DBAdmin/inbox/content_seo.md

SEO Machine — Dokumentacja architektury i workflow

1. Cel systemu (Problem → Rozwiązanie) SEO Machine to wyspecjalizowany workspace Claude Code do produkcji i utrzymania treści long‑form, zoptymalizowanych pod SEO i pod konwersję, w oparciu o:

komendy workflow (np. /research, /write, /optimize, /publish-draft, itd.) agentów specjalistycznych (analiza SEO, meta, linkowanie, CRO, nagłówki, performance) moduły analityczne w Pythonie (intencja, keyword density, readability, SEO score, itp.) integracje danych (GA4, Search Console, DataForSEO) pliki kontekstu (brand voice, styl, guidelines, keywords, internal links map…)

2. Diagram systemu (logic + przepływ danych) 2.1 Diagram architektury (warstwy)

flowchart TB
  U[User / Operator] --> CC[Claude Code Workspace]

  CC --> CMD[Custom Commands\n/research /write /rewrite /optimize\n/analyze-existing /priorities /publish-draft\n/landing-* /research-*]
  CC --> CTX[Context Pack\nbrand-voice\nwriting-examples\nstyle-guide\nseo-guidelines\ninternal-links-map\ntarget-keywords\ncompetitor-analysis]
  CC --> DIR[Workspace Directories\ntopics/\nresearch/\ndrafts/\nrewrites/\npublished/\nlanding-pages/\naudits/\nreview-required/]

  CMD --> AG[Specialized Agents\nContent Analyzer\nSEO Optimizer\nMeta Creator\nInternal Linker\nKeyword Mapper\nEditor\nPerformance\nHeadline Generator\nCRO Analyst\nLanding Page Optimizer]
  AG --> PY[Python Analysis Modules\nintent/keywords/readability/seo-score\ncontent-length comparator\nCRO analyzers\nopportunity scoring]
  PY --> DATA[Data Integrations\nGA4\nGSC\nDataForSEO]

  DATA --> PY
  PY --> AG
  AG --> OUT[Outputs\nbriefs\nreports\noptimized drafts\nmeta packs\nlink maps\npublish payloads]
  OUT --> DIR
  DIR --> CC
  CC --> U

2.2 Diagram “happy path” dla nowego artykułu

sequenceDiagram
  participant U as User
  participant C as Claude Code
  participant R as /research
  participant W as /write
  participant A as Agents (auto)
  participant O as /optimize
  participant P as /publish-draft
  participant FS as Workspace folders

  U->>C: /research [topic]
  C->>R: run research workflow
  R->>FS: save research brief -> research/brief-*.md

  U->>C: /write [topic|brief]
  C->>W: generate 2000-3000+ word article
  W->>FS: save draft -> drafts/*.md
  W->>A: auto-run SEO/meta/link/keyword agents
  A->>FS: save agent reports -> drafts/*-report.md

  U->>C: /optimize [draft file]
  C->>O: final audit + readiness score
  O->>FS: save optimization report -> drafts/optimization-report-*.md

  U->>C: /publish-draft [file]
  C->>P: WordPress REST + Yoast metadata
  P->>FS: move/copy -> published/*.md

1. Agenci (Specialized Agents) — zakres i odpowiedzialność

Reguła jakości: jeden agent = jedna odpowiedzialność + output jako raport/rekomendacje.

4.1 Content Analyzer (modułowy, data‑driven) Daje “Executive summary” i plan działań:

intent (informational / navigational / transactional / commercial) keyword density/clustering + stuffing risk długość vs top SERP readability (Flesch, FK grade, sentence complexity, passive voice) SEO quality rating 0–100 + breakdown

Output: raport + metryki + benchmarki + priorytety zmian. 4.2 SEO Optimizer

struktura nagłówków, snippet chances, linki, meta, UX wynik 0–100 + konkretne poprawki

4.3 Meta Creator

5 tytułów + 5 opisów + rekomendacje testowe + SERP preview

4.4 Internal Linker

3–5 linków wewnętrznych + anchor text + mapowanie ścieżki użytkownika bazuje na context/internal-links-map.md

4.5 Keyword Mapper

rozmieszczenie keywords, checklist placementu, LSI coverage, cannibalization risk

