claude-plugin-wordpress-manager 2.14.0 → 2.14.1

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "wordpress-manager",
-  "version": "2.13.0",
-  "description": "Unified WordPress management plugin for Claude Code. Orchestrates Hostinger MCP, WP REST API bridge (148 tools incl. WooCommerce, Multisite, Mailchimp, Buffer, SendGrid, GSC, GA4, Plausible, CWV, Slack, Workflows, LinkedIn, Twitter, Schema), and WordPress.com MCP with 45 skills, 12 agents, and security hooks. v2.13.0 adds Content Framework: 3-phase pipeline from signals intelligence to editorial calendar to WordPress publishing.",
+  "version": "2.14.1",
+  "description": "Unified WordPress management plugin for Claude Code. Orchestrates Hostinger MCP, WP REST API bridge (148 tools incl. WooCommerce, Multisite, Mailchimp, Buffer, SendGrid, GSC, GA4, Plausible, CWV, Slack, Workflows, LinkedIn, Twitter, Schema), and WordPress.com MCP with 46 skills, 12 agents, and security hooks. v2.14.1 adds Editorial Kanban Dashboard and Context Snippet (Step 0) in content skills.",
   "author": {
     "name": "vinmor",
     "email": "morreale.v@gmail.com"
@@ -42,7 +42,10 @@
     "content-pipeline",
     "editorial-calendar",
     "signals-intelligence",
-    "structured-data"
+    "structured-data",
+    "dashboard",
+    "kanban",
+    "context-snippet"
   ],
   "mcpServers": "./.mcp.json"
 }

package/CHANGELOG.md

@@ -2,6 +2,28 @@
 
 All notable changes to the WordPress Manager plugin for Claude Code.
 
+## [2.14.1] — 2026-03-02
+
+### Added — Context Snippet (Step 0)
+
+Contesto editoriale automatico nelle content skill. Ogni invocazione di `wp-content-pipeline` o `wp-editorial-planner` mostra 3-5 righe di stato operativo prima di iniziare il workflow — "ecco cosa sto per toccare".
+
+**Step 0: CONTEXT nelle skill**
+- `wp-content-pipeline/SKILL.md` — Step 0 con pipeline slice: mostra brief attivi con status
+- `wp-editorial-planner/SKILL.md` — Step 0 con calendar slice: mostra prossima scadenza
+- Workflow aggiornato: `CONTEXT → SCAN → CONFIG → ...` e `CONTEXT → PLAN → BRIEF → ...`
+- Step 0 è informational-only: non blocca mai il workflow, gestisce file mancanti gracefully
+
+**CLI snippet in `context-scanner.mjs`**
+- Nuovo entry point: `node scripts/context-scanner.mjs --snippet --site=mysite [--slice=pipeline|calendar|signals]`
+- `renderContextSnippet()` enhanced: nuovo parametro `activeBriefs` per pipeline slice
+- Output su stderr (coerente con pattern MCP server)
+
+**Design document**
+- `docs/plans/2026-03-02-context-snippet-design.md` — architettura, formato, vincoli
+
+---
+
 ## [2.14.0] — 2026-03-02
 
