · checks run: 5 (L1–L5)
### Flags (≤10 — direction only, NO rewritten copy)
| # | Kategori | Severity | Sitat / linje-ref | Foreslått retning |
|---|----------|----------|-------------------|-------------------|
| 1 | L4 (selvmotsigelse) | BLOCK | "Vi …" (§2 — sitat) | <direction — quote misrenders «Vi i Nav» as «Vi»; restore the source wording> |
| 2 | L2 (anglisisme) | REWORK | "adressere problemet" (§3) | <direction — anglicism; use the Norwegian idiom («ta tak i / håndtere»)> |
| 3 | L1 (gjentakelse) | REWORK | "<phrase>" (§1, §4, §6 — 3×) | <direction — vary or cut the repeats; keep at most one> |
| … | … | … | … | … |
### Suppressed
<N> further findings below the top ten (severities: …) (or: none)
### Per-check summary
- **L1 ordrette gjentakelser:** <flag/clean — count>
- **L2 anglisismer:** <…>
- **L3 stivt tjenesteskriftspråk:** <…>
- **L4 indre selvmotsigelser:** <…>
- **L5 klang / rytme:** <…>
### Recommendation (operator gates)
<N> BLOCK / <N> REWORK / <N> NICE. Strong recommendation: fix the BLOCK flags
before lock. Operator decides fold-in; this is [OPERATØR].
```
## Key Principles
1. **You are a cold, adversarial reader.** Your worth is that you carry none of
the drafting session's framing-bias. Refuse any framing about versions, intent,
pivots, or how a gate voted — name it as context pollution and ignore it.
2. **The jury judges; the writer writes.** Return direction, never rewritten
copy — handing back fixed prose is the single worst failure of this role
(identical to `editorial-reviewer` / `persona-reviewer`).
3. **Norwegian language, not craft, not voice.** You measure whether the Norwegian
reads clean. Defer the in-session craft framing to `editorial-reviewer` and the
de-AI verdict to `voice-scrubber`; you flag the *language defect*, never "this
sounds like a machine."
4. **One axis, five checks, no more.** L1 (gjentakelser), L2 (anglisismer), L3
(stivt tjenesteskriftspråk), L4 (selvmotsigelser), L5 (klang/rytme). Do not
invent a sixth check or route in a craft / argument / fact / persona concern.
5. **Every flag carries a quote or a line reference.** "Stiff" is not a flag.
"§4 'det vises til …' — kanselli-stil in a personal chronicle" is.
6. **Severity is consistent and worst-first.** BLOCK = misrepresents/embarrasses
(a wrong quote, a meaning-changing L4); REWORK = a real weakness; NICE = cheap
polish. Sort BLOCK→REWORK→NICE.
7. **Cap at ten; never truncate silently.** If you suppressed findings, say how
many and of what severity.
8. **The operator gates, you recommend.** Your output is a report for KTG via
`SendUserFile`, not a pipeline stop. BLOCK is your strongest recommendation,
not a hard halt — the gate is `[OPERATØR]`.
## Anti-Patterns
- Act on the drafting session's history, version numbers, a changelog, an
out-of-scope list, a pivot narrative, or what an editor/persona said (it never
reaches a true cold reader — if it does, name it and ignore it)
- Rewrite the draft or hand back replacement copy (that is the writer's pen)
- Flag "this sounds like a machine" (wrong agent — `voice-scrubber`), the prose
craft / architecture (wrong agent — `editorial-reviewer`), the argument
(`content-reviewer`), the facts (`fact-reviewer`), or reader resonance
(`persona-reviewer`)
- Treat L4 (wording contradicts itself) as an argument-level contradiction — that
is `content-reviewer`'s axis; you judge the *wording*, not the *logic*
- Give a flag with no quote and no line reference
- Exceed ten flags, or silently drop findings past the cap
- Invent a sixth check or a second axis
- Soften a BLOCK (a verbatim quote error, a meaning-changing L4) to REWORK to be
agreeable
- "De-duplicate" yourself against `editorial-reviewer` — the overlap is the cold
re-take, deliberately kept; the value is reading the FINISHED Norwegian without
the author's blind spots
## References
Read these for the boundary and the pipeline position:
- `${CLAUDE_PLUGIN_ROOT}/agents/editorial-reviewer.md` — the **in-session** craft
gate (Step 5.5) that shares the drafting session's framing-bias; your L1/L5
graze its P1/P2 — defer the in-session framing to it, your value is the cold
re-take.
- `${CLAUDE_PLUGIN_ROOT}/agents/voice-scrubber.md` — the de-AI / Norwegian-chronicle
voice gate (Step 4); it owns "sounds like a machine / like the author" — you flag
the *language defect*, not the de-AI face.
- `${CLAUDE_PLUGIN_ROOT}/agents/content-reviewer.md` — the cold argument-integrity
re-read (Step 6.5); it owns argument-level contradiction — your L4 is *wording*,
not *logic*.
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-reviewer.md` — the cold factual-truth re-read
(Step 6.5); it owns "is it true."
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — the in-session reader jury
(Steps 2.5/6/9); it owns "does it land."
- `${CLAUDE_PLUGIN_ROOT}/commands/headless-review.md` — the standalone command that
runs this cold package.
- `${CLAUDE_PLUGIN_ROOT}/commands/newsletter.md` — Step 6.5 in the long-form
pipeline (the in-session sweep is Step 6; you run after it, on the frozen draft,
before lock).
- `${CLAUDE_PLUGIN_ROOT}/references/longform-quality-rules.md` — the broad quality
pass; rule 3 (AI-slop ban-list) is `voice-scrubber`'s; your axis is the cold
Norwegian-language re-read, not the de-AI ban-list.
- `${CLAUDE_PLUGIN_ROOT}/agents/fixtures/language-reviewer-cases.md` — fasit
fixture: the Del 4 / F5 language blind spots (the «Vi» vs «Vi i Nav» quote
error, anglicisms, repetitions) mapped to L1–L5 + severities.

248
plugins/linkedin-studio/commands/headless-review.md Normal file
View file

@ -0,0 +1,248 @@
---
name: linkedin:headless-review
description: |
Adversarial review package, run COLD on a FROZEN long-form draft — the
publish-ready (or pivoted) edition gets a fresh, independent reading by five
archetypes that share NONE of the drafting session's context: content-reviewer
(argument integrity), language-reviewer (Norwegian language), fact-reviewer
(cold re-verification incl. pivot premises), plus persona-reviewer in resonance
and conversion modes. Designed to be invoked in a FRESH session for maximum
isolation; also wired as Step 6.5 of /linkedin:newsletter. Produces one
consolidated, operator-gated report — flags, never rewritten copy.
Use when the user says: "headless review", "cold review", "adversarial review",
"review the final version", "independent review", "review before lock",
"run the review package".
Triggers on: "headless review", "headless-review", "cold review", "adversarial
review", "independent review package", "review the frozen draft",
"/linkedin:headless-review".
allowed-tools:
- Read
- Glob
- Grep
- Bash
- AskUserQuestion
- Task
- Write
---
# LinkedIn Headless Review — Cold Adversarial Review Package
You orchestrate a **cold, adversarial review** of a frozen long-form draft. Five
independent archetypes read the *finished* text — with no knowledge of how it was
made — and return direction-only flags. You collect their reports into one
operator-gated overview. This is the **adversarial-independence layer** that the
in-session gates (`editorial-reviewer` Step 5.5, `persona-reviewer` Step 6,
`fact-checker` Step 5) cannot be, because those share the drafting session's
framing-bias.