4.6 Editor

“humanity score” + before/after rewrites + eliminacja patternów AI

4.7 Performance

GA4/GSC/DataForSEO → priorytety, quick wins (pozycje 11–20), low CTR, spadki

4.8 CRO / Landing Page Optimizer

above‑the‑fold, CTA, trust signals, friction points + scoring 0–100 A/B testing recommendations

5. Moduły Python — jak wchodzą w workflow 5.1 SEO Analysis (5 modułów “core”)

search_intent_analyzer.py keyword_analyzer.py content_length_comparator.py readability_scorer.py seo_quality_rater.py

Zasada: Agent “Content Analyzer” jest UI/orkiestratorem; Python moduły dają liczby i benchmarki. 5.2 CRO Analysis (landing pages)

above_fold_analyzer.py, cta_analyzer.py, trust_signal_analyzer.py landing_page_scorer.py, landing_performance.py, cro_checker.py

5.3 Dodatkowe

opportunity scoring, engagement, competitor gaps, article planner, section writer, social research aggregator

6. Integracje danych (GA4/GSC/DataForSEO) — użycie 6.1 Jakie decyzje wspierają

wybór tematów i priorytetów identyfikacja quick wins ocena, czy rewrite ma sens (ROI) walidacja trendu i sezonowości wykrywanie spadków performance

6.2 Cache i rate‑limits

trzymanie cache w data_sources/cache/ rekomendacja: cache per zapytanie + TTL (np. 24h), żeby nie palić limitów i budżetu

7. Organizacja workspace (folders + “single source of truth”) Proponowany kontrakt folderów:

topics/ — surowe pomysły + notatki research/ — briefy i analizy istniejących treści drafts/ — artykuły w toku + raporty agentów review-required/ — do ręcznego review (editorial/SME/legal) rewrites/ — zaktualizowane wersje + diff/changelog published/ — final + gotowe pod WP landing-pages/ — LP content audits/ — CRO/SEO audyty

Reguła: każdy artifact powinien zawierać: topic, datę, wersję.

8. “Definition of Done” (DoD) dla artykułu 8.1 Content

2000+ słów (preferowane 2500–3000) unikalna wartość vs competitors konkretne przykłady, dane i źródła CTA i logiczny flow

8.2 SEO

primary keyword: 1–2% + H1 + pierwsze 100 słów + 2–3 H2 3–5 internal links, 2–3 external authority links meta title 50–60 chars, meta description 150–160 chars poprawna hierarchia H1>H2>H3 wynik SEO score ≥ (np.) 85/100

8.3 Readability

8–10 grade 15–20 słów / zdanie (średnio) akapity 2–4 zdania heading co 300–400 słów

9. Getting Started (instalacja + first run) 9.1 Prereqs

Claude Code zainstalowany Anthropic API account Python + pip

9.2 Instalacja Shellgit clone https://github.com/[your-username]/seomachine.gitcd seomachinepip install -r data_sources/requirements.txtclaude-code .``Show more lines 9.3 Uzupełnij context (najważniejsze) Wypełnij szablony:

context/brand-voice.md context/writing-examples.md context/style-guide.md context/seo-guidelines.md context/target-keywords.md context/internal-links-map.md context/competitor-analysis.md

9.4 Quick Start

/research [topic] /write [topic] /optimize drafts/.md /publish-draft drafts/.md (opcjonalnie)

10. Dodatkowy diagram: “moduły analizy” (SEO + CRO)

flowchart LR
  A[Content Analyzer Agent] --> SI[Search Intent Analyzer]
  A --> KA[Keyword Analyzer\nDensity + Clustering + LSI]
  A --> CL[Content Length Comparator\nSERP Benchmarks]
  A --> RS[Readability Scorer]
  A --> SQ[SEO Quality Rater\nScore 0-100 + breakdown]

  LP[Landing Page Optimizer] --> AF[Above-the-fold Analyzer]
  LP --> CTA[CTA Analyzer]
  LP --> TS[Trust Signal Analyzer]
  LP --> LPS[Landing Page Scorer]
  LP --> LPP["Landing Performance (GA4/GSC)"]

Chcesz, żebym zrobił to “w wersji repo-ready”? Mogę przygotować:

docs/architecture.md (ten tekst) docs/workflows.md (same flowy i checklisty) docs/commands.md (referencja komend + inputs/outputs) docs/agents.md (agent responsibilities + output templates) docs/data-integrations.md (GA4/GSC/DataForSEO + env vars)

Powiedz tylko: czy SEO Machine ma być stricte “Claude Code workspace” (jak w opisie), czy chcesz też wariant pod OpenClaw skill-suite (skill + sub-skille + slash dispatch)?