 ### Added — Editorial Kanban Dashboard

package/docs/plans/2026-03-02-context-snippet-design.md

@@ -0,0 +1,232 @@
+# Context Snippet — Design Document (Fase B)
+
+**Data**: 2026-03-02
+**Versione**: 1.0.0
+**Stato**: Implementazione
+**Prerequisito**: [Dashboard Strategy](2026-03-02-dashboard-strategy.md) (Fase A completata in v2.14.0)
+**Deriva da**: Dashboard Strategy, sezione 6 — Fase B
+
+---
+
+## 1. Obiettivo
+
+Ogni content skill stampa automaticamente 3-5 righe di contesto editoriale operativo **prima** di iniziare il proprio workflow. Questo "Step 0" fornisce a Claude e all'operatore umano una fotografia istantanea dello stato corrente senza richiedere un'azione esplicita (come invocare `wp-dashboard`).
+
+```
+┌─────────────────────────────────────────────────────────────┐
+│                    LIVELLO 2: CONTESTO OPERATIVO            │
+│                                                             │
+│  Step 0 inline nelle content skill                          │
+│  Output: testo terminale (3-5 righe)                        │
+│  Trigger: automatico, ogni invocazione skill                │
+│  Contenuto: solo lo slice rilevante per la skill attiva     │
+│                                                             │
+│  "Ecco cosa sto per toccare"                                │
+└─────────────────────────────────────────────────────────────┘
+```
+
+---
+
+## 2. Architettura
+
+### 2.1 Componenti
+
+```
+scripts/context-scanner.mjs
+  ├── scanContentState()       ← legge .content-state/ (già esistente)
+  ├── aggregateMetrics()       ← calcola metriche (già esistente)
+  ├── renderContextSnippet()   ← formatta 3-5 righe terminal (già esistente)
+  └── CLI: --snippet --site=X  ← NUOVO: entry point per invocazione da skill
+
+skills/wp-content-pipeline/SKILL.md
+  └── Step 0: CONTEXT          ← NUOVO: istruzioni per Claude
+
+skills/wp-editorial-planner/SKILL.md
+  └── Step 0: CONTEXT          ← NUOVO: istruzioni per Claude
+```
+
+### 2.2 Flusso di Esecuzione
+
+```
+Utente invoca skill
+  → Claude legge SKILL.md
+    → Step 0: CONTEXT
+      → Claude legge .content-state/{site_id}.config.md
+      → Claude legge .content-state/{YYYY-MM}-editorial.state.md
+      → Claude legge .content-state/pipeline-active/*.brief.md
+      → Claude formatta e stampa snippet (3-5 righe)
+    → Step 1+: workflow normale della skill
+```
+
+### 2.3 Due Modalità di Rendering
+
+| Modalità | Quando | Come |
+|----------|--------|------|
+| **Claude-native** | Sempre (Step 0 nelle skill) | Claude legge i file direttamente e formatta il testo seguendo le istruzioni SKILL.md |
+| **Script CLI** | Opzionale, per automazione | `node scripts/context-scanner.mjs --snippet --site=mysite` |
+
+La modalità Claude-native è il default. La modalità CLI serve come backup e per uso in script/automazione.
+
+---
+
+## 3. Formato Snippet
+
+### 3.1 Formato Base (tutte le skill)
+
+```
+── Editorial Context ──────────────────────
+  mysite.example.com | 2026-03-01 → 2026-03-31
+  Pipeline: 2 draft → 1 ready → 0 scheduled
+  Posts: 3/12 pubblicati
+───────────────────────────────────────────
+```
+
+### 3.2 Slice: Pipeline (wp-content-pipeline)
+
+Aggiunge dettaglio sui brief attivi:
+
+```
+── Editorial Context ──────────────────────
+  mysite.example.com | 2026-03-01 → 2026-03-31
+  Pipeline: 2 draft → 1 ready → 0 scheduled
+  Briefs: BRF-2026-005 (ready), BRF-2026-006 (draft)
+  Posts: 3/12 pubblicati
+───────────────────────────────────────────
+```
+
+### 3.3 Slice: Calendar (wp-editorial-planner)
+
+Aggiunge prossima scadenza:
+
+```
+── Editorial Context ──────────────────────
+  mysite.example.com | 2026-03-01 → 2026-03-31
+  Pipeline: 2 draft → 1 ready → 0 scheduled
+  Next: Mar 18 — "Acqua Premium: La Rivoluzione..."
+  Posts: 3/12 pubblicati
+───────────────────────────────────────────
+```
+
+### 3.4 Slice: Signals (future — wp-analytics)
+
+Aggiunge anomalie rilevanti:
+
+```
+── Editorial Context ──────────────────────
+  mysite.example.com | 2026-03-01 → 2026-03-31
+  Pipeline: 2 draft → 1 ready → 0 scheduled
+  Signals: 3 anomalie | Top: /blog/benefici +45%
+  Posts: 3/12 pubblicati
+───────────────────────────────────────────
+```
+
+---
+
+## 4. Step 0 nelle Skill
+
+### 4.1 Template Step 0
+
+Ogni content skill riceve un blocco Step 0 identico nella struttura, con `sliceType` diverso:
+
+```markdown
+## Step 0: CONTEXT
+
+**What it does:** Provides editorial context before the main workflow begins.
+
+**Trigger:** Automatic — run this step at every skill invocation, before Step 1.
+
+**Procedure:**
+1. Determine the active site from user context or ask the user
+2. Read `.content-state/{site_id}.config.md` for site URL
+3. Read the most recent `.content-state/*-editorial.state.md`
+4. Count entries by status (planned, draft, ready, scheduled, published)
+5. Read `.content-state/pipeline-active/*.brief.md` for active brief IDs and statuses
+6. Display the context snippet
+
+**Format:**
+[specific to slice type]
+
+**If files are missing:**
+- If no config file exists → skip snippet, report "No site config found"
+- If no editorial calendar exists → show partial snippet with "no calendar"
+- If no active briefs → show "0 active briefs"
+- Never stop the workflow because of missing context — Step 0 is informational only
+```
+
+### 4.2 Differenze per Skill
+
+| Skill | sliceType | Riga Extra |
+|-------|-----------|------------|
+| `wp-content-pipeline` | `pipeline` | Lista brief attivi con status |
+| `wp-editorial-planner` | `calendar` | Prossima scadenza (data + titolo troncato) |
+| Future: `wp-analytics` | `signals` | Anomalie top dal signals-feed |
+
+---
+
+## 5. CLI Entry Point
+
+### 5.1 Aggiunta a context-scanner.mjs
+
+```javascript
+// CLI entry point (when run directly, not imported)
+if (process.argv[1] && process.argv[1].endsWith('context-scanner.mjs')) {
+  const args = process.argv.slice(2);
+  if (args.includes('--snippet')) {
+    // Parse --site=X and --slice=Y
+    // Run scanContentState + aggregateMetrics + renderContextSnippet
+    // Output to stderr (consistent with MCP server logging pattern)
+  }
+}
+```
+
+### 5.2 Interfaccia CLI
+
+```
+node scripts/context-scanner.mjs --snippet --site=mysite [--slice=pipeline|calendar|signals] [--month=2026-03]
+```
+
+| Flag | Required | Default | Description |
+|------|----------|---------|-------------|
+| `--snippet` | Yes | — | Attiva modalità snippet (senza questo flag, il modulo è solo libreria) |
+| `--site` | Yes | — | Site ID per il contesto |
+| `--slice` | No | `pipeline` | Tipo di slice: `pipeline`, `calendar`, `signals` |
+| `--month` | No | Mese corrente | Mese del calendario editoriale |
+
+---
+
+## 6. Vincoli
+
+| Vincolo | Valore | Motivazione |
+|---------|--------|-------------|
+| Overhead < 500ms | Sì | Step 0 non deve rallentare la skill |
+| Informational only | Sì | Step 0 non blocca mai il workflow |
+| No nuovi file .content-state/ | Sì | Step 0 legge, non scrive |
+| Stesso formato di renderContextSnippet() | Sì | Consistenza tra modalità |
+
+---
+
+## 7. Deliverable
+
+| File | Azione | Descrizione |
+|------|--------|-------------|
+| `scripts/context-scanner.mjs` | Modifica | Aggiunta CLI `--snippet` + rimozione label "stub" |
+| `skills/wp-content-pipeline/SKILL.md` | Modifica | Aggiunta Step 0: CONTEXT (pipeline slice) |
+| `skills/wp-editorial-planner/SKILL.md` | Modifica | Aggiunta Step 0: CONTEXT (calendar slice) |
+| `docs/plans/2026-03-02-context-snippet-design.md` | Nuovo | Questo documento |
+| `package.json` | Modifica | Bump `v2.14.0` → `v2.14.1` |
+| `CHANGELOG.md` | Modifica | Entry per context snippet |
+
+---
+
+## 8. Catena Documentale
+
+```
+docs/plans/2026-03-02-dashboard-strategy.md
+  ├── docs/plans/2026-03-02-dashboard-kanban-design.md        (Fase A — completata)
+  │     └── docs/plans/2026-03-02-dashboard-kanban-implementation.md
+  └── docs/plans/2026-03-02-context-snippet-design.md         ← QUESTO DOCUMENTO (Fase B)
+```
+
+---
+
+*Design document per Fase B del sistema dashboard — context snippet inline nelle content skill.*