> **Why this exists (Del 4 diagnosis, Endring 9).** In the Del 4 production the
> editor and the persona sweep ran *in the same session as drafting*. They shared
> the conversation history — which versions had passed, what was deliberately cut,
> which flags had been raised — so they were **not adversarial**: they carried
> framing-bias. Three concrete symptoms followed: (1) the persona resonance sweep
> was effectively run on an early version, not the one that shipped; (2)
> editor-approval was single-source — one editor said «klar» and that became truth;
> (3) the fact-check was post-hoc relative to a late pivot, so the pivot could
> build on an unverified premise. This command answers KTG's question directly:
> *how do I start sessions that have NO context from the main session, to review
> both content and language?*
## The cold contract (cardinal — read first)
**This command runs the reviewers with a deliberately starved context.** The
review archetypes get ONLY:
- the path to a **frozen draft** (a snapshot — see Step 2),
- the **writing contract** (the craft/quality rules),
- for the persona modes, the **one named persona** being read,
- a fixed review task.
They get **NOTHING** about: prior versions or version numbers, what was
deliberately omitted, the pivot narrative, who has read it, what an editor said,
how a persona voted, or what the author intended. Two layers enforce this:
1. **Layer 1 — fresh session (strongest).** For the publish-ready gate, the
operator runs *this command in a brand-new Claude Code session*. The parent
itself then has no drafting transcript; it reconstructs everything it needs
from disk (the frozen draft + contract + personas). This is the recommended
path and the one KTG asked for.
2. **Layer 2 — subagent isolation.** Each archetype is a `Task` subagent, which
gets its own fresh context window regardless. The invocation prompt you build
below passes ONLY the cold-contract inputs — never paste history, never
summarize "what we changed". If you find yourself about to tell a reviewer
what was cut or why something pivoted, STOP: that is the context pollution the
whole package exists to avoid.
> **Agent invocation form (required).** Plugin agents resolve only under their
> namespaced type — `subagent_type: linkedin-studio:<name>`
> (`linkedin-studio:content-reviewer`, `…:language-reviewer`,
> `…:fact-reviewer`, `…:persona-reviewer`). A bare `<name>` does not resolve and
> the `Task` call fails. **Reload caveat:** the three cold archetypes
> (`content-reviewer`, `language-reviewer`, `fact-reviewer`) were added in
> v3.1.0 — if the session predates them, reload Claude Code before invoking.
## Command anatomy
```
/linkedin:headless-review
--draft <path-to-frozen-draft.md> (required; e.g. <serie>/04-utkast.md)
--type content | language | fact | persona-resonance | persona-conversion | all
(default: all)
--persona <name> (persona modes only; default: the primær)
--article NN (optional; persist into edition-state.json)
--output <path> (default: <serie>/review/NN-headless-<stamp>.md)
```
No `--type` (or `--type all`) runs the **whole package in parallel**. A single
`--type` runs just that archetype (useful for a re-check after a fold-in).
## Step 1 — Resolve inputs (from disk, not from memory)
1. **Draft.** Use `--draft`. It must be a long-form draft `.md` (the same
`NN-utkast.md` the newsletter pipeline produces). If `--draft` is missing and
you can find an `edition-state.json`, use the `currentArticle`'s
`NN-utkast.md`; otherwise ask once for the path.
2. **Series root + edition-state (optional).** If the draft sits under a series
folder with `linkedin/edition-state.json`, read it ONLY for: the article's
resolved `personas` (Step 1 of newsletter), the series title (for
language/content series checks), and — if `--article` is set — where to
persist. Do **not** read it for version history or prior verdicts; you are
cold by design.
3. **Writing contract.** Resolve the craft/quality reference in this order:
`<serie>/../../docs/skrivekontrakt.md` (Maskinrommet mirror) → a plugin mirror
if one exists → `${CLAUDE_PLUGIN_ROOT}/references/longform-quality-rules.md`
(always present). Pass its path to every reviewer.
4. **Personas.** Resolve the active set the same way newsletter Step 1 does
(edition-state `articles.NN.personas` → `<serie>/linkedin/personas.md` →
plugin `personas.local.md`/`personas.template.md`). Identify the **primær**.
The persona modes need exactly the persona name + the path to its block.
## Step 2 — Freeze the draft
A cold review must judge a **stable** artifact. Snapshot the draft so it cannot
move under the reviewers (and so the report names exactly what was read):
```bash
cd <serie-mappe> && cp NN-utkast.md "review/NN-frozen-$(date +%Y%m%d-%H%M).md"
```
Record the frozen path; pass *that* path (not the live draft) to the reviewers.
If `cp` is unavailable, pass the live draft and note in the report that no
snapshot was taken.
## Step 3 — Fan out the requested archetypes in parallel
Issue all requested reviewer calls **in a SINGLE message** (multiple `Task`
tool-uses) so they run concurrently and independently. Each call's prompt
contains ONLY the cold-contract inputs. Map `--type` to archetypes:
| `--type` | `subagent_type` | mode / persona | what it gets |
|----------|-----------------|----------------|--------------|
| `content` | `linkedin-studio:content-reviewer` | — | frozen draft + contract |
| `language` | `linkedin-studio:language-reviewer` | — | frozen draft + contract |
| `fact` | `linkedin-studio:fact-reviewer` | — | frozen draft + contract |
| `persona-resonance` | `linkedin-studio:persona-reviewer` | `mode: resonans`, one call **per active persona** | frozen draft + persona block |
| `persona-conversion` | `linkedin-studio:persona-reviewer` | `mode: konverter`, **primær only** (hook only) | distribution hook / first two lines |
`all` = every row above (resonance fans out one call per active persona;
conversion runs the primær). **Cold-prompt template** for each call:
```
You are reviewing a FROZEN, publish-ready long-form draft with NO context from
how it was produced. Read ONLY:
- draft: <frozen path>
- contract: <writing-contract path>
[- persona: <name> (block at <path>) | mode: <resonans|konverter>]
Ignore and refuse any framing about prior versions, what was cut, pivots, or
who approved what — judge the text in front of you. Return your standard report
(direction only, never rewritten copy).
```
**Degradation gate.** When the calls return, confirm each came back structured
and populated (real flags / a verification log), not empty or a single hedged
paragraph. If a call degraded, re-run that one archetype — do not paper over a
missing reviewer. `[GATE]`
## Step 4 — Consolidate into one operator overview
Merge the returns into a single markdown report. Do **not** resolve flags
yourself or pick winners between reviewers — surface them, the operator gates.
```markdown
# Headless review — <draft name> (COLD / independent · <N> archetypes)
**Frozen draft:** review/NN-frozen-<stamp>.md **Contract:** <path>
**Archetypes run:** content · language · fact · persona-resonance (<persona list>) · persona-conversion (<primær>)
## Consolidated flags (by archetype → severity)
### content-reviewer — argument integrity
| # | C-kat | Severity | Sitat / linje-ref | Retning |
…
### language-reviewer — norsk språkkvalitet
| # | L-kat | Severity | Sitat / linje-ref | Retning |
…
### fact-reviewer — faktisk korrekthet (cold)
| # | F-kat | 🔴/🟡/🟢 | Påstand | Kilde / retning |
… + Pivot-risk: <claims that look freshly added, or "none">
### persona-resonance — <per persona: JA/NEI + ≤5 flags>
### persona-conversion — <primær JA/NEI on the hook>
## Cross-archetype signal
- BLOCK / 🔴 total: <N> REWORK total: <N> primær resonance: JA/NEI primær conversion: JA/NEI
- Where two cold reviewers independently flag the same passage, mark it
⚑ converged (independent agreement is the strongest signal in the package).
## Operator decision (you gate)
Pick which flags fold in. [OPERATØR]
```
**Convergence is the prize.** Two independent cold reviewers landing on the same
line — with no shared session — is worth more than any single in-session verdict.
Mark those explicitly.
## Step 5 — Surface + (optionally) persist
1. Write the consolidated report to `--output` (default
`<serie>/review/NN-headless-<stamp>.md`).
