Trim the front-door README to essentials: value prop, two engines, quick start, full command overview, compact 19-agent roster (lint-required model surface), merged boundaries, quality rules, 2 examples, license. Deep reference (knowledge base, hooks/skills/personalization/config/analytics internals, agent pipeline) now points to references/ and CLAUDE.md; full version history points to CHANGELOG. The pre-1.0 note is gap-framed (what is not done yet), not a promised roadmap. No functionality changed; lint green (74/0).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Reset version declarations to v0.4.0 across plugin.json, README badge + version-history, CLAUDE.md header, and the root marketplace README entry. The 1.0.0–4.1.0 numbering reflected ambition, not maturity: user data still lives inside the plugin tree (.gitignore-defended), no command has passed the hardening quality-gate, command testing is incomplete, and there is no GUI. A deliberate marketplace downgrade (existing installs will not auto-pull a lower number). Path to v1.0.0 = architecture (M0 per-user data-dir migration) + hardening + command testing + GUI. No functionality changed — only version declarations; all prior changelog/history preserved. Lint green (Passed 74 / Failed 0).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
v1 (autonomous 5-step + reviewer swarm + /trekreview gate) produced confident
fabrications and burned context for what are usually 0-1 line edits. v2 is a
slow, dialogic, one-command-per-session gate with the operator as truth source
and every mechanical claim tool-grounded. Method, command-class predicates,
session queue, and STATE.md end-of-session handoff are documented in plan.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New phase after the baseline-audit remediation (S1-S17, 2633d32, complete):
a command-hardening pass that simulates each of the 29 commands and tightens
it to its stated intention (intention-fidelity + prompt-quality only — no
structural redesign, no new features, no GUI/M0). Runs over ~8 journey-grouped
Voyage sessions, gated by lint + /trekreview ALLOW before push.
Foundation laid this session (execution starts next session):
- docs/hardening/brief.md (valid; 3 locked forks: hybrid simulation,
intention-fidelity+prompt-quality, per-journey cadence; research skipped —
the 2026 bar is frozen in algorithm-signals-reference.md)
- docs/hardening/plan.md (8 sessions, Grade B+ 87)
Adversarial review: brief-reviewer PROCEED_WITH_RISKS (3 findings folded:
self-graded quality axis, SC-H deferral contradiction, no stopping rule);
plan-critic REVISE (2 blockers + 10 major — all addressed: anchored coverage
predicate kills substring false-positives, full-session cold-reviewer oracle
on concrete before/after output, per-type mechanical predicate, Failed:0 gate
not literal-74, S1 HALT circuit-breaker); scope-guardian ALIGNED.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Cold-read triage of the ~34 uncalibrated baseline-audit findings (C13–C46)
that never got a second hostile pass. An independent Opus reader classified
each against the current code; every disposition was grep-verified in the
main session.
Result: 0 still-real, 23 already-fixed, 1 outdated-drop (24 grouped/sub-claim
entries). No inline code fix needed — the v4.0.0 + S13–S16 remediation had
already closed every still-real item (dead lint, 11 orphan agents, carousel
full-deck clipboard, router tiering, onboarding inline, de-AI gate, video gate,
post-feedback-monitor->Opus, series-path parameterization, SKILL roster).
Deliverable: docs/remediation/c13-c46-triage.md (disposition record) +
docs/remediation/review.md (S17 review, ALLOW). /trekreview: brief-conformance
0 findings; code-correctness 2 MAJOR in the triage doc's own prose (one
overclaim, one line-pointer) FIXED in-session — no false-green disposition.
Gate: test-runner.sh 74/0/0, hooks node --test 98/98, analytics 116/116.
M0 (per-user data-dir migration) deferred to the UI track. Remediation COMPLETE.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
B2 (router tiering) was already delivered in S14, so S15 = B1 + B3 only.
No surface/count/version change -> within-v4.1.0 refinement (S11-S13 precedent).
- B1 (commands/onboarding.md): replace the "Run /linkedin:first-post" dead-end
hand-off in Phase 3 with the first-post drafting steps embedded inline (3.1 topic
-> 3.2 3-line draft -> 3.3 QC -> 3.4 present+clipboard -> 3.5 state-update that sets
first_post_date). Wizard now yields a draft in-flow; 0 dead-end strings. Stays within
the existing allowed-tools (Read/Bash/AskUserQuestion); UI-brief §12b scope-guard
honored (no provider seams / progressive-disclosure added).
- B3 (commands/carousel.md): Step 6 now assembles the ENTIRE deck (every slide's copy
+ the caption) into the clipboard payload, not just the caption; full-deck assembly
precedes the clipboard-helper.mjs call.
Independent /trekreview (2 Opus reviewers): brief-conformance 0 findings; code-correctness
1 MAJOR that is PRE-EXISTING and out of S15 scope (onboarding Phase 2 saves need Write in
allowed-tools; lines 142/157, untouched by the S15 diff) -> DEFERRED to next session per
"ekte design-funn -> neste sesjon". Verdict ALLOW for the delivered scope (not a WARN-override).
Gate: test-runner.sh 74/0/0; node --test 98/98; commands=29; v4.1.0 unchanged.
See docs/remediation/review.md for the full record (ALLOW + 1 deferred MAJOR).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the 2 grep/Read-verified findings from the S12 cold full-brief re-review
(docs/remediation/review.md, WARN 0/1/1/0, 0 dropped) and closes the $-injection
CLASS — not the line — across the whole state-updater.mjs mutation surface.
See docs/remediation/review.md (S13 ALLOW, 0/0/0/0) for the full closure record:
replaceField -> replacement function; the 3 additive-insert sites -> functions
(m === $1, behavior-preserving); a scalar assert.match pins last_post_topic; and a
behavioral, coverage-complete, self-testing Section 12 guard (check-replace-safety.mjs)
that is mutation-proven. Docs three-doc + residuals updated. test-runner.sh 71/0/0,
node --test 98/98.