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "claude-plugin-wordpress-manager",
-  "version": "2.14.0",
-  "description": "Unified WordPress management and development plugin for Claude Code. Orchestrates Hostinger MCP, WP REST API bridge (148 tools incl. 30 WooCommerce + 10 Multisite + 4 Webhooks + 7 Mailchimp + 5 Buffer + 6 SendGrid + 8 GSC + 6 GA4 + 4 Plausible + 4 CWV + 3 Slack + 4 Workflows + 5 LinkedIn + 5 Twitter + 3 Schema), and WordPress.com MCP with 46 skills, 12 agents, and security hooks. v2.14.0 adds Editorial Kanban Dashboard: self-contained HTML report generated from .content-state/ files.",
+  "version": "2.14.1",
+  "description": "Unified WordPress management and development plugin for Claude Code. Orchestrates Hostinger MCP, WP REST API bridge (148 tools incl. 30 WooCommerce + 10 Multisite + 4 Webhooks + 7 Mailchimp + 5 Buffer + 6 SendGrid + 8 GSC + 6 GA4 + 4 Plausible + 4 CWV + 3 Slack + 4 Workflows + 5 LinkedIn + 5 Twitter + 3 Schema), and WordPress.com MCP with 46 skills, 12 agents, and security hooks. v2.14.1 adds Context Snippet (Step 0) to content skills for automatic editorial context display.",
   "author": {
     "name": "vinmor",
     "email": "morreale.v@gmail.com"