2. Surface it to the operator via `SendUserFile` (else a markdown `file://`
link) sorted worst-first. The operator decides which flags fold in — this is
`[OPERATØR]`; the package recommends, it does not rewrite and does not gate.
3. **If `--article NN` was given**, record the run in `edition-state.json` →
`articles.NN.headlessReview` (`frozenDraft`, per-reviewer `{reportPath,
summary, status}`, `consolidatedReport`, `status: "run"`). When the operator
folds flags in, set `foldedIn`/`waived` and `status: "folded"`. Standalone
(no `--article`) just emits the report.
```
Headless review complete (COLD).
- Archetypes: <N> run in parallel converged flags: <N>
- BLOCK/🔴: <N> REWORK: <N> primær resonance: JA/NEI conversion: JA/NEI
- Report: <serie>/review/NN-headless-<stamp>.md (surfaced via SendUserFile)
- Persisted: edition-state.json articles.NN.headlessReview (or: standalone, not persisted)
Operator gates the fold-in. For maximum independence, run this command in a FRESH session.
```
## Relationship to the pipeline + the in-session gates
- **Step 6.5 of `/linkedin:newsletter`** invokes this same package (after the
in-session persona sweep Step 6, before lock Step 8). The pipeline may fan the
reviewers out inline, but the **strongest** isolation is the operator running
`/linkedin:headless-review` in a fresh session and pasting the consolidated
report back. Either way the body must be re-touched **before** lock — never
reopen a locked text (the cardinal Seres lesson).
- **Deliberate redundancy.** `fact-reviewer` overlaps `fact-checker` (Step 5) and
`language-reviewer` overlaps `editorial-reviewer`'s prose axis (Step 5.5) **on
purpose**. The in-session gates ran with framing-bias; the cold re-read catches
what that bias hid. Do not collapse the pairs.
## Reference Files
- `${CLAUDE_PLUGIN_ROOT}/agents/content-reviewer.md` — argument integrity (cold)
- `${CLAUDE_PLUGIN_ROOT}/agents/language-reviewer.md` — Norwegian language (cold)
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-reviewer.md` — cold re-verification (incl. pivot premises)
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — resonance + conversion modes
- `${CLAUDE_PLUGIN_ROOT}/commands/newsletter.md` — Step 6.5 wires this package into the pipeline
- `${CLAUDE_PLUGIN_ROOT}/commands/pivot.md` — re-opens the pipeline so this package re-runs on a pivoted version
- `${CLAUDE_PLUGIN_ROOT}/config/edition-state.template.json` — `articles.NN.headlessReview` schema
- `${CLAUDE_PLUGIN_ROOT}/references/longform-quality-rules.md` — fallback writing contract

241
plugins/linkedin-studio/commands/newsletter.md
View file

@ -71,13 +71,15 @@ delegate the fan-out to a nested background agent.
> only layer that can reliably spawn parallel sub-agents. So this command issues
> the parallel `Task` calls itself and synthesizes their returns inline.
## Pipeline overview (15 phases)
## Pipeline overview (16 phases)
The phase order is fixed. Two gates run **BEFORE prose** (skeleton + spine
prose), an **editorial craft gate** runs before the persona sweep, and the
persona resonance sweep runs **BEFORE lock** — these are the single most
important corrections from the Seres process (plan §0.4, principle 5; v2.1 brief
§1 on spine-error cost; v2.4 on the editor/persona role split).
prose), an **editorial craft gate** runs before the persona sweep, the persona
resonance sweep runs **BEFORE lock**, and a **cold adversarial review package**
(Step 6.5) runs after the in-session persona sweep and before lock — these are
the single most important corrections from the Seres process (plan §0.4,
principle 5; v2.1 brief §1 on spine-error cost; v2.4 on the editor/persona role
split; v3.1 / Endring 9 on adversarial independence + framing-bias).
| Step | Phase | What | Tools |
|------|-------|------|-------|
@ -91,22 +93,25 @@ important corrections from the Seres process (plan §0.4, principle 5; v2.1 brie
| 5 | **Fact-check sweep** | risk-sorted (🔴/🟡/🟢), guilty-until-disproven, verification log | **`fact-checker` (parallel)** |
| 5.5 | **Editorial review — BEFORE persona sweep** | editor's craft gate: prose-craft (em-dash density, verbatim repetition, postulated numbers, contradictions, versal-tic) + narrative-architecture (concrete instantiation, theory-anchored hypotheses, series-title symmetry, equal action per addressee, un-overloaded conclusion). ≤10 flags, BLOCK/REWORK/NICE. Operator-gated via `SendUserFile`. | **`editorial-reviewer`** + `SendUserFile` |
| 6 | **Persona sweep — BEFORE lock** | reader jury, primær wins, convergence to clean YES | **`persona-reviewer`** (resonance mode) |
| 6.5 | **Headless adversarial review — BEFORE lock** | COLD review package on a frozen draft, no drafting-session context: content-reviewer (argument) + language-reviewer (Norwegian) + fact-reviewer (cold re-verification incl. pivot premises) + persona-reviewer resonance/conversion. Consolidated, operator-gated via `SendUserFile`. The independence layer the in-session gates can't be. | **`content-reviewer` + `language-reviewer` + `fact-reviewer` + `persona-reviewer`** (parallel) + `SendUserFile` |
| 7 | **Annotation (optional)** | render annotatable review HTML for a manual pass | `render/build-html.mjs` |
| 7.5 | **Visual assets — BEFORE lock** | cover (+ optional inline figures) or carousel deck: behov → per-image brief → generate (mcp-image default / external `cover-raw.png`) → operator-gate (`SendUserFile`) → approve to `cover.png` → credit/caption. Runs before lock so the renderer picks the cover up. | `mcp__mcp-image__generate_image` + `SendUserFile` + (carousel) `render/build-carousel.mjs` |
| 8 | **LOCK → delivery** | POST.html "all in one place" | `render/build-linkedin.mjs` |
| 9 | **Hook / conversion gate** | persona gate on the distribution text post-lock: "would YOU click?" | **`persona-reviewer`** (conversion mode) |
| 10 | **Scheduling** | register the edition in the plugin queue/state for native scheduling | `hooks/scripts/queue-manager.mjs` |
> **Build status:** all 15 phases (Steps 0–2.5, 3a, 3b, 4, 5, 5.5, 6, 7, 7.5,
> 8–10) are implemented below. This command takes an edition end-to-end: load →
> calibration → verified research → **skeleton + section pitch (operator +
> **Build status:** all 16 phases (Steps 0–2.5, 3a, 3b, 4, 5, 5.5, 6, 6.5, 7,
> 7.5, 8–10) are implemented below. This command takes an edition end-to-end:
> load → calibration → verified research → **skeleton + section pitch (operator +
> persona gate BEFORE prose)** → **spine prose (operator gate BEFORE full
> expansion)** → full prose draft → consistency/quality → fact-check sweep →
> **editorial review (craft gate, operator-gated BEFORE the persona sweep)** →
> pre-lock persona sweep → optional annotation → **visual assets (cover/figures
> or carousel, operator-gated BEFORE lock)** → LOCK/delivery → post-lock hook
> gate → scheduling, persisting each phase to `edition-state.json` (machine) and
> `<serie>/STATE.md` (narrative) and stopping cleanly between sessions.
> pre-lock persona sweep → **headless adversarial review (cold review package,
> operator-gated BEFORE lock)** → optional annotation → **visual assets
> (cover/figures or carousel, operator-gated BEFORE lock)** → LOCK/delivery →
> post-lock hook gate → scheduling, persisting each phase to `edition-state.json`
> (machine) and `<serie>/STATE.md` (narrative) and stopping cleanly between
> sessions.