Closes the 2 grep-verified findings from the S11 cold full-brief re-review
(docs/remediation/review.md, BLOCK 1/0/1/0, 0 dropped). Both were the NEXT
RING of the meta-class S10/S11 converged: a propagation miss — the fix had
landed where the SC named the file, not in the render-source it depends on.
BLOCKER (command->reference propagation): references/ab-testing-framework.md:166
still shipped the banned A/B "Significant? (>20%)" Yes/No verdict column while
commands/ab-test.md (which RENDERS from it, inlined at :30, presented at :69)
had been cleaned to the honest "Directional?" framing. Re-framed the reference
result template to match the command verbatim (header + the directional note)
and retuned :38 "20% significance threshold" -> "minimum-meaningful-difference
threshold". The whole render chain is now significance-verdict-free.
MINOR ($-replacement, class-closed not line-patched): the newest-first section
appends/rewrites in hooks/scripts/state-updater.mjs passed a replacement STRING
embedding untrusted user content to String.replace, so dollar-sequences
($1 / $& / dollar-backtick / dollar-apostrophe / $$) in a topic/hook/partner
(e.g. "$100 budget cut") re-injected the captured heading and dropped
characters, silently corrupting state. Converted all 5 content-bearing sites
(Recent Posts, prune, Milestone Log, First-Hour, Outreach) to replacement
FUNCTIONS; the 3 remaining $1 sites only interpolate date scalars. +4
$-bearing regression tests — incl. the prune fixture, which itself had to
switch to a function (the bug bit the fixture as it was being written).
META (generalize the lint to the propagation-miss class): new test-runner.sh
Section 11 — render-chain propagation guard. Forbids the significance-verdict
column (Significant? adjacent to "(" or a table pipe) across the WHOLE render
chain (commands + every inlined reference + adjacent surfaces), with a
permanent non-vacuity self-test (3 verdict forms caught, 6 legitimate
Significant/significance/Directional? forms ignored) and an e2e mutation-proof.
Generalizes S10/S11's "fix the class, not the line" to command->reference.
Pre-patch render-chain sweep confirmed ab-testing-framework.md was the SOLE
propagation survivor (so a 6th review finds no 3rd). test-runner.sh 70/0/0;
node --test 98/98. CLAUDE.md lint enumeration synced.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Cold full-brief re-review (S10) reached a class the S7->S9 algorithm-stat lens
never did:
BLOCKER — post-feedback-monitor published as Haiku in four surfaces
(README:259, skills/linkedin-studio:159 with wrong color Green too,
skills/linkedin-analytics:41, agents-capability-matrix:20) while
agents/post-feedback-monitor.md runs Opus. v4.0.0's Opus promotion never
reached the user-facing tables. Synced all to Opus/Lime. Refreshed
agents-capability-matrix.md (frozen at the v2.0 14-agent era): header 14->19,
+5 missing longform agents, tier counts Opus 2->8 / Haiku 1->0, longform-gate
diagram updated to the real 8-Opus-agent chain.
MAJOR — de-branded docs/plan-fullspektrum-innholdsmotor.md:70 (model brand +
jan-2026 asserted as fact -> no-name/no-month relevance-model phrasing). It was
the only tracked survivor; the rest live in gitignored ROADMAP.md /
.claude/research/ (not shipped, out of honesty scope).
META — added Section 10 model-consistency guard
(scripts/check-model-consistency.mjs): each agents/*.md model: must match every
surface declaration AND the canonical rosters must list all 19 agents.
Permanent non-vacuity self-test + e2e mutation-proven.
Pre-patch sweep confirmed post-feedback-monitor was the sole drifted agent
(89 model rows, 0 other mismatches). test-runner.sh 68/0/0, node --test 94/94.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the S9 re-review (1 BLOCKER + 2 MAJOR, all grep-verified). The survivor
set converged 8 -> 6 -> 2; this closes the meta-problem behind the convergence,
not just the two lines.
BLOCKER — references/glossary.md:10: drop the fabricated "150-parameter
foundation model" (a garbled 150B that the S9 enumerative grep/lint, requiring a
"B"/"billion", could not match). Reframe to "a real input to LinkedIn's 2026
relevance-ranking model" with no parameter count, citing
algorithm-signals-reference.md inline — which makes the :12 "Used in" provenance
accurate (the reference does state the relevance-ranking framing; it never stated
"150-parameter").
MAJOR — CHANGELOG.md:308: de-brand "360Brew profile optimization (January 2026
algorithm update)" -> "Profile/topic-relevance optimization". Removes the
unpublishable brand + asserted Jan-2026 date, honouring v4.0.0's "removed
everywhere" claim. It was the only STALE_STATS hit in CHANGELOG.
MAJOR — scripts/test-runner.sh: the rebuilt lint was enumerative on surface form.
Generalize it to the PATTERN CLASS so the same grep that defines the SC fails on
any surface form, present or future:
- STALE_STATS model token: "150 ?B param|150 billion param"
-> "[0-9]+[ -]?(B|billion)?[ -]?param"
(covers 150-parameter / 150B param / 150 billion param). This robustifies the
review's literal suggestion "[0-9]+[ -]?(B|billion )?param", which missed the
space form "150B param"; the separator is moved out of the group.
- STAT_HITS grep scope += CHANGELOG.md (the 360Brew survivor lived outside it).
- Permanent non-vacuity SELF-TEST before the real scan: 13 forbidden probes must
match (incl. the exact "150-parameter" survivor), 8 legitimate probes must not
("Language parameter", "parameterized", "different parameters",
"175-milliarders parametermodell", 5x5x5, cadence, pixel dims, "10x your
reach"). S7->S9 each shipped a green lint because the proof was run by hand and
never committed; this makes narrowing STALE_STATS fail the suite.