1. Diagram architektury (warstwy + przepływ danych)

flowchart TB
  %% =========================
  %% GLOBAL ENTRYPOINT
  %% =========================
  U[User / Operator] --> CC[Claude Code Workspace\nSEO Machine Runtime]
  %% CONTEXT PACK (STATIC)
  %% =========================
  subgraph CTX["Context & Guidelines (Single Source of Truth)"]
    direction TB
    BV[context/brand-voice.md\nVoice pillars • tone • messaging]
    WE[context/writing-examples.md\n3–5 exemplary posts + notes]
    SG[context/style-guide.md\nEditorial rules • formatting]
    SEO[context/seo-guidelines.md\nLength • density • meta • links]
    KW[context/target-keywords.md\nClusters • intent • long-tails]
    IL[context/internal-links-map.md\nKey pages • anchors • journeys]
    CA[context/competitor-analysis.md\nCompetitors • gaps • angles]
    FT[context/features.md\nBenefits • differentiation]
  end

  %% =========================
  %% WORKSPACE DIRECTORIES (ARTIFACTS)
  %% =========================
  subgraph FS["Workspace Directories (Artifacts)"]
    direction TB
    TOP[topics/\nRaw ideas]
    RES[research/\nBriefs + analyses]
    DRA[drafts/\nWIP articles + agent reports]
    REV[review-required/\nHuman review queue]
    REW[rewrites/\nUpdated posts + change summary]
    PUB["published/\nFinal content (ready)"]
    LP[landing-pages/\nLanding pages]
    AUD[audits/\nSEO/CRO audits]
  end

  %% =========================
  %% COMMANDS (ORCHESTRATION)
  %% =========================
  subgraph CMD["Custom Commands (Orchestration Layer)"]
    direction TB
    C1[research]
    C2[write]
    C3[optimize]
    C4[analyze-existing]
    C5[rewrite]
    C6[performance-review]
    C7[priorities]
    C8[publish-draft]
    C9["article (simplified flow)"]
    C10["scrub (de-AI pattern cleanup)"]

    subgraph RC["Research Commands"]
      direction TB
      R1[research-serp]
      R2[research-gaps]
      R3[research-trending]
      R4[research-performance]
      R5[research-topics]
    end

    subgraph LC["Landing Page Commands"]
      direction TB
      L1[landing-write]
      L2[landing-audit]
      L3[landing-research]
      L4[landing-competitor]
      L5[landing-publish]
    end
  end

  %% =========================
  %% SPECIALIZED AGENTS
  %% =========================
  subgraph AG["Specialized Agents (Expert Reviews)"]
    direction TB
    A1["Content Analyzer\n(Executive summary + metrics)"]
    A2["SEO Optimizer\n(On-page score + fixes)"]
    A3["Meta Creator\n(Title/desc variants + tests)"]
    A4["Internal Linker\n(placements + anchors)"]
    A5["Keyword Mapper\n(density + distribution)"]
    A6["Editor\n(humanity score + rewrites)"]
    A7["Performance\n(priorities + roadmap)"]
    A8["Headline Generator\n(variants + A/B plan)"]
    A9["CRO Analyst\n(CRO scoring + friction)"]
    A10["Landing Page Optimizer\n(CRO + SEO for LP)"]
  end