package/scripts/context-scanner.mjs

@@ -421,9 +421,9 @@ export function aggregateMetrics(rawData, viewType = 'kanban') {
   };
 }
 
-// ── RENDER: Context Snippet (Fase B stub) ──────────────────────────
+// ── RENDER: Context Snippet (Livello 2 — inline nelle content skill) ──
 
-export function renderContextSnippet(metrics, sliceType = 'pipeline') {
+export function renderContextSnippet(metrics, sliceType = 'pipeline', activeBriefs = []) {
   const m = metrics;
   const period = m.calendarPeriod ? m.calendarPeriod.replace('..', ' → ') : 'no calendar';
   const lines = [
@@ -434,6 +434,15 @@ export function renderContextSnippet(metrics, sliceType = 'pipeline') {
     `───────────────────────────────────────────`,
   ];
 
+  if (sliceType === 'pipeline' && activeBriefs.length > 0) {
+    const briefSummary = activeBriefs
+      .slice(0, 3)
+      .map(b => `${b.briefId} (${b.status})`)
+      .join(', ');
+    const extra = activeBriefs.length > 3 ? ` +${activeBriefs.length - 3} more` : '';
+    lines.splice(3, 0, `  Briefs: ${briefSummary}${extra}`);
+  }
+
   if (sliceType === 'calendar' && m.nextDeadline) {
     lines.splice(3, 0, `  Next: ${m.nextDeadline.date} — "${m.nextDeadline.title?.substring(0, 40)}..."`);
   }
@@ -444,3 +453,43 @@ export function renderContextSnippet(metrics, sliceType = 'pipeline') {
 
   return lines.join('\n');
 }
+
+// ── CLI Entry Point ─────────────────────────────────────────────────
+// Usage: node scripts/context-scanner.mjs --snippet --site=mysite [--slice=pipeline] [--month=2026-03]
+
+async function cli() {
+  const args = process.argv.slice(2);
+  if (!args.includes('--snippet')) return;
+
+  const siteArg = args.find(a => a.startsWith('--site='));
+  if (!siteArg) {
+    process.stderr.write('Error: --site=<site_id> is required\n');
+    process.exit(1);
+  }
+  const siteId = siteArg.split('=')[1];
+
+  const sliceArg = args.find(a => a.startsWith('--slice='));
+  const sliceType = sliceArg ? sliceArg.split('=')[1] : 'pipeline';
+
+  const monthArg = args.find(a => a.startsWith('--month='));
+  const month = monthArg ? monthArg.split('=')[1] : undefined;
+
+  try {
+    const rawData = await scanContentState('.content-state', siteId, month);
+    const metrics = aggregateMetrics(rawData);
+    const snippet = renderContextSnippet(metrics, sliceType, rawData.briefs.active);
+    process.stderr.write(snippet + '\n');
+  } catch (err) {
+    process.stderr.write(`Context snippet error: ${err.message}\n`);
+    process.exit(1);
+  }
+}
+
+// Only run CLI when executed directly (not imported as module)
+const isMainModule = process.argv[1] && (
+  process.argv[1].endsWith('context-scanner.mjs') ||
+  process.argv[1].endsWith('context-scanner')
+);
+if (isMainModule) {
+  cli();
+}

package/skills/wp-content-pipeline/SKILL.md

@@ -26,13 +26,44 @@ The content pipeline orchestrates a structured workflow from content brief files
 ## Pipeline Workflow
 
 ```
-SCAN → CONFIG → VALIDATE → PUBLISH → DISTRIBUTE → UPDATE → ARCHIVE
+CONTEXT → SCAN → CONFIG → VALIDATE → PUBLISH → DISTRIBUTE → UPDATE → ARCHIVE
 ```
 