Verification: test-runner.sh 67/0/0 exit 0 (was 66/0/0; +1 self-test);
node --test 94/94; broadened exhaustive grep across the tree -> zero survivors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Closes the S8 re-review (BLOCK 3/4/1). The S8 fix patched only the 2 strings S7 named; the re-review found 6 more same-class survivors. Per the systemic read, this is a comprehensive sweep, not a per-line patch.
Reconciled every retired engagement-coefficient + model-fact survivor against the canonical references/algorithm-signals-reference.md (order, not coefficients; comment ≈ 2x a like; no model name/params):
- glossary.md: coefficient table + Save-Signal '10x weight' → canonical ordering (citation now true)
- engagement-frameworks.md, analytics-interpreter.md, content-optimizer.md, pipeline.md, engagement-coach.md: the 10x/8x/7-9x/2.5x/0.2x system (incl. 4 survivors the re-review did not cite) → ordering
- playbook: '15x more algorithmic boost' + video '5x more conversations' → directional, sourced
- profile.md + linkedin-voice/SKILL.md: '150B parameter foundation model' → '2026 relevance-ranking model'
- quality-scorecard.md: '360Brew Validation' → topic-relevance framing
- setup.md: 'thought leadership plugin' → 'LinkedIn Studio plugin'
Lint (MAJOR 4): rebuilt scripts/test-runner.sh STALE_STATS to forbid EVERY retired-class phrasing (not the 2 S7 strings) + widened scope to assets/checklists/. Targets retired phrasings (7-9x, (10x), '10x weight', '5x more conversations'), NOT bare 10x/15x/5x (legit 5x5x5 / cadence / pixel-dims / '10x your reach' hyperbole). Proven non-vacuous: catches all 10 retired strings, ignores all 10 legit uses.
Tests (MAJOR 7): added no-anchor fall-through tests for recordFirstHourPlan + recordOutreachContact (date scalar not written/reported, section still appended). MINOR 8: reflowed newsletter.md content-repurposer wiring onto one line.
test-runner.sh 66/0/0; node --test 94/94 (was 92, +2). NO push until /trekreview re-clears the gate.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Close the 5 findings from the S7 /trekreview release gate (review.md, verdict BLOCK):
- BLOCKER: comment-multiplier "5x" reconciled to the canonical order-only framing
(no fixed multiplier) in agents/engagement-coach.md, linkedin-growth-playbook,
linkedin-formats.md — per algorithm-signals-reference.md ("do not quote a comment
multiplier").
- BLOCKER: carousel rate "6.60%/6.6% (highest)" reconciled to "~7% top organic
format" in linkedin-formats.md:42 (was self-contradicting :50) and
assets/templates/carousel-templates.md.
- Lint hardening: test-runner.sh STALE_STATS now matches 6.60% + the 5x comment
folklore and scans agents/ + assets/templates/ — the grep that defines the
Phase-0 criterion now catches both BLOCKERs.
- MAJOR: onboarding.md command count 26 -> 27.
- MAJOR: add section-append-branch (production-path) tests for recordFirstHourPlan
+ recordOutreachContact against a template-layout fixture.
- MINOR: move date-scalar changes.push inside the write branch in state-updater.mjs.
Verify: node --test hooks/scripts/__tests__/*.test.mjs -> 92/92; bash
scripts/test-runner.sh -> 66/0/0. NO push until /trekreview re-confirms ALLOW/WARN.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Step 21 (remediation Wave 4 / S6, SOLO): finalize the audit-remediation as
v4.0.0. Version 3.1.0 -> 4.0.0 across all current-version declarations; counts
reconciled to the ls-derived source of truth (19 agents / 27 commands / 6 skills
/ 9 hooks / 25 reference docs / 16 newsletter phases — Step 20 confirmed NO
TRIM); three-doc sync (plugin README + plugin CLAUDE.md + root README) clears the
[skip-docs] debt accumulated across Wave 1-4; CHANGELOG v4.0.0 entry summarizing
Steps 1-20.
Scope additions beyond the plan's literal Files list, all version-sync or
[skip-docs]-debt in nature (flagged, not feature creep):
- ../../CLAUDE.md (root marketplace): linkedin-studio entry v3.1.0 -> v4.0.0
(the version-sync invariant mandates updating every version reference; leaving
it stale is a real inconsistency).
- scripts/test-runner.sh: added the version-consistency grep the file's own
Step-1 comment promised ("added in Step 21") — plugin.json version must match
the README badge, the CLAUDE.md header, and the CHANGELOG top entry — and
folded the Wave 2 lint gap (plugin.json now covered by the stat-consistency
scan). 66/0/0, exit 0.
- plugin README: added the missing /linkedin:firsthour command row (Step 16
[skip-docs] debt) and the 25th reference doc (longform-quality-rules.md) to
the knowledge-base table; fixed body counts ("All 26 commands" -> 27, "24
reference documents" -> 25); badges + intro 26 -> 27 commands, 24 -> 25 refs.
- root README + marketplace.json: dropped the unpublishable model brand/date
("360Brew" / "January 2026") the algorithm-signal reconciliation already
removed everywhere inside the plugin.
Surviving "3.1.0" strings are intentional history, not stale declarations: the
README version-history table row, the "vX added Y" attributions in
plugin.json/CLAUDE.md, and the headless-review reload caveat are all
changelog-genre. Every current-version declaration (plugin.json version, README
badge, CLAUDE.md header, root README marker, marketplace.json) reads 4.0.0.
The major bump reflects the remediation's scope plus the reinstall/reload the
newly-wired agents need to register, and consolidates — does not repeat — the
v3.0.0 identity break; it is not a fresh breaking API change (locked operator
decision).
Pre-existing and out of scope (flagged, untouched): a duplicate /linkedin:setup
row in the README command tables.