  %% =========================
  %% PYTHON MODULES (ANALYTICS ENGINE)
  %% =========================
  subgraph PY["Python Modules (Analysis Engine)"]
    direction TB

    subgraph SEO5["SEO Analysis Modules (Core 5)"]
      direction TB
      P1[search_intent_analyzer.py\nintent + confidence]
      P2[keyword_analyzer.py\ndensity + clustering + LSI + heatmap]
      P3[content_length_comparator.py\nSERP length benchmarks]
      P4[readability_scorer.py\nFlesch/FK + passive voice + complexity]
      P5[seo_quality_rater.py\nscore 0-100 + breakdown]
    end

    subgraph CRO6["CRO Modules (Landing Pages)"]
      direction TB
      CP1[above_fold_analyzer.py\nheadline/value prop/CTA/trust]
      CP2[cta_analyzer.py\nCTA quality + distribution]
      CP3[trust_signal_analyzer.py\nsocial proof + risk reversal]
      CP4[landing_page_scorer.py\nscore 0-100 + categories]
      CP5[landing_performance.py\nGA4/GSC performance trends]
      CP6[cro_checker.py\nbest-practice checklist]
    end

    subgraph AUX["Additional Modules"]
      direction TB
      X1["opportunity_scorer.py\nimpact score (0-100)"]
      X2[content_scorer.py\nhumanity/specificity/structure/SEO/readability]
      X3[engagement_analyzer.py\npatterns + hooks]
      X4[competitor_gap_analyzer.py\ngaps + opportunities]
      X5[article_planner.py\nplan from data]
      X6[section_writer.py\nsection-level guidance]
      X7[social_research_aggregator.py\nsocial inputs]
      X8[wordpress_publisher.py\npublish payload + Yoast fields]
    end
  end

  %% =========================
  %% DATA INTEGRATIONS
  %% =========================
  subgraph DATA["Data Integrations (Real-time Signals)"]
    direction TB
    GA4["(Google Analytics 4)"]
    GSC["(Google Search Console)"]
    DFS["(DataForSEO)"]
  end

  %% =========================
  %% WORDPRESS PUBLISHING
  %% =========================
  subgraph WP["Publishing Target"]
    direction TB
    WPR["(WordPress REST API)"]
    YO["(Yoast SEO Metadata\nvia MU-plugin bridge)"]
  end

  %% =========================
  %% MARKETING SKILLS (26)
  %% =========================
  subgraph MS["Marketing Skills (26 slash skills)"]
    direction TB
    MS1[Copywriting\n/copywriting /copy-editing]
    MS2[CRO\n/page-cro /form-cro /signup-flow-cro\n/onboarding-cro /popup-cro /paywall-upgrade-cro]
    MS3[Strategy\n/content-strategy /pricing-strategy\n/launch-strategy /marketing-ideas]
    MS4[Channels\n/email-sequence /social-content /paid-ads]
    MS5[SEO\n/seo-audit /schema-markup /programmatic-seo\n/competitor-alternatives]
    MS6[Analytics\n/analytics-tracking /ab-test-setup]
    MS7[Other\n/referral-program /free-tool-strategy /marketing-psychology]
  end

  %% =========================
  %% WORKFLOWS AS SUBGRAPHS
  %% =========================
  subgraph WF1["Workflow A — New Long-form Content"]
    direction TB
    WFA1[research topic] --> WFA2[Research Brief\nkeywords • SERP • gaps • outline]
    WFA2 --> WFA3[write topic brief]
    WFA3 --> WFA4[Draft Article\n2000–3000+ words\nH1/H2/H3 + links + meta]
    WFA4 --> WFA5[Auto Agent Reviews\nSEO/meta/link/keyword/editor/headlines]
    WFA5 --> WFA6[optimize draft]
    WFA6 --> WFA7[Publishing Readiness Score\n+ final fix list]
    WFA7 --> WFA8["/publish-draft (optional)"]
  end

  subgraph WF2["Workflow B — Update Existing Content"]
    direction TB
    WFB1[analyze-existing url file] --> WFB2[Content Health Score 0–100\n+ quick wins + strategic fixes]
    WFB2 --> WFB3[rewrite topic analysis]
    WFB3 --> WFB4[Updated Article + Change Summary\nstats refresh • gap fill • keep winners]
    WFB4 --> WFB5[optimize rewrite]
    WFB5 --> WFB6[Final SEO Audit + Readiness]
    WFB6 --> WFB7["/publish-draft (optional)"]
  end

  subgraph WF3["Workflow C — Landing Pages (CRO-driven)"]
    direction TB
    WFC1[landing-research topic] --> WFC2[Positioning + competitor LP insights]
    WFC2 --> WFC3[landing-write topic]
    WFC3 --> WFC4[Landing Page Draft\nvalue prop • CTA • trust]
    WFC4 --> WFC5[landing-audit file]
    WFC5 --> WFC6[CRO Score 0–100\n+ A/B test plan + priority fixes]
    WFC6 --> WFC7["landing-publish (optional)"]
  end