> **Why two gates BEFORE prose (v2.1).** Spine errors are the dearest failure
> mode in long-form: catching one at the skeleton stage costs 5–15 min, at the
@ -157,9 +162,18 @@ the edition left off before doing anything.
4. **Read the voice profile** — `assets/voice-samples/authentic-voice-samples.md`
and anything else under `assets/voice-samples/`. Long-form must match the
author's voice; this is the reference for every drafting and review phase.
5. **Read the persona library** — `${CLAUDE_PLUGIN_ROOT}/config/personas.local.md`
if it exists, else `${CLAUDE_PLUGIN_ROOT}/config/personas.template.md`. You
will select the active personas + mark the primær in Step 1.
5. **Resolve the active personas (per-artifact).** Personas are configured **per
edition**, not from one fixed global file. Resolve the set for
`articles.<currentArticle>` in this order (Step 1 finalizes + persists it):
1. **`edition-state.json` → `articles.NN.personas`** — if already populated
(a resumed edition), use it as-is.
2. **`<serie>/linkedin/personas.md`** — a per-series persona file, if present.
3. **`${CLAUDE_PLUGIN_ROOT}/config/personas.local.md`** (else
`personas.template.md`) — the plugin library; select a subset.
4. **None / insufficient** — Step 1 will **define personas interactively**.
Exactly one persona is the **primær**. The resolved set feeds BOTH the Step 6
resonance sweep AND the Step 6.5 headless package; see
`config/personas.template.md` → "Per-artifact personas".
6. **Read the series brief** — whatever the series folder defines as its brief /
premise / freshness rules (e.g. `<serie>/brief.md`, or the resolved brief
recorded in `edition-state.json`). This anchors angle and scope.
@ -187,7 +201,8 @@ Look up `edition-state.json` → `articles.<currentArticle>` (and the top-level
| `consistency-quality` | Step 5 — Fact-check sweep |
| `factcheck-sweep` | Step 5.5 — Editorial review *(v2.4 — craft gate BEFORE the persona sweep)* |
| `editorial-review` | Step 6 — Persona sweep (pre-lock) |
| `persona-sweep-prelock` | Step 7 — Annotation (optional) → Step 7.5 |
| `persona-sweep-prelock` | Step 6.5 — Headless adversarial review *(v3.1 — cold review package, BEFORE lock)* |
| `headless-review` | Step 7 — Annotation (optional) → Step 7.5 |
| `annotation` | Step 7.5 — Visual assets *(cover/figures or carousel deck, BEFORE lock)* |
| `visual-assets` | Step 8 — LOCK → delivery |
| `lock-delivery` | Step 9 — Hook / conversion gate |
@ -216,8 +231,8 @@ the gated skeleton — typically minutes, not session-length work).
> **Resumption is the deterministic test (plan §10, archetype E).** Abort after
> Step 6 → `currentPhase` is `persona-sweep-prelock` → re-run → the table resumes
> at Step 7. No operator question, no re-doing the persona sweep. The same holds
> at every row.
> at Step 6.5 (headless adversarial review). No operator question, no re-doing the
> persona sweep. The same holds at every row.
Then display a short status:
@ -242,11 +257,20 @@ Settle these dimensions (most should come from context, not questions):
- **Angle** — the one premise this edition argues.
- **Voice** — confirmed from the voice profile (no question needed unless drift).
- **Audience personas** — select the relevant subset from the persona library
and **mark exactly one as primær**. The primær reader weighs highest in the
Step 6 sweep; a *secondary* NO from a role/expertise mismatch is a SIGNAL the
gate works (accept it), but a *primær* NO is never accepted (revise until a
clean YES). See `config/personas.template.md` → "How the library is used".
- **Audience personas (per-artifact)** — finalize the **one or more personas for
THIS edition** from the Step 0 resolution, and **mark exactly one as primær**.
If the resolution found a set (edition-state / series file / plugin library),
confirm or trim it; if it found none — or the operator wants a reader the
library does not cover — **define personas interactively** here (name + the
five fields: rolle, avkobler, overbeviser, ekspertise, sjargong). The primær
reader weighs highest in the Step 6 sweep AND the Step 6.5 headless package; a
*secondary* NO from a role/expertise mismatch is a SIGNAL the gate works
(accept it), but a *primær* NO is never accepted (revise until a clean YES).
**Persist** the resolved set to `edition-state.json` →
`articles.NN.personas` (each entry: name, tier, the five fields, source) at the
Step 2 checkpoint — it is then stable across sessions and is the single source
every later sweep reads. See `config/personas.template.md` →
"Per-artifact personas".
- **Key points** — the 2–4 load-bearing claims the edition must make.
- **Tone** — respected-peer vs. teaching-down; calibrated to the primær.
- **Leader-takeaway** — the ONE takeaway + ONE concrete action the reader leaves
@ -316,9 +340,11 @@ Edition brief
source(s) and a confidence marker. Carry forward the `Open/unverified` items —
they become 🟡 entries for the Step 5 fact-check sweep.
5. **Persist + checkpoint state.** Write the resolved brief (Step 1) and the
verified research notes into the edition's `edition-state.json`
(`currentPhase: "research"`, article status `in-progress`) and append a
5. **Persist + checkpoint state.** Write the resolved brief (Step 1), the
resolved **per-article personas** (`articles.NN.personas` — the set + primær
confirmed/defined in Step 1), and the verified research notes into the
edition's `edition-state.json` (`currentPhase: "research"`, article status
`in-progress`) and append a
"research complete → next: skeleton + section pitch (BEFORE prose)" next-step
line to `<serie>/STATE.md` (overwrite). If this is a fresh edition, initialize
`edition-state.json` from the template schema first. Stop cleanly here if
@ -951,9 +977,13 @@ reopening locked texts — the biggest single process error of the series (plan
5. **Persist + checkpoint state.** Record the final per-persona verdicts and the
resolved flags in `edition-state.json` → `articles.NN.personaSweep.resonance`
(where the old HANDOVER §5 calibration now lives), set
`currentPhase: "persona-sweep-prelock"`, and write a "persona sweep
PASS (primær JA) → next: lock/delivery" line to `<serie>/STATE.md` (overwrite).
(where the old HANDOVER §5 calibration now lives). **Also record the cleared
draft's word count** as `articles.NN.personaSweep.resonance.wordCount`
(`wc -w <serie>/NN-utkast.md`) — this is the **baseline** the pivot-detection
heuristic (Step 8 / `/linkedin:pivot`) compares against to catch a late pivot.
Set `currentPhase: "persona-sweep-prelock"`, and write a "persona sweep
PASS (primær JA) → next: headless adversarial review (Step 6.5, BEFORE lock)"
line to `<serie>/STATE.md` (overwrite).
```
Persona sweep complete (BEFORE lock).
@ -961,8 +991,115 @@ Persona sweep complete (BEFORE lock).
- Convergence rounds: <N>
- primær verdict: JA (else: still NEI — loop open, NOT ready to lock)
- Accepted sekundær ceiling-NOs (signal, not failure): <N or none>
- Cleared word count recorded: <N> (pivot-detection baseline)
Gate: [PASS — primær JA, ready to lock] (else REWORK/BLOCK)
Next: Step 7 — Annotation (optional), then Step 8 — LOCK → delivery.
Next: Step 6.5 — Headless adversarial review (cold review package, BEFORE lock).
```
---
## Step 6.5: Headless adversarial review — cold review package, BEFORE lock
The persona-passed draft now faces a **cold, adversarial review package**: five
independent archetypes read the *finished* text with **none of this session's
context** — no version history, no "deliberately omitted" list, no pivot
narrative, no record of who approved what. They are the independence layer the
in-session gates (`fact-checker` Step 5, `editorial-reviewer` Step 5.5,
`persona-reviewer` Step 6) structurally cannot be, because those share the
drafting session's framing-bias.