Verify: bash scripts/test-runner.sh exit 0 (66/0/0); plugin.json + marketplace.json
parse; counts consistent README == CLAUDE.md == root README; stale-count sweep
clean. NO push — /trekreview (S7) is the release gate.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Steg 20 (remediation Wave 4 / S5, SOLO): measure whether the 7-agent long-form
review stack carries redundant gates. Method: cross-reference each agent's check
taxonomy against its in-repo fasit fixture; four fixtures (editorial, content,
language, fact-reviewer) target the SAME Del 4 edition, enabling a real
cross-gate overlap comparison on one piece (not a live run — fixtures' own
live-run notes require a reload + cross-repo Maskinrommet access, out of scope).
Finding: every gate has >=1 unique catch on Del 4. The four genuine overlaps
(verbatim repetition, the Vi/Vi-i-Nav quote, the postulated number, the
small-orgs thread) are each justified — a cold re-take (Endring 9's reason to
exist), the same symptom via a different operation (flag-absence vs web-verify),
or two distinct defects sharing a surface topic — with no subsumption either way.
The fact-checker <-> fact-reviewer overlap is load-bearing (the pivot premise
arrived after Step 5, so only the cold re-run caught it).
Decision: NO TRIM. voice-scrubber has no fixture -> inconclusive; redundancy
retained (Step 20 On-failure = skip). Counts unchanged 19 agents / 27 commands;
count contract (EXPECT_AGENTS=19) untouched. test-runner 62/62 green.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Step 19 (Wave 4 S4). Adds an expectation banner at the very top of
/linkedin:newsletter: a multi-session, multi-gate process (16 phases),
~4-8+ hours across several sessions, state maintained between sessions, with
a pointer to the short-form commands for feed posts. Sets realistic
expectations before the user starts. Pure docs/content.
Verify: grep -niE 'multi-session|multi-gate|16 phases|~[0-9].*hour' near top
→ matches; test-runner.sh exit 0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Step 17 (Wave 4 S4) of the audit remediation. Applies research/03 §D5 + the
two S2 residual fixes folded in. No new commands/agents (counts stay 27/19).
Newsletter (commands/newsletter.md): new "Distribution channel" section after
Step 10 teaching the HONEST native-newsletter mechanics — bypasses organic feed
ranking via ONE deduplicated notification per subscriber per edition (NOT a
three-touchpoint blast); the mass invite fires once → ~1-2K follower floor
(wait until you can spend it); realistic cold-start 0-100 subs months 1-3;
discloses non-export / no-canonical / no-read-analytics / per-subscriber decay;
explicit below-vs-above-floor decision rule. Sourced to research/03 D5.
Profile (commands/profile.md): new "Profile SEO" section — headline as the
highest-weight search field + a per-section keyword-target table
(headline/about/experience/skills/featured), consistency-over-stuffing rule.
Outreach (commands/outreach.md): Step 8c persists the pipeline board to tracked
state via the new recordOutreachContact mutation (mirrors Step 16's
recordFirstHourPlan): additive last_outreach_date/outreach_active scalars + a
non-R-initial ## Outreach Pipeline section in state-updater.mjs +
config/state-file.template.md + --record-outreach CLI branch. +7 tests
(state-updater 26→33, full hook suite 83→90).
Residual 1 (growth-playbook:216): 9:16 "distribution boost" → 4:5/1:1 guidance
(9:16 mobile-only opt-in; "immersive distribution" = uncorroborated heuristic).
Residual 2 (video-strategy-guide:300): "3-second test determines 70% retention"
→ "front-load value for muted autoplay" (three-second hook is folklore, not a
LinkedIn signal).
Verify: grep checks 1-5 pass; test-runner.sh exit 0 (stat-consistency green);
state-updater 33/33. [skip-docs] — tre-doc + version bump deferred to Step 21.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
post-feedback-monitor was the lone non-Opus/Sonnet agent (model: haiku). It
does human-facing, real-time post-publish coaching — exactly the workload the
standing Opus-default targets — so the Haiku setting both contradicts the
default and underpowers the surface the new /linkedin:firsthour command hands
off to (remediation Step 18).
- agents/post-feedback-monitor.md: model haiku -> opus.
- CLAUDE.md: agent-table model column Haiku -> Opus. No Haiku agent remains.
Verify: grep 'model: haiku' agents/ -> none; CLAUDE.md row shows Opus;
structure lint -> 62/62, counts unchanged.
Plan Step 18 (Wave 4 S3). Counts unchanged.
[skip-docs]: tre-doc + version bump deferred to Step 21 per remediation plan.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wire orphan agent #11 (engagement-coach) by giving it a command surface, and
add the tracked first-hour state the plan calls for (remediation Step 16).
- commands/firsthour.md (new, 27th command): post-publish first-hour /
reply-loop sprint. Delegates plan construction to engagement-coach via
Task (subagent_type: linkedin-studio:engagement-coach) — returns a grouped
target list (whales/inner-circle/ICPs/new connections), 2-3 seed
self-comments + 3-5 CEA replies in the user's voice, and a minute-by-minute
timeline anchored to publish time. Presents timeline/targets/drafts +
velocity checkpoints, auto-copies the drafts to clipboard, persists the
plan, then hands off to post-feedback-monitor for the 48h window.
- hooks/scripts/state-updater.mjs: new pure mutation recordFirstHourPlan()
mirroring updatePostTracking — additive by contract (inserts
last_firsthour_date after last_post_date when absent, creates the
## First-Hour Plans section when absent, never touches existing fields).
Section name is deliberately non-R-initial so it stays outside
pruneContentHistory's "## Recent Posts ... (?=\n## [^R])" capture window.
+ a --record-firsthour CLI branch for parity with the other mutations.
- config/state-file.template.md: additive scalars (last_firsthour_date,
firsthour_active) + the ## First-Hour Plans section.