 Each step reads specific files, calls specific tools, and has defined decision points. If any step fails or requires user input, the pipeline stops and reports status -- it never silently skips steps.
 
 ---
 
+## Step 0: CONTEXT
+
+**What it does:** Provides editorial context before the main workflow begins. This is an informational step -- it never blocks the workflow.
+
+**Trigger:** Automatic -- run this step at every skill invocation, before Step 1.
+
+**Procedure:**
+
+1. Determine the target `site_id` from user context (explicit mention, active site, or ask the user)
+2. Read `.content-state/{site_id}.config.md` -- extract `site_url` from frontmatter
+3. Find the most recent `.content-state/*-editorial.state.md` file -- parse the editorial table and count entries by status (`planned`, `draft`, `ready`, `scheduled`, `published`). Extract `goals.posts_target` from frontmatter
+4. List `.content-state/pipeline-active/*.brief.md` files -- read each brief's `brief_id` and `status` from frontmatter
+5. Display the context snippet:
+
+```
+── Editorial Context ──────────────────────
+  {site_url} | {calendar period}
+  Pipeline: {draft_count} draft → {ready_count} ready → {scheduled_count} scheduled
+  Briefs: {brief_id} ({status}), {brief_id} ({status})
+  Posts: {published_count}/{posts_target} pubblicati
+───────────────────────────────────────────
+```
+
+**If files are missing:**
+- No config file → skip snippet, note "No site config found for {site_id}" and continue to Step 1
+- No editorial calendar → show partial snippet with "no calendar" in place of period
+- No active briefs → omit the Briefs line
+- Never stop the workflow because of missing context -- Step 0 is informational only
+
+---
+
 ## Step 1: SCAN
 
 **What it does:** Discovers brief files that are ready for processing.

package/skills/wp-editorial-planner/SKILL.md

@@ -26,11 +26,42 @@ The editorial planner manages monthly content calendars as `.state.md` files sto
 ## Workflow Overview
 
 ```
-PLAN --> BRIEF --> SCHEDULE --> SYNC
+CONTEXT --> PLAN --> BRIEF --> SCHEDULE --> SYNC
 ```
 
 ---
 
+## Step 0: CONTEXT
+
+**What it does:** Provides editorial context before the main workflow begins. This is an informational step -- it never blocks the workflow.
+
+**Trigger:** Automatic -- run this step at every skill invocation, before Step 1.
+
+**Procedure:**
+
+1. Determine the target `site_id` from user context (explicit mention, active site, or ask the user)
+2. Read `.content-state/{site_id}.config.md` -- extract `site_url` from frontmatter
+3. Find the most recent `.content-state/*-editorial.state.md` file -- parse the editorial table and count entries by status (`planned`, `draft`, `ready`, `scheduled`, `published`). Extract `goals.posts_target` from frontmatter
+4. Identify the next upcoming deadline: first non-published entry with a title, sorted by date
+5. Display the context snippet:
+
+```
+── Editorial Context ──────────────────────
+  {site_url} | {calendar period}
+  Pipeline: {draft_count} draft → {ready_count} ready → {scheduled_count} scheduled
+  Next: {date} — "{title truncated to 40 chars}..."
+  Posts: {published_count}/{posts_target} pubblicati
+───────────────────────────────────────────
+```
+
+**If files are missing:**
+- No config file → skip snippet, note "No site config found for {site_id}" and continue to Step 1
+- No editorial calendar → show partial snippet with "no calendar" in place of period
+- No upcoming deadline → omit the Next line
+- Never stop the workflow because of missing context -- Step 0 is informational only
+
+---
+
 ## Step 1: PLAN -- Create or update editorial calendar
 
 **When to run:** User asks to create a new plan or update an existing one.