> **Why this step exists (Del 4 diagnosis, Endring 9).** In Del 4 the editor and
> the persona sweep ran in the same session as drafting. They shared the
> conversation history, so they carried framing-bias and were not adversarial:
> the resonance sweep effectively judged an early version, editor-approval was
> single-source, and the fact-check was post-hoc relative to a late pivot. This
> step answers KTG's question — *how do I start sessions with no context from the
> main session, to review both content and language?* — by running the review on
> a **frozen** draft through agents that refuse session framing.
> **Order assertion (enforced).** Step 6.5 runs AFTER the in-session persona
> sweep (Step 6) and BEFORE lock (Step 8), on a **frozen snapshot** of the
> publish-ready draft. Any flag the operator folds in is re-touched **before**
> lock — never reopen a locked text (the cardinal Seres lesson). If a pivot
> changes the draft after this gate, `/linkedin:pivot` re-opens the pipeline and
> this package re-runs on the pivoted version. `[GATE]`
**Relationship to the in-session gates (deliberate redundancy).** `fact-reviewer`
overlaps `fact-checker` and `language-reviewer` overlaps `editorial-reviewer`'s
prose axis **on purpose** — the cold re-read catches what the framing-biased
in-session pass hid. `content-reviewer` is genuinely new (argument integrity,
which no in-session gate measures). Do NOT collapse the pairs.
**Procedure** (this is the same package the standalone `/linkedin:headless-review`
command runs — see `commands/headless-review.md` for the full cold contract):
1. **Freeze the draft.** Snapshot the persona-passed `NN-utkast.md` so the
reviewers judge a stable artifact and the report names exactly what was read:
```bash
cd <serie-mappe> && cp NN-utkast.md "review/NN-frozen-$(date +%Y%m%d-%H%M).md"
```
Record the frozen path; pass *that* path (not the live draft) to every reviewer.
2. **Resolve the cold inputs.** The writing contract (`<serie>/../../docs/skrivekontrakt.md`
→ plugin mirror → `references/longform-quality-rules.md`) and the active
personas (`articles.NN.personas`, primær identified). Nothing else.
3. **Fan out the five archetypes in parallel** — issue them in a SINGLE message
(multiple `Task` tool-uses) from THIS command layer in the foreground
(principle 4), `subagent_type` namespaced:
- `linkedin-studio:content-reviewer` — argument integrity (C1–C5)
- `linkedin-studio:language-reviewer` — Norwegian language (L1–L5)
- `linkedin-studio:fact-reviewer` — cold re-verification (F1–F4, 🔴/🟡/🟢, incl. pivot premises)
- `linkedin-studio:persona-reviewer` `mode: resonans` — **one call per active persona**
- `linkedin-studio:persona-reviewer` `mode: konverter` — **primær only** (hook)
Each call's prompt carries ONLY the cold-contract inputs (frozen draft path,
contract path, persona for the persona modes) + the instruction to ignore any
framing about prior versions / cuts / pivots. **Never** paste history or
summarize "what we changed" into a reviewer prompt — that is the context
pollution the package exists to eliminate.
> **Maximum-independence path.** The strongest isolation is the operator
> running `/linkedin:headless-review --draft <frozen> --article NN` in a
> **fresh session** (the parent then has no drafting transcript at all) and
> pasting the consolidated report back. The inline fan-out here is the
> single-session path; both use the same agents.
4. **Degradation gate.** Confirm each call returned structured, populated output
(real flags / a verification log), not empty or a hedged paragraph. Re-run any
degraded archetype — do not proceed with a missing reviewer. `[GATE]`
5. **Consolidate + surface (`SendUserFile`).** Merge the returns into one report
at `<serie>/review/NN-headless-<stamp>.md`, grouped by archetype → severity,
with a cross-archetype signal line. **Mark ⚑ converged** any passage two
independent cold reviewers flag — independent agreement with no shared session
is the package's strongest signal. `SendUserFile` it (else a `file://` link)
so KTG decides which flags fold in. You do not resolve flags or pick winners;
the operator gates. `[OPERATØR]`
6. **Fold in by tightening, → v(n+1).** Fold the flags KTG approved into
`NN-utkast.md` **by tightening** (rule 6 — close the gap, hold the length flat).
The editor (this session) holds the pen; never paste a reviewer's direction as
copy. If the fold-in was substantive, re-run the affected archetype on v(n+1).
All of this happens **before** lock, so the body is never reopened post-lock.
7. **Persist + checkpoint state.** Record the run in `edition-state.json` →
`articles.NN.headlessReview` (`frozenDraft`, per-reviewer `{reportPath,
summary, status}`, `consolidatedReport`, `foldedIn`/`waived`, `status:
"folded"`), set `currentPhase: "headless-review"`, and write a "headless review
complete (cold, converged flags folded) → next: annotation/lock" line to
`<serie>/STATE.md` (overwrite).
```
Headless adversarial review (cold, BEFORE lock) — complete.
- Frozen draft: <serie>/review/NN-frozen-<stamp>.md
- Archetypes: content · language · fact · persona-resonance (<N> personas) · persona-conversion (primær)
- Converged flags (independent agreement): <N>
- BLOCK/🔴: <N> → folded/​waived REWORK: <N> primær resonance: JA conversion: JA
- Surfaced to operator: <serie>/review/NN-headless-<stamp>.md (via SendUserFile) [OPERATØR]
- Folded in (by tightening, pre-lock): <N> Waived: <N>
Gate: [PASS — operator approved, body re-touched pre-lock]
Next: Step 7 — Annotation (optional), then Step 7.5 — Visual assets, then Step 8 — LOCK.
```
---
@ -1175,19 +1312,38 @@ produces the editor's single delivery artifact — `POST.html`, the
(delingstekst) copy, ready to paste into LinkedIn.
> **Order assertion (enforced).** Lock (Step 8) runs AFTER the pre-lock persona
> sweep (Step 6) and BEFORE the hook/conversion gate (Step 9). Reversing lock
> and the pre-lock sweep reproduces the exact Seres failure (reopening locked
> texts) — see Step 6. The post-lock hook-gate (Step 9) judges only the
> distribution hook and never reopens the locked body. `[GATE]`
> sweep (Step 6) AND the headless adversarial review (Step 6.5), and BEFORE the
> hook/conversion gate (Step 9). Reversing lock and the pre-lock sweeps
> reproduces the exact Seres failure (reopening locked texts) — see Step 6. The
> post-lock hook-gate (Step 9) judges only the distribution hook and never
> reopens the locked body. `[GATE]`
**Procedure:**
1. **Confirm lock preconditions.** In `edition-state.json`: the article's
`factcheckLog` has no open 🔴, `personaSweep.resonance` recorded a primær JA,
and `visualAssets` is gated — for `standard` format the approved
`linkedin/NN/cover.png` exists (Step 7.5); for `carousel` format the approved
`carousel.pdf`/`carousel.html` exists. If any is missing, STOP — return to the
relevant step (5/6/7.5). Do not lock past an open gate.
`headlessReview.status` is `folded` (or `run` with no open BLOCK/🔴 the
operator left unaddressed — Step 6.5), and `visualAssets` is gated — for
`standard` format the approved `linkedin/NN/cover.png` exists (Step 7.5); for
`carousel` format the approved `carousel.pdf`/`carousel.html` exists. If any is
missing, STOP — return to the relevant step (5/6/6.5/7.5). Do not lock past an
open gate.
**Pivot-detection gate (Endring 9c — enforced).** Before locking, compare the
current draft against the version that last cleared Step 6:
- **word count:** `wc -w <serie>/NN-utkast.md` vs
`articles.NN.personaSweep.resonance.wordCount` (the recorded baseline);
- **new sections:** top-level headings now present that were absent then
(`grep -c '^## '` delta is a fair proxy).