- hooks/scripts/__tests__/state-updater.test.mjs: extend (existing file) with
7 recordFirstHourPlan tests — section creation, field insertion vs in-place
update (no duplication), round-trip non-interference, graceful empty
defaults, changes array.
- CLAUDE.md: register the command (## Commands 26 -> 27, table row).
- scripts/test-runner.sh: EXPECT_COMMANDS 26 -> 27 (registration guard).
Verify: grep 'subagent_type: linkedin-studio:engagement-coach' commands/ ->
firsthour.md; node --test state-updater -> 26/26; full hook suite -> 83/83;
bash scripts/test-runner.sh -> exit 0 (62 passed, commands 27/27).
Plan Step 16 (Wave 4 S3).
[skip-docs]: tre-doc + version bump deferred to Step 21 per remediation plan.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Resolve the video-advice contradiction on the two surfaces this step owns
(research/03 D1-D3):
- commands/video.md: rename the "3-Second Test" to a "Muted-Autoplay Test"
(front-load value for ~85% muted viewing; the "three-second hook" is
cross-platform folklore, not a LinkedIn signal); make captions the
enforceable spec (SRT or native auto-captions, indexed for search);
aspect ratio as guidance (4:5/1:1 preferred for broad distribution,
9:16 opt-in for the vertical video tab, crops to 1:1 on desktop); add
MP4-default + warn-only MOV/AVI + official upload limits to the
pre-recording reminder.
- references/linkedin-formats.md: reframe the 9:16 "distribution boost"
as uncorroborated heuristic; 4:5 "deprioritized" -> 4:5/1:1 preferred;
drop the "3-second hook"; MP4 safe-default + warn-only MOV/AVI; replace
"good video rewarded more than ever" with the honest "per-video reach
declining; documents/carousels out-engage video."
algorithm-signals-reference.md (the canonical magnitude source) was
already reconciled in Phase 0 (line 56 reads "declining ... 4:5/1:1
preferred, captions are the enforceable spec") and carries no boost copy,
so it is intentionally untouched here — and it is not in this step's
manifest.
Verify: grep 'must be 9:16|9:16 (1080|3-second hook' video.md
linkedin-formats.md -> none; 'captions' video.md -> 5; 'deprioritized'
linkedin-formats.md -> none; structure lint 61/61.
KNOWN RESIDUAL (flagged for follow-up, NOT in this step's scope):
- references/linkedin-growth-playbook-2025-2026.md:216 still carries a
"9:16 ... distribution boost" line. That file is owned by Step 17, but
Step 17's changes are newsletter-distribution — the video line would
fall through. Fold this into Step 17 or the review gate.
- references/video-strategy-guide.md:300 still says "the 3-second test
determines 70% of retention." That file is owned by NO plan step
(orphaned) — needs a home. Surface at session end for an operator
decision.
Plan Step 15 (Wave 4 S2). Counts unchanged.
[skip-docs]: tre-doc + version bump deferred to Step 21 per remediation plan.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wire the orphan differentiation-checker (#10) into the five short-form
creation commands (post/quick/react/carousel/video) as a De-AI /
Differentiation Gate at each command's quality-check step: confirm the
LinkedIn-named substance signals (personal substance, original thinking,
concrete specifics, genuine voice) + a soft engagement-bait check, and
delegate an originality pass to linkedin-studio:differentiation-checker
when the angle risks commodity content. Add Task to allowed-tools in
quick/react/carousel (post/video already had it from Step 13).
Extend (not duplicate) hooks/prompts/voice-guardian.md's AI-pattern
section with the same named signals from research/01 D8 + research/03 D4.
Runtime-loaded prompt — no compile-hooks.py, no hooks.json change
(verified: compile-hooks --check reports no drift).
Test: new hooks/scripts/__tests__/linkedin-content-filter.test.mjs pins
the content/non-content boundary the gate is scoped by (14 tests).
Full hook suite 76/76, structure lint 61/61.
Plan Step 14 (Wave 4 S2). Counts unchanged (26 commands / 19 agents).
[skip-docs]: tre-doc + version bump deferred to Step 21 per remediation plan.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Resolves the orphan-agent audit finding by the locked default: wire all, no deletions, so the agent count stays 19. Per agent, added Task to the target command's allowed-tools and a coherent 'subagent_type: linkedin-studio:<name>' delegation at a real point in the command's flow (not a token grep-match).
Wired (agents 1-9 of 11): video-scripter -> video.md (Step 4); content-optimizer -> post.md (Step 7 refinement) + ab-test.md (2a.4 optimized challenger); analytics-interpreter -> report.md (Step 7, report mode) + analyze.md (Step 2, interpret mode); content-planner -> batch.md (Step 2) + pipeline.md (Step 1); trend-spotter -> batch.md (Step 1) + pipeline.md (Step 1); network-builder -> outreach.md (Step 3a); strategy-advisor -> strategy.md (Step 3); voice-trainer -> setup.md (Step 3a); post-feedback-monitor -> calendar.md (publish action, 48h monitor).
Deferred to their dedicated steps: #10 differentiation-checker -> Step 14 (short-form de-AI gate), #11 engagement-coach -> Step 16 (first-hour command). Namespaced subagent_type form requires a session reload before the wired agents are invokable.
Verify: each of the 9 has >=1 invocation in commands/; structural lint 61/61 (counts 19/26/25/6 intact); agent-fixtures 35/35; hook tests 62/62. Three-doc + version reconciliation deferred to Step 21 per the locked plan [skip-docs].
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 3 / Step 12 of the remediation plan (Phase 1 — usable by a non-author).
Fix the discoverability defects the audit flagged:
- skills/linkedin-studio/SKILL.md (the auto-activating router): self-naming 'the
LinkedIn thought leadership plugin' -> 'LinkedIn Studio' (v3.0 rename leftover);
the 'All Agents' table corrected from 14 to the real 19 (added editorial-reviewer,
voice-scrubber, content-reviewer, language-reviewer, fact-reviewer); the 'All
Commands' table completed to 26 (added headless-review, pivot, carousel) so it
routes to newsletter/headless-review/pivot/react.