  %% =========================
  %% CONNECTIONS (WHO USES WHAT)
  %% =========================
  CC --> CTX
  CC --> CMD
  CC --> AG
  CC --> PY
  CC --> FS
  CC --> MS

  %% Commands drive workflows
  C1 --> WF1
  C2 --> WF1
  C3 --> WF1
  C8 --> WF1

  C4 --> WF2
  C5 --> WF2
  C3 --> WF2
  C8 --> WF2

  L3 --> WF3
  L1 --> WF3
  L2 --> WF3
  L5 --> WF3

  %% Agents consume context + produce artifacts
  CTX --> AG
  AG --> FS

  %% Python modules power agents
  AG --> PY
  PY --> AG

  %% Data feeds Python/Performance
  DATA --> PY
  GA4 --> DATA
  GSC --> DATA
  DFS --> DATA

  %% Saving artifacts to folders
  WFA2 --> RES
  WFA4 --> DRA
  WFA5 --> DRA
  WFA7 --> DRA
  WFB2 --> RES
  WFB4 --> REW
  WFC4 --> LP
  WFC6 --> AUD

  %% Review stage (optional)
  DRA --> REV
  REW --> REV
  REV --> PUB

  %% Publishing
  C8 --> WP
  WFA8 --> WP
  WFB7 --> WP
  WFC7 --> WP

  FS --> WPR
  WPR --> YO
  YO --> PUB

  %% Marketing skills feed drafts and landing pages
  MS --> DRA
  MS --> LP

2. Workflow: nowy artykuł (Research → Write → Optimize → Publish)

sequenceDiagram
  autonumber
  participant U as User
  participant C as Claude Code (SEO Machine)
  participant R as /research
  participant W as /write
  participant A as Auto Agents
  participant O as /optimize
  participant P as /publish-draft
  participant FS as Workspace (folders)

  U->>C: /research [topic]
  C->>R: Keyword + SERP + competitor analysis
  R->>FS: research/brief-[topic]-[date].md

  U->>C: /write [topic|brief]
  C->>W: Generate long-form article (2000-3000+)
  W->>FS: drafts/[topic]-[date].md
  W->>A: Trigger SEO/meta/link/keyword agents
  A->>FS: drafts/*-seo-report.md
  A->>FS: drafts/*-meta-options.md
  A->>FS: drafts/*-internal-links.md
  A->>FS: drafts/*-keyword-map.md

  U->>C: /optimize [draft file]
  C->>O: Final audit + publishing readiness score
  O->>FS: drafts/optimization-report-[topic]-[date].md

  U->>C: /publish-draft [file]
  C->>P: Publish to WP + Yoast fields
  P->>FS: published/[topic]-[date].md

3. Workflow: aktualizacja istniejącej treści (Analyze → Rewrite → Optimize → Publish)

sequenceDiagram
autonumber
participant U as User
participant C as Claude Code (SEO Machine)
participant AE as /analyze-existing
participant RW as /rewrite
participant O as /optimize
participant P as /publish-draft
participant FS as Workspace (folders)

U->>C: /analyze-existing [url|file]
C->>AE: Fetch + performance + content health scoring
AE->>FS: research/analysis-[topic]-[date].md

U->>C: /rewrite [topic|analysis file]
C->>RW: Update stats, gaps, SEO, keep what's working
RW->>FS: rewrites/[topic]-rewrite-[date].md
RW->>FS: rewrites/[topic]-change-summary-[date].md

U->>C: /optimize [rewrite file]
C->>O: Final polish + readiness score
O->>FS: drafts/optimization-report-[topic]-[date].md

U->>C: /publish-draft [file]
C->>P: Publish updated content to WP
P->>FS: published/[topic]-[date].md