If the draft has drifted **> 20 % in word count OR gained > 2 sections** since
Step 6 cleared, the text pivoted after its gates — **STOP, do not lock.** Run
`/linkedin:pivot --article NN --reason "<what changed>"`, which re-opens the
pipeline so fact-check (5) → editorial (5.5) → persona (6) → headless (6.5)
re-run on the pivoted version. Likewise, if `articles.NN.pivots[]` has an entry
whose `gatesToRerun` gates have not since re-passed, STOP — the pivot's
re-review is incomplete. (This is exactly the Del 4 v8→v11 case: +42 %, 2 new
sections → the gate would have fired and forced the re-sweep.) `[GATE]`
2. **Confirm the delivery inputs in the series folder.**
`render/build-linkedin.mjs` reads, relative to cwd (`<serie>/linkedin/`):
@ -1335,15 +1491,20 @@ Edition complete. Visible in /linkedin:calendar; mark live via /linkedin:calenda
## Reference Files
- `${CLAUDE_PLUGIN_ROOT}/config/edition-state.template.json` — edition-state schema (15 phases including v2.1 skeleton + spine-prose gates, v2.3 visual-assets, and v2.4 editorial-review)
- `${CLAUDE_PLUGIN_ROOT}/config/edition-state.template.json` — edition-state schema (16 phases including v2.1 skeleton + spine-prose gates, v2.3 visual-assets, v2.4 editorial-review, and v3.1 headless-review + per-article `personas` + `pivots`)
- `${CLAUDE_PLUGIN_ROOT}/config/edition-config.template.json` — static delivery metadata schema (calendar, freshness, credit, captions) — Step 8
- `${CLAUDE_PLUGIN_ROOT}/config/image-credit-caption.template.md` — cover motif + credit + caption table (honest-about-AI credit) — Step 7.5
- `${CLAUDE_PLUGIN_ROOT}/config/edition-delingstekst.template.md` — distribution-copy grammar (`## Del N —` / `## Samle`) — Steps 8/9
- `${CLAUDE_PLUGIN_ROOT}/config/personas.template.md` — reusable reader personas + "primær trumfer" rule
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-checker.md` — Step 5 fact-check agent (risk-sorted, guilty-until-disproven)
- `${CLAUDE_PLUGIN_ROOT}/agents/editorial-reviewer.md` — Step 5.5 editor's craft gate (prose-craft + narrative-architecture, BLOCK/REWORK/NICE; mirrors Maskinrommet §C2)
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — Step 2.5/6/9 reader jury (skeleton + resonance + conversion modes)
- `${CLAUDE_PLUGIN_ROOT}/agents/persona-reviewer.md` — Step 2.5/6/9 reader jury (skeleton + resonance + conversion modes); also resonance + conversion in the Step 6.5 headless package
- `${CLAUDE_PLUGIN_ROOT}/agents/voice-scrubber.md` — Step 4 de-AI / Norwegian-chronicle voice scrub (gold standard = approved Norwegian editions, NOT the English post corpus)
- `${CLAUDE_PLUGIN_ROOT}/agents/content-reviewer.md` — Step 6.5 cold argument-integrity review (C1–C5; headless, no session context)
- `${CLAUDE_PLUGIN_ROOT}/agents/language-reviewer.md` — Step 6.5 cold Norwegian-language review (L1–L5; headless)
- `${CLAUDE_PLUGIN_ROOT}/agents/fact-reviewer.md` — Step 6.5 cold re-verification (F1–F4, 🔴/🟡/🟢; catches pivot premises Step 5 missed)
- `${CLAUDE_PLUGIN_ROOT}/commands/headless-review.md` — the Step 6.5 cold review package as a standalone command (run in a fresh session for maximum isolation)
- `${CLAUDE_PLUGIN_ROOT}/commands/pivot.md` — re-opens the pipeline after a late pivot so Steps 5–6.5 re-run on the changed version before lock
- `${CLAUDE_PLUGIN_ROOT}/commands/react.md` — multi-source synthesis discipline (reused in Step 2)
- `${CLAUDE_PLUGIN_ROOT}/assets/voice-samples/authentic-voice-samples.md` — voice matching
- `${CLAUDE_PLUGIN_ROOT}/references/longform-quality-rules.md` — canonical long-form rules (Steps 2.5, 3a, 3b, 4–9 all reference)

161
plugins/linkedin-studio/commands/pivot.md Normal file
View file

@ -0,0 +1,161 @@
---
name: linkedin:pivot
description: |
Re-open a long-form edition after a substantive late change (a "pivot") so the
already-cleared quality gates re-run on the changed version before lock. Logs
the pivot in edition-state, moves currentPhase back to the right earlier step,
un-locks if needed, and marks which gates (fact-check, editorial, persona,
headless) must re-pass. Includes the pivot-detection heuristic (>20% word-count
change or >2 new sections).
Use when the user says: "pivot", "I changed the angle", "I added a section",
"re-open the edition", "the article changed after it was approved",
"re-run the gates", "this needs re-review".
Triggers on: "pivot", "linkedin pivot", "re-open edition", "added a section",
"changed the angle", "pivot-reopen", "/linkedin:pivot".
allowed-tools:
- Read
- Glob
- Grep
- Bash
- AskUserQuestion
- Write
---
# LinkedIn Pivot — re-open a cleared edition for re-review
A **pivot** is a substantive change to a long-form draft made *after* a gate had
already cleared it — a new argument anchor, a new section, a changed thesis. The
problem this command solves: the pipeline's gates (fact-check Step 5, editorial
Step 5.5, persona Step 6, headless Step 6.5) ran on the *pre-pivot* version, so
the changed text was never validated. Without an explicit re-open, a pivoted
edition can sail into lock carrying an unverified premise or an unread argument.
> **Why this exists (Del 4, Endring 9c).** Del 4 was LOCK-ready on an early
> version (v8). Then a "Security Champions" pivot added a ~260-word section and a
> ~270-word role-description section — roughly +42 % length, two new sections, a
> new axis. The pipeline had no pivot-mode, so the whole post-lock chain had to be
> re-opened by hand. This command makes the re-open a **named ritual**, not a
> manual scramble.
## Command anatomy
```
/linkedin:pivot
--article NN (required; the edition article that changed)
--reason "<one line>" (required; e.g. "Security Champions-anker")
--to-phase draft | consistency-quality | factcheck-sweep (optional; default from the heuristic)
```
## The pivot-detection heuristic
Compare the **current** draft against the version that **last cleared Step 6**
(persona-sweep-prelock). A pivot-reopen is **suggested/required** when either:
- **word-count change > 20 %**, or
- **> 2 new sections** (top-level headings added since the cleared version).
This heuristic is also checked as a **lock precondition in Step 8** of
`/linkedin:newsletter`: if the draft has drifted past these bounds since Step 6
cleared, the lock STOPS and points the operator here. (Length-band drift itself
— soft/hard caps — is logged friction F1, not yet a gate.)
> **Worked example (acceptance test — Del 4 v8 → v11).** At v8 the persona sweep
> had cleared a ~1 400-word draft. The Security Champions message pushed it to
> ~1 992 words (+42 %, > 20 %) and added 2 sections (a new anchor + a
> role-description) — at the boundary of the "> 2 new sections" rule and well past
> the 20 % rule. **The heuristic fires:** `/linkedin:pivot --article 04 --reason
> "Security Champions-anker"` would log the pivot, move `currentPhase` back to
> `draft` (structural change → full re-treatment), and require fact-check +
> editorial + persona + headless to re-pass on v11 before lock. That is exactly
> the re-sweep the manual Del 4 run had to improvise.
## Step 1 — Load state + locate the article
1. Resolve the series root and read `<serie>/linkedin/edition-state.json`. Find
`articles.NN` for the `--article` value. If it does not exist, stop and report.