- commands/onboarding.md: '25 commands' -> '26 commands' (x2); pillar count '3-5
expertise areas' -> '5 expertise areas' (reconciles onboarding's 3-5 with
setup.md's '5 core topics' and the CLAUDE.md '5 core expertise areas' rule).
- commands/linkedin.md: router table + numbered option list gain headless-review
and pivot.
Scope note (operator decision, this session): the plan's Verify grep
'grep -rni "thought leadership" skills/ -> no matches' is broader than the brief's
actual criterion (which targets only the router skill's self-naming). 'thought
leadership' is a plugin-wide DOMAIN term — the '8 Thought Leadership Angles'
framework lives in references/thought-leadership-angles.md and is referenced by ~40
files (glossary, agents, post/video/batch commands). Renaming it only inside skills/
would create cross-file inconsistency; renaming it plugin-wide is a separate
vocabulary migration outside Step 12's discoverability scope. Per operator choice,
the router SELF-NAMING is fixed (brief criterion met) and the 4 remaining skills/
hits (linkedin-content-creation headings + one linkedin-strategy phase cell) are
legitimate domain usage kept consistent with the rest of the tree.
setup.md needed no edit: its pillar number was already '5'; reconciliation was a
one-sided fix in onboarding.
Verify: router no longer says 'thought leadership plugin'; grep -nc 'headless-review'
commands/linkedin.md -> 2; onboarding pillar count '5' matches setup.md; SKILL.md
agent table 19 rows, command table 26 rows; structural lint exit 0 (61 passed).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 3 / Step 11 of the remediation plan (Phase 1 — usable by a non-author).
The README claimed 'the version that ships is the version that's actually been
independently reviewed' — an overclaim the operator correction flagged: the cold
review is a CAPABILITY the pipeline can run, not a guarantee every shipped edition
received. Reframe to the honest capability:
- Long-Form Engine bullet: the draft CAN be frozen and re-read by reviewers with no
drafting-session context (argument, language, facts, reader-fit) before lock, so a
cold pass catches what the framing-biased in-session gates miss; run inline (Step
6.5) or via /linkedin:headless-review for maximum independence. No claim about what
every shipped edition went through.
- Pipeline diagram annotation '(independent re-read)' -> '(cold, headless)'.
Verify: grep -ni 'independently reviewed' README.md -> no matches.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 3 / Step 10 of the remediation plan (Phase 1 — usable by a non-author), per
research/02.
State what the plugin can and cannot do for a PERSONAL profile, dated 'as of
2026-05', without replacing one false claim with another:
- README.md: new 'Boundaries (as of 2026-05)' section. (a) post-level analytics API
EXISTS but is partner-gated (vetted Community Management app + verified org +
Page) — not self-serve; CSV is the practical floor; saves visible natively
(count-only, ~Sept 2025) with no self-serve API pull. (b) auto-publish is
technically POSSIBLE self-serve (w_member_social) but deliberately NOT built — a
design + ToS choice, explicitly 'not an impossibility' (avoids the new false
'cannot auto-publish' claim). (c) dwell internal-only for organic, not exportable.
- commands/calendar.md: the publish action now states plainly it marks a post YOU
posted manually as published — the tool does not post on your behalf.
- commands/report.md: dated the saves/no-self-serve-API note ('as of 2026-05').
- commands/import.md: dated 'Why CSV' note — API is partner-gated, CSV is the floor.
Verify: grep -niE 'cannot auto-publish|only way to (get|access)' README.md
commands/*.md -> no matches; grep -ni 'as of 2026' README.md -> 2.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 3 / Step 9 of the remediation plan (Phase 1 — usable by a non-author).
The long-form review layer shipped Norwegian-locked: language-reviewer graded
unconditional Norwegian, voice-scrubber's gold standard was 'approved Norwegian
editions', and the editorial/content craft gates pointed at a 'skrivekontrakt §C2'
that does not ship. A non-Norwegian adopter would get English prose graded against
Norwegian idiom and a gate that depends on an unshipped contract.
- config/edition-state.template.json: add additive 'language' field (top-level,
default 'en') + a _doc entry. Threads into the language-dependent agents.
- agents/language-reviewer.md: new 'Language parameter' section — Norwegian-specific
checks (anglicism->Norwegian idiom, kanselli-stil) apply only when language=='no';
any other value grades that language's equivalents and never flags idiomatic
English as an anglicism. Default 'en'.
- agents/voice-scrubber.md: gold standard reframed to 'approved editions in the
configured language'; the Norwegian-chronicle calibration is the language=='no'
instantiation.
- agents/editorial-reviewer.md + agents/content-reviewer.md: the in-tree checklist
is now the operative, self-contained source of truth; Maskinrommet §C2 is an
optional upstream contract that does NOT ship (available only on the author's
runs). The gates work for an adopter without it.
- commands/newsletter.md: thread 'language' through the Step 6.5 cold-inputs and the
per-reviewer call inputs; the writing contract is now 'if it ships'.
Norwegian remains fully working when language: no (the author's case).
fact-reviewer.md was in the plan's file list but needed no change on inspection:
its F1-F4 checks (claims/quotes/numbers/sources) are language-agnostic; its
'Norwegian' mentions are boundary notes vs language-reviewer, which stay correct.
[skip-docs]: three-doc + version reconciliation is Step 21 (pre-review-gate); these
intermediate Wave commits are not pushed before the /trekreview gate.