4. Moduły SEO Analysis (5 modułów) + wynik “Content Analyzer”

flowchart LR
  CA[Content Analyzer Agent\nExecutive summary + action plan] --> SI["Search Intent Analyzer\n(intent + confidence)"]
  CA --> KA["Keyword Analyzer\n(density + clustering + LSI + heatmap)"]
  CA --> CL["Content Length Comparator\n(SERP benchmarks: median, p75, target)"]
  CA --> RS["Readability Scorer\n(Flesch, FK grade, passive voice, complexity)"]
  CA --> SQ["SEO Quality Rater\n(score 0-100 + category breakdown)"]

  SI --> CA
  KA --> CA
  CL --> CA
  RS --> CA
  SQ --> CA

5. Landing Page / CRO pipeline (audit → rekomendacje → test plan)

flowchart TB
  LWC[/landing-write or /landing-audit/] --> CRO[CRO Analyst + Landing Page Optimizer]
  CRO --> AF[above_fold_analyzer.py\nheadline/value prop/CTA/trust]
  CRO --> CTA[cta_analyzer.py\nCTA quality + distribution]
  CRO --> TS[trust_signal_analyzer.py\nsocial proof + risk reversal]
  CRO --> LPS[landing_page_scorer.py\nscore 0-100 + breakdown]
  CRO --> LPPerf[landing_performance.py\nGA4/GSC performance trends]
  CRO --> CK[cro_checker.py\nbest-practices checklist]

  CRO --> OUT[Outputs\nCRO score\npriority fixes\nA/B testing plan\ncopy suggestions\nwireframe notes]
  OUT --> FS["(audits/ + landing-pages/)"]

flowchart TB

  U["User / Operator"] --> CC["Claude Code Workspace\nSEO Machine Runtime"]

  subgraph CTX["Context & Guidelines\nSingle Source of Truth"]
    direction TB
    BV["context/brand-voice.md"]
    WE["context/writing-examples.md"]
    SG["context/style-guide.md"]
    SEOG["context/seo-guidelines.md"]
    KW["context/target-keywords.md"]
    IL["context/internal-links-map.md"]
    COMP["context/competitor-analysis.md"]
    FT["context/features.md"]
  end

  subgraph FS["Workspace Directories\nArtifacts + Outputs"]
    direction TB
    TOP["topics/"]
    RES["research/"]
    DRA["drafts/"]
    REV["review-required/"]
    REW["rewrites/"]
    PUB["published/"]
    LP["landing-pages/"]
    AUD["audits/"]
  end

  subgraph CMD["Custom Commands\nOrchestration Layer"]
    direction TB
    C1["/research"]
    C2["/write"]
    C3["/optimize"]
    C4["/analyze-existing"]
    C5["/rewrite"]
    C6["/performance-review"]
    C7["/priorities"]
    C8["/publish-draft"]
    C9["/article"]
    C10["/scrub"]

    subgraph RC["Research Commands"]
      direction TB
      R1["/research-serp"]
      R2["/research-gaps"]
      R3["/research-trending"]
      R4["/research-performance"]
      R5["/research-topics"]
    end

    subgraph LC["Landing Page Commands"]
      direction TB
      L1["/landing-write"]
      L2["/landing-audit"]
      L3["/landing-research"]
      L4["/landing-competitor"]
      L5["/landing-publish"]
    end
  end

  subgraph AG["Specialized Agents\nExpert Reviews"]
    direction TB
    A1["Content Analyzer"]
    A2["SEO Optimizer"]
    A3["Meta Creator"]
    A4["Internal Linker"]
    A5["Keyword Mapper"]
    A6["Editor"]
    A7["Performance"]
    A8["Headline Generator"]
    A9["CRO Analyst"]
    A10["Landing Page Optimizer"]
  end

  subgraph PY["Python Modules\nAnalysis Engine"]
    direction TB

    subgraph SEO5["SEO Analysis\nCore 5 Modules"]
      direction TB
      P1["search_intent_analyzer.py"]
      P2["keyword_analyzer.py"]
      P3["content_length_comparator.py"]
      P4["readability_scorer.py"]
      P5["seo_quality_rater.py"]
    end