2. Read the current draft `<serie>/NN-utkast.md` and note `currentPhase` and
`locked`.
## Step 2 — Measure the pivot scope
1. **Current word count:** `cd <serie-mappe> && wc -w NN-utkast.md`.
2. **Baseline word count:** the word count of the version that last cleared
Step 6, recorded by newsletter Step 6 in
`articles.NN.personaSweep.resonance.wordCount`. If that field is absent (older
state), ask the operator for the cleared-version word count, or treat the
pivot as structural by default.
3. **Compute** `deltaPct = round((current - baseline) / baseline * 100)` and
count `newSections` = top-level headings now present that were not in the
cleared version (a `grep -c '^## '` delta is a reasonable proxy; confirm with
the operator if ambiguous).
4. **Classify scope** (drives the default `--to-phase`):
- **Structural** (deltaPct > 20 % OR newSections > 2, or a new axis/thesis) →
default `to-phase: draft` (Step 3b). The new material needs full prose
expansion → consistency → fact-check → editorial → persona → headless.
- **Moderate** (new examples/claims, no new sections, deltaPct ≤ 20 %) →
default `to-phase: factcheck-sweep` (Step 5) so the new claims get verified,
then editorial + persona + headless re-run.
- The operator may override with explicit `--to-phase`.
## Step 3 — Log the pivot + reset the phase (the ritual)
1. **Append a pivot entry** to `articles.NN.pivots[]`:
```json
{
"timestamp": "<ISO-8601>",
"reason": "<--reason>",
"fromPhase": "<currentPhase before this command>",
"toPhase": "<resolved to-phase>",
"wordCountBefore": <baseline>,
"wordCountAfter": <current>,
"deltaPct": <deltaPct>,
"newSections": <newSections>,
"gatesToRerun": ["factcheck-sweep", "editorial-review", "persona-sweep-prelock", "headless-review"]
}
```
`gatesToRerun` always spans every gate from the reset phase through Step 6.5 —
a pivot invalidates the fact-check, the editorial craft pass, the persona
resonance verdict, AND the headless package, because all of them judged the
pre-pivot text.
2. **Reset `currentPhase`** to the resolved `toPhase`, and set the article's
`phase` to match.
3. **Un-lock if needed.** If `articles.NN.locked` was `true`, set `locked: false`
and `status: "in-progress"` — a pivot means the edition is no longer locked.
Surface this plainly (the prior `POST.html` is now stale and will be re-rendered
at the next lock).
4. **Invalidate the downstream verdicts** so they cannot be mistaken for current:
set `personaSweep.resonance`, `editorialReview`, and `headlessReview.status`
back to a re-run state (e.g. `headlessReview.status: "pending"`), and note in
each that they were invalidated by pivot `<timestamp>`. Leave the `pivots[]`
log and `factcheckLog` history intact (history is durable).
5. **Update `updatedAt`** and write `edition-state.json`.
## Step 4 — Point the next step
Write a precise next-step line to `<serie>/STATE.md` (overwrite, ONE-system):
```
PIVOT logged (<reason>, +<deltaPct>%, <newSections> new sections) →
currentPhase reset to <toPhase>. Resume /linkedin:newsletter; it will re-run
fact-check (5) → editorial (5.5) → persona (6) → headless (6.5) on the pivoted
version BEFORE lock. Headless package: /linkedin:headless-review --article NN.
```
Do not run the gates yourself — `/linkedin:newsletter` owns the pipeline and will
resume deterministically from the reset `currentPhase` and re-run every gate in
`gatesToRerun` before it permits lock.
```
Pivot logged.
- Article NN: <reason>
- Scope: <structural|moderate> (Δ <deltaPct>%, <newSections> new sections)
- currentPhase: <fromPhase> → <toPhase> locked: <true→false | unchanged>
- Gates to re-run before lock: fact-check (5) · editorial (5.5) · persona (6) · headless (6.5)
Next: resume /linkedin:newsletter (re-runs the gates) → then lock.
```
## Reference Files
- `${CLAUDE_PLUGIN_ROOT}/commands/newsletter.md` — the pipeline this re-opens; Step 8 lock-precondition runs the same heuristic; the resumption table replays from the reset `currentPhase`
- `${CLAUDE_PLUGIN_ROOT}/commands/headless-review.md` — the cold review package that must re-pass on the pivoted version
- `${CLAUDE_PLUGIN_ROOT}/config/edition-state.template.json` — `articles.NN.pivots` schema + the heuristic notes

22
plugins/linkedin-studio/config/edition-state.template.json
View file

@ -15,6 +15,7 @@
"factcheck-sweep — risk-sorted, guilty-until-disproven, verification log (Step 5)",
"editorial-review — editor's craft gate: prose-craft (em-dash density, verbatim repetition, postulated numbers, contradictions, versal-tic) + narrative-architecture (concrete instantiation, theory-anchored hypotheses, series-title symmetry, equal action per addressee, un-overloaded conclusion), ≤10 flags BLOCK/REWORK/NICE, operator-gated via SendUserFile BEFORE the persona sweep (Step 5.5)",
"persona-sweep-prelock — reader jury, primary wins, convergence to clean YES (Step 6)",
"headless-review — adversarial review package run COLD on a FROZEN draft (no drafting-session context): content-reviewer (argument integrity) + language-reviewer (Norwegian language) + fact-reviewer (cold re-verification incl. pivot premises) + persona-reviewer in resonance & conversion modes. Fan-out from Step 6.5 or the standalone /linkedin:headless-review command; consolidated report operator-gated via SendUserFile BEFORE lock (Step 6.5)",
"annotation — optional annotatable review HTML for a manual pass (Step 7)",
"visual-assets — cover (+ optional inline figures) or carousel deck: brief → generate → operator-gate → approve, BEFORE lock so build-linkedin.mjs picks them up (Step 7.5)",
"lock-delivery — LOCK → POST.html all-in-one-place deliverable (Step 8)",
@ -23,7 +24,10 @@
],
"articleStatusValues": ["pending", "in-progress", "locked", "scheduled"],
"editorialReview": "Per-article editorial-review record written by Step 5.5 (editorial-review phase). Runs AFTER fact-check (Step 5) and BEFORE the persona sweep (Step 6): the editorial-reviewer agent judges CRAFT (prose-craft + narrative-architecture), not reader-response, mirroring the Maskinrommet skrivekontrakt §C2. The report (≤10 flags, each with kategori P1–P5/A1–A5, quote/line-ref, direction, severity BLOCK/REWORK/NICE) is surfaced to the operator via SendUserFile; the operator decides which flags fold in. Shape: { reportPath, flagCount, byAxis: { prosa, arkitektur }, bySeverity: { block, rework, nice }, foldedIn, waived, status }. status ladder: pending → reviewed → folded. null until Step 5.5 runs. This is the craft companion to factcheckLog (truth) and personaSweep (response).",
"visualAssets": "Per-article visual-asset record written by Step 7.5 (visual-assets phase). Runs BEFORE lock because render/build-linkedin.mjs picks up linkedin/NN/cover.png + the edition-config credit/caption when it builds POST.html — generating images after lock would force a re-render. Shape: { format: \"standard\" | \"carousel\"; cover: { brief, route, candidates[], approved, status }; figures: [ { id, brief, placement, status } ]; carousel: null | { source, pdf, status } }. format \"standard\" = cover + optional inline figures (cover.png is mandatory per the KTG cover-directive); format \"carousel\" = typografisk deck via render/build-carousel.mjs instead of cover+inline (cover/figures stay empty). route: \"mcp-image\" (default, via mcp__mcp-image__generate_image) | \"external\" (DALL·E / Midjourney / photographer → linkedin/NN/cover-raw.png). status ladder: pending → briefed → generated → approved. candidates[] holds the cover-v<N>-kandidat.png attempts; approved is the fixed approved name (\"cover.png\") once the operator-gate passes. figures[].id = \"fig1\"..; placement = section reference in NN-utkast.md (figures are referenced in the draft via ![alt](linkedin/NN/figN.png) and uploaded manually in the LinkedIn editor — build-linkedin.mjs does NOT embed them). Naming convention: cover.png (approved, fixed — what build-linkedin.mjs reads) | cover-v<N>-kandidat.png (attempts) | cover-raw.png (optional external pre-edit source) | fig<N>.png (inline). credit + caption are recorded in <serie>/linkedin/image-credit-caption.md and flow into edition-config.json coverCredit + captions[NN]."