Verify: edition-state JSON parses + has top-level language 'en'; language-reviewer
has 'language ==' references and no unconditional-Norwegian assertion; editorial
§C2 reframed to in-tree fallback ('operative source', 'does not ship'); agent
fixtures 35/35 pass; structural lint exit 0 (61 passed).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 3 / Step 8 of the remediation plan (Phase 1 — usable by a non-author).
The flagship long-form engine shipped bespoke-as-general: a maintainer-private
absolute series path and a hardcoded 'Maskinrommet' brand in the renderers. Both
are now generalized so a non-author can run the pipeline without inheriting the
author's filesystem or publication identity.
- commands/newsletter.md + config/edition-state.template.json: series-root default
/Users/ktg/repos/maskinrommet/serier -> $HOME/linkedin-series; reconciled the
prose that called the maskinrommet folder 'the default' so it no longer
contradicts the new neutral default. LTL_SERIES_ROOT override contract preserved.
- render/build-linkedin.mjs + render/build-carousel.mjs: brand is now an LTL_BRAND
env-var (empty default -> generic). Samle-post title, carousel footer brand-span,
and cover-eyebrow fallback are de-branded; empty brand renders clean chrome. The
operator sets LTL_BRAND=Maskinrommet in their own env to re-brand (same pattern
as LTL_SERIES_ROOT).
- config/image-credit-caption.template.md: 'Maskinrommet-/serie-badge' -> generic
'serie-badge'.
Out of scope (Step 9): the residual 'Maskinrommet skrivekontrakt §C2/§A'
references in newsletter.md are the writing-contract generalization, handled next.
[skip-docs]: three-doc + version reconciliation is Step 21 (pre-review-gate, per
plan: 'LAST so it captures everything'). These intermediate Wave commits are NOT
pushed before the /trekreview gate, so the three-doc obligation (which governs
pushed changes) is satisfied at Step 21, not per local checkpoint commit.
Verify: grep -rIn '/Users/ktg' config/ commands/ render/ (excl .local) -> no
matches; grep -rn 'Maskinrommet' render/ -> no matches (de-branded); node --check
on both render scripts -> OK; LTL_SERIES_ROOT still present in newsletter.md.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 2 / Step 7 of the remediation plan (research/02 D2/D4).
The tool parses the LinkedIn analytics CSV export, which has no saves/dwell — and
there is no self-serve API to pull them. Stop implying it tracks them:
- report.md: replace "Saves (10x weight) and expert comments (7-9x) are the
highest-impact signals" with honest wording — saves rank highest in the
engagement ORDER (cite references/algorithm-signals-reference.md, not a restated
coefficient), but are visible only in native LinkedIn post analytics (count-only,
~Sept 2025+) with no self-serve API, so this tool does not auto-track them; dwell
is internal to LinkedIn for organic posts.
- types.ts: document why PostMetrics intentionally omits saves/dwell (no ingest
source) so a future contributor does not "add the missing fields".
- strategy.md: reconcile two saves references (signature-content criterion +
authority scorecard) to say the count is read from native LinkedIn analytics, not
captured by this tool.
No new metric field, no manual-entry feature (operator Q3).
Verify: report.md has no "Saves (10x"/"highest-impact signals" and does say "no
self-serve API"; tsc --noEmit clean; analytics suite 106/106; structural lint
0 failed (report.md cites the reference, so stat-consistency stays green).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 2 / Step 6 of the remediation plan.
Organic personal-post A/B tests gather a handful of posts per variant — far below
the volume a significance test needs — so the tool must not imply statistical
significance:
- Rename the results-table "Significant?" column to "Directional?" and define it
as "clears the ~20% minimum-meaningful-difference AND points the same way across
most posts" — a direction to test further, not a significant result.
- Reword the "20% significance rule" to a minimum-meaningful-difference effect-size
heuristic (explicitly NOT statistical significance).
- Replace the "3 = Medium, 5+ = High" confidence ladder with a directional-only
confidence section: treat every result as directional (not significant) given
realistic volume is well under ~50 conversions/variant; name a direction, not a
winner.
The 20% minimum-meaningful-difference threshold itself stays — it is a legitimate
effect-size heuristic; only the significance framing was the false claim.
Verify: no "Significant?"/"20% significance" remain; "directional" present;
structural lint 0 failed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 2 / Step 5 of the remediation plan (coupled criticals: voice-leak +
placeholder-detection).
Voice profile (the adopter-default leak):
- Ship a PII-free placeholder at authentic-voice-samples.md carrying a
<!-- VOICE_PLACEHOLDER --> sentinel + neutral default voice principles.
- Migrate the author's real profile to gitignored authentic-voice-samples.local.md
(already matched by *.local.md; added an explicit, commented .gitignore entry so
the intent is unmissable). NO git-history rewrite — the historical file is
attributed authorship, not a secret (per the plan threat model).
- Add authentic-voice-samples.template.md — a clean fill-in template for adopters.
- personalization-score.mjs: detect the sentinel (deterministic) instead of the
unreliable `[Your Name]` heuristic, so the placeholder scores 0 voice points and
a populated profile (sentinel removed) earns the 25.
- Both voice writers replace-not-append on the placeholder: setup.md (merge ->
replace-if-placeholder) and onboarding.md (append -> replace-if-placeholder), so
populating removes the sentinel; updated setup.md's stale heuristic table.
Operator decisions (deviations from plan-literal, approved this session):
- KEEP the plugin.json author name. The plan said scrub author -> neutral/org, but
that contradicts its own LICENSE reasoning (intentional MIT attribution) and all
5 sibling plugins keep author = the author; scrubbing only this one would create
inconsistency for zero security gain (the name is public-by-design). The voice
placeholder fully fixes the adopter-inheritance bug.
- Scrub the stale "January 2026 360Brew" brand from the plugin.json description and
the "360brew" keyword (locked decision: no publishable model name/date). This is
a Wave-1 propagation miss surfaced here because plugin.json was in Step 5's
touch-scope.