    subgraph CRO6["CRO Modules\nLanding Pages"]
      direction TB
      CP1["above_fold_analyzer.py"]
      CP2["cta_analyzer.py"]
      CP3["trust_signal_analyzer.py"]
      CP4["landing_page_scorer.py"]
      CP5["landing_performance.py"]
      CP6["cro_checker.py"]
    end

    subgraph AUX["Additional Modules"]
      direction TB
      X1["opportunity_scorer.py"]
      X2["content_scorer.py"]
      X3["engagement_analyzer.py"]
      X4["competitor_gap_analyzer.py"]
      X5["article_planner.py"]
      X6["section_writer.py"]
      X7["social_research_aggregator.py"]
      X8["wordpress_publisher.py"]
    end
  end

  subgraph DATA["Data Integrations\nSignals"]
    direction TB
    GA4["Google Analytics 4"]
    GSC["Google Search Console"]
    DFS["DataForSEO"]
  end

  subgraph WP["Publishing Target\nWordPress + Yoast"]
    direction TB
    WPR["WordPress REST API"]
    YO["Yoast SEO fields\nMU-plugin bridge"]
  end

  subgraph MS["Marketing Skills\n26 slash skills"]
    direction TB
    MS1["Copywriting\n/copywriting\n/copy-editing"]
    MS2["CRO skills\n/page-cro\n/form-cro\n/onboarding-cro"]
    MS3["Strategy\n/content-strategy\n/pricing-strategy"]
    MS4["Channels\n/email-sequence\n/social-content\n/paid-ads"]
    MS5["SEO skills\n/seo-audit\n/schema-markup\n/programmatic-seo"]
    MS6["Analytics\n/analytics-tracking\n/ab-test-setup"]
    MS7["Other\n/referral-program\n/free-tool-strategy"]
  end

  subgraph WF1["Workflow A\nNew Long-form Content"]
    direction TB
    WF1a["/research topic"] --> WF1b["Research brief"]
    WF1b --> WF1c["/write topic or brief"]
    WF1c --> WF1d["Draft article\n2000–3000+ words"]
    WF1d --> WF1e["Auto agent reviews"]
    WF1e --> WF1f["/optimize draft"]
    WF1f --> WF1g["Readiness score"]
    WF1g --> WF1h["/publish-draft optional"]
  end

  subgraph WF2["Workflow B\nUpdate Existing Content"]
    direction TB
    WF2a["/analyze-existing url or file"] --> WF2b["Health score 0–100"]
    WF2b --> WF2c["/rewrite topic or analysis"]
    WF2c --> WF2d["Updated article\nChange summary"]
    WF2d --> WF2e["/optimize rewrite"]
    WF2e --> WF2f["Final audit\nReadiness"]
    WF2f --> WF2g["/publish-draft optional"]
  end

  subgraph WF3["Workflow C\nLanding Pages (CRO)"]
    direction TB
    WF3a["/landing-research topic"] --> WF3b["Positioning insights"]
    WF3b --> WF3c["/landing-write topic"]
    WF3c --> WF3d["Landing page draft"]
    WF3d --> WF3e["/landing-audit file"]
    WF3e --> WF3f["CRO score 0–100\nA/B plan"]
    WF3f --> WF3g["/landing-publish optional"]
  end

  %% Wiring
  CC --> CTX
  CC --> CMD
  CC --> AG
  CC --> PY
  CC --> FS
  CC --> MS

  CTX --> AG
  AG --> PY
  PY --> AG
  DATA --> PY

  %% Commands to workflows
  C1 --> WF1
  C2 --> WF1
  C3 --> WF1
  C8 --> WF1

  C4 --> WF2
  C5 --> WF2
  C3 --> WF2
  C8 --> WF2

  L3 --> WF3
  L1 --> WF3
  L2 --> WF3
  L5 --> WF3

  %% Artifacts to folders
  WF1b --> RES
  WF1d --> DRA
  WF1e --> DRA
  WF1g --> DRA

  WF2b --> RES
  WF2d --> REW

  WF3d --> LP
  WF3f --> AUD

  %% Review handoff
  DRA --> REV
  REW --> REV
  REV --> PUB

  %% Publishing
  C8 --> WP
  WF1h --> WP
  WF2g --> WP
  WF3g --> WP

  FS --> WPR
  WPR --> YO
  YO --> PUB