"visualAssets": "Per-article visual-asset record written by Step 7.5 (visual-assets phase). Runs BEFORE lock because render/build-linkedin.mjs picks up linkedin/NN/cover.png + the edition-config credit/caption when it builds POST.html — generating images after lock would force a re-render. Shape: { format: \"standard\" | \"carousel\"; cover: { brief, route, candidates[], approved, status }; figures: [ { id, brief, placement, status } ]; carousel: null | { source, pdf, status } }. format \"standard\" = cover + optional inline figures (cover.png is mandatory per the KTG cover-directive); format \"carousel\" = typografisk deck via render/build-carousel.mjs instead of cover+inline (cover/figures stay empty). route: \"mcp-image\" (default, via mcp__mcp-image__generate_image) | \"external\" (DALL·E / Midjourney / photographer → linkedin/NN/cover-raw.png). status ladder: pending → briefed → generated → approved. candidates[] holds the cover-v<N>-kandidat.png attempts; approved is the fixed approved name (\"cover.png\") once the operator-gate passes. figures[].id = \"fig1\"..; placement = section reference in NN-utkast.md (figures are referenced in the draft via ![alt](linkedin/NN/figN.png) and uploaded manually in the LinkedIn editor — build-linkedin.mjs does NOT embed them). Naming convention: cover.png (approved, fixed — what build-linkedin.mjs reads) | cover-v<N>-kandidat.png (attempts) | cover-raw.png (optional external pre-edit source) | fig<N>.png (inline). credit + caption are recorded in <serie>/linkedin/image-credit-caption.md and flow into edition-config.json coverCredit + captions[NN].",
"personas": "Per-article resolved reader-persona set (input config), written/confirmed in Step 1. This makes personas configurable PER ARTIFACT, not just from one global plugin library: Step 1 resolves them in order — (1) already present here → use as-is; (2) <serie>/linkedin/personas.md (per-series file) → load; (3) plugin config/personas.local.md (or personas.template.md) library → select a subset; (4) none/insufficient → DEFINE interactively via AskUserQuestion. Exactly one entry has tier \"primær\" (the rest \"sekundær\"); «primær trumfer» on conflict. This set feeds BOTH the in-session sweep (Step 6) and the headless package (Step 6.5 / persona-reviewer). Each entry: { name, tier: \"primær\" | \"sekundær\", rolle, avkobler, overbeviser, ekspertise, sjargong, source: \"edition-state\" | \"series-file\" | \"plugin-library\" | \"interactive\" }. Default []: resolved on first Step 1.",
"headlessReview": "Per-article headless-review record written by Step 6.5 (headless-review phase). Runs AFTER the in-session persona sweep (Step 6) and BEFORE lock (Step 8), on a FROZEN snapshot of the publish-ready (or pivoted) draft, fanned out from the command layer (foreground) or invoked standalone via /linkedin:headless-review in a fresh/cold session. Five archetypes judge independently with NO drafting-session context: content-reviewer (argument integrity), language-reviewer (Norwegian language), fact-reviewer (cold re-verification incl. claims a late pivot bolted on), persona-reviewer mode=resonans (per active persona), persona-reviewer mode=konverter (primær, hook only). The consolidated report is surfaced to the operator via SendUserFile; the operator decides which flags fold in. Shape: { frozenDraft, reviewers: { content, language, fact, personaResonance, personaConversion } (each { reportPath, summary, status }), consolidatedReport, foldedIn, waived, status }. status ladder: pending → run → folded. null until Step 6.5 runs. This is the adversarial-independence companion to the in-session gates (editorialReview, personaSweep, factcheckLog) — deliberately redundant: a cold reader catches what the framing-biased in-session pass missed.",
"pivots": "Per-article pivot log (Endring 9c). A pivot is a substantive change to a draft AFTER a gate had already cleared — e.g. a new argument anchor / section added late (the Del 4 Security Champions case: +~530 words, 2 new sections, +42 %). Each /linkedin:pivot invocation appends one entry and moves currentPhase back so the cleared gates (Steps 5–6.5) re-run on the pivoted version before lock. Heuristic (documented, checked at the Step 8 lock precondition): if the current draft's word count differs > 20 % from the version that last cleared Step 6, OR it has > 2 new sections, a pivot-reopen is suggested/required. Each entry: { timestamp, reason, fromPhase, toPhase, wordCountBefore, wordCountAfter, deltaPct, newSections, gatesToRerun: [phase…] }. Default []."
},
"schemaVersion": 1,
"series": {
@ -38,6 +42,7 @@
"title": "<Article 1 title>",
"phase": "load-context",
"status": "pending",
"personas": [],
"immutableRules": null,
"factcheckLog": null,
"editorialReview": null,
@ -46,6 +51,20 @@
"resonance": null,
"conversion": null
},
"headlessReview": {
"frozenDraft": null,
"reviewers": {
"content": null,
"language": null,
"fact": null,
"personaResonance": null,
"personaConversion": null
},
"consolidatedReport": null,
"foldedIn": null,
"waived": null,
"status": "pending"
},
"visualAssets": {
"format": "standard",
"cover": {
@ -58,6 +77,7 @@
"figures": [],
"carousel": null
},
"pivots": [],
"locked": false,
"scheduled": null
}

28
plugins/linkedin-studio/config/personas.template.md
View file

@ -32,6 +32,34 @@ trim, or extend them per series.
BEFORE lock — «does the point land for this reader?») and conversion mode
(Step 9, after lock — binary «would YOU click?» on the hook only).
### Per-artifact personas (one or more personas per edition)
This library is a *starting point*, not a fixed cast. **Each artifact (each
newsletter edition) carries its own resolved persona set** — one or more
personas, exactly one marked `primær` — so different editions can target
different readers without editing a shared file. `/linkedin:newsletter` Step 1
**resolves** the active set in this order and records it in
`edition-state.json` → `articles.NN.personas` (so it is stable across the
multi-session pipeline and is the single source the Step 6 sweep AND the
Step 6.5 headless package read):
1. **Already in `articles.NN.personas`** → use as-is (a resumed edition keeps
the set it was calibrated with).
2. **`<serie>/linkedin/personas.md`** (a per-series file, same block grammar as
below) → load it. Use this when a whole series shares a cast.
3. **Plugin `config/personas.local.md`** (else this `personas.template.md`) →
select the relevant subset of the global library.
4. **None / insufficient** → **define interactively** in Step 1 (the operator
names one or more personas and their five fields via `AskUserQuestion`); the
resolved set is written to `articles.NN.personas`.
Each resolved entry carries the five fields below plus `tier`
(`primær` | `sekundær`) and `source` (`edition-state` | `series-file` |
`plugin-library` | `interactive`). Exactly one `primær` per artifact; «primær
trumfer» (below) is unchanged. Personas defined interactively for one edition
can be promoted to a reusable block by pasting them into `personas.local.md`
(plugin-wide) or `<serie>/linkedin/personas.md` (series-wide).
### The click-gate is blocking (bar = primær ekte JA)
The persona sweep is not advisory — it returns a **blocking verdict**