Flagged for follow-up (NOT done here — out of Session 2 scope):
- The lint's stat-consistency grep (scripts/test-runner.sh) scans references/,
commands/, skills/, hooks/prompts/, CLAUDE.md, README.md — but NOT
.claude-plugin/plugin.json, which is why the 360Brew brand slipped Wave 1.
Needs a Session-1-scoped lint extension to add plugin.json to the scan set.
- Readers (user-prompt-context.mjs, voice-guardian.md, state-update-reminder.md)
read the tracked .md (placeholder), per the plan. The operator's real voice now
lives in the gitignored .local.md, which nothing reads. To use it, readers + the
voice score should prefer .local.md (matching the user-profile.local.md
precedent). Deferred as a coherence follow-up for operator review.
Test-first: hooks/scripts/__tests__/personalization-score.test.mjs (red on the
placeholder scoring 25 under the old heuristic, green after the sentinel fix). Hook
suite 62/62, structural lint 0 failed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Wave 2 / Step 4 of the remediation plan (docs/remediation/plan.md).
PRIMARY (the real fresh-clone failure):
- scripts/analytics/node_modules is gitignored, so a fresh clone has neither
tsx nor csv-parse. Surface an idempotent `npm install --silent` prerequisite
at point-of-use in report.md (Step 1b) and import.md (Step 4).
DEVIATION FROM PLAN (correction-in-scope, to satisfy the plan's own Verify gate):
- The plan assumed prepending `npm install` was sufficient. Verified it is NOT:
the commands invoke the CLI with an absolute src/cli.ts path but from the
user's arbitrary CWD, and `node --import tsx` resolves the `tsx` specifier
relative to CWD, not the script. There is no global tsx, so the call still
fails with ERR_MODULE_NOT_FOUND from any CWD other than scripts/analytics.
- Complete fix: invoke the locally-installed tsx by its absolute
node_modules/.bin/tsx path in all CLI calls (report.md x10, import.md x3), so
they resolve from any working directory once the install above has run.
Verified: 0 ERR_MODULE_NOT_FOUND running `report` from /tmp.
SECONDARY (latent correctness / hardening):
- Add findPluginRoot(): walks up to the dir holding .claude-plugin/plugin.json
and anchors getAnalyticsRoot() on it, falling back to the legacy 4-up count.
MEASURED that ../../../../ already resolved to the plugin root from BOTH
src/utils and build/utils (both 4 levels deep), so the plan's "src-vs-build
depth miscalibration" premise was false — this is correct-by-construction
hardening (survives a future source move), not a live-bug fix.
- Reconcile cli.ts usage/help text: `node build/cli.js` -> `node --import tsx
src/cli.ts` (the real runtime).
- Fix report.md troubleshooting: "Verify tsx is available" -> the actual
install command on ERR_MODULE_NOT_FOUND.
Test-first: scripts/analytics/tests/storage-root.test.ts (red on missing
findPluginRoot export, green after). Full suite 106/106, tsc --noEmit clean,
structural lint 0 failed.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Set expectations up front: the plugin systematizes the work, it doesn't
substitute for the human effort. Letting AI drive lands you in the
forgettable middle; the judgment, genuine engagement, and effort that
make content worth reading stay with the user. Reinforces the README's
no-hype tone and filters for serious adopters over a viral shortcut.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The standalone "What's New in v3.1.0" section opened with internal
development narrative (Endring 9, Del 4 run, framing-bias) that means
nothing to a prospective adopter — still leading with a changelog, the
exact pattern the rewrite set out to fix. Follow llm-security: motivation
first, version detail only in the Version History table at the bottom.
- Remove the "## What's New in v3.1.0" section and its TOC entry
- README now flows intro → Why LinkedIn Studio Exists → What Is This?
→ Two Engines → Quick Start
- Nothing lost: v3.1.0 remains in the Version History table
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The README led with a ~600-word version-by-version wall and stacked
seven "What's New in vX.Y.Z" sections before reaching "What Is This?",
reading as a changelog rather than motivating adoption.
Restructure to the sibling-plugin pattern (config-audit / llm-security
/ voyage):
- Concise motivating intro paragraph (no version enumeration)
- Keep only the latest "What's New" (v3.1.0); full history stays in the
Version History table at the bottom
- New "Why LinkedIn Studio Exists" section: the real pain (blank-page
paralysis, inconsistency, algorithm opacity, generic AI slop, the
360Brew shift) mapped to what the plugin does about each
- New "Two Engines" section framing feed vs. long-form, with the
pre-lock quality gauntlet as the long-form differentiator
- Refresh "What Is This?" to the LinkedIn Studio / content-engine framing
- Swap the unverified "40+ ideas / Content Matrix" claim for the
documented "8 universal angles" (verification duty)
All accurate tables (26 commands, 19 agents, 9 hooks, 24 reference
docs) and the full Version History preserved verbatim.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
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>
BREAKING CHANGE: the marketplace slug, the agent namespace
(linkedin-studio:<agent>), and the runtime state-file path
(~/.claude/linkedin-studio.local.md) all change. Reinstall required;
existing state migrated in place (post metrics, streak, history preserved).
The /linkedin:* commands are unchanged — the command namespace is set
per-command in frontmatter and was always independent of the plugin slug.
Functionality is byte-identical to v2.4.0; this release is pure identity.
- dir + manifests: plugins/linkedin-studio + plugin.json + root marketplace.json
- agent namespace updated in commands/newsletter.md (only functional invoker)
- state path updated in 4 hook scripts + topic-rotation prompt + state template
- catch-all skill dir renamed skills/linkedin-studio (5 functional skills unchanged)
- docs + version bump to 3.0.0 across README badge, CHANGELOG, root README/CLAUDE.md
- historical records (CHANGELOG past entries, docs/ build artifacts,
config-audit v5.0.0 snapshots) intentionally retain the old slug
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
