claude-plugin-wordpress-manager 2.12.1 → 2.12.2

package/CHANGELOG.md

@@ -2,6 +2,29 @@
 
 All notable changes to the WordPress Manager plugin for Claude Code.
 
+## [2.12.2] — 2026-03-01
+
+### Fixed — Structured Data Tools Rewrite
+
+**sd_inject: content-block approach (replaces broken meta approach)**
+- JSON-LD now injected as `<!-- wp:html -->` block in post content
+- No mu-plugin or `register_post_meta()` required — works out of the box
+- Idempotent: re-injecting same @type replaces existing block, no duplicates
+
+**sd_list_schemas: content scanning (replaces broken meta scanning)**
+- Scans post content for `<script type="application/ld+json">` blocks
+- No longer depends on unregistered `_schema_json_ld` meta field
+
+**sd_validate: multi-block support**
+- Now finds ALL JSON-LD scripts on a page (not just the first)
+- Reports per-schema issues with type labels
+
+### Validated end-to-end on opencactus.com
+- sd_inject Article on post 2348: ✅
+- sd_list_schemas finds injected schema: ✅
+- sd_validate on live URL: ✅ (2 schemas detected)
+- Idempotency verified: ✅
+
 ## [2.12.1] — 2026-03-01
 
 ### Fixed — Critical MCP Tool Parameter Passing

package/docs/GUIDE.md

@@ -1,6 +1,6 @@
 # WordPress Manager - Guida Completa per Utenti e Amministratori
 
-**Versione:** 2.12.1
+**Versione:** 2.12.2
 **Ultimo aggiornamento:** 2026-03-01
 **Repository:** https://github.com/morrealev/wordpress-manager
 
@@ -110,7 +110,7 @@ WordPress Manager e un plugin per **Claude Code** (la CLI ufficiale di Anthropic
 ### Componenti del Plugin
 
 ```
-wordpress-manager/                          # v2.12.1
+wordpress-manager/                          # v2.12.2
 +-- .claude-plugin/plugin.json              # Manifest
 +-- .mcp.json                               # Server MCP bundled
 +-- LICENSE                                 # MIT + GPL-2.0-or-later
@@ -181,8 +181,8 @@ wordpress-manager/                          # v2.12.1
 |   +-- [DISTRIBUZIONE DIRETTA + CONTENT FACTORY - 4 skill]
 |   +-- wp-linkedin/                            # LinkedIn direct posting e analytics (v2.10.0)
 |   +-- wp-twitter/                             # Twitter/X direct posting e thread (v2.10.0)
-|   +-- wp-structured-data/                     # Schema.org/JSON-LD validation e injection (v2.12.1)
-|   +-- wp-content-generation/                  # AI content generation pipeline (v2.12.1)
+|   +-- wp-structured-data/                     # Schema.org/JSON-LD validation e injection (v2.12.2)
+|   +-- wp-content-generation/                  # AI content generation pipeline (v2.12.2)
 +-- hooks/                                  # 12 hook di sicurezza
 |   +-- hooks.json                              # 10 prompt + 2 command
 |   +-- scripts/                                # Script per hook command-type
@@ -619,9 +619,9 @@ IDEAZIONE -> BOZZA -> REVISIONE -> OTTIMIZZAZIONE -> PUBBLICAZIONE -> MONITORAGG
 
 **AI Content Optimization** (v2.6.0): Pipeline AI-driven in 5 step per ottimizzazione contenuti: headline scoring → readability analysis (Flesch-Kincaid) → SEO scoring → meta optimization → content freshness check. Include bulk triage per analisi rapida di tutti i contenuti con classificazione Quick Wins / Maintain / Deep Review / Outdated. Usa la skill `wp-content-optimization`.
 
-**AI Content Generation** (v2.12.1): Pipeline completa 7 step per generare contenuti da zero: brief → keyword research (GSC se disponibile) → outline (4 pattern: standard, tutorial, listicle, FAQ) → draft (calibrato sulla voce del sito) → SEO optimize → structured data injection → publish as draft. Procedure-based, usa tool MCP esistenti (`wp/v2`, `gsc_*`, `sd_*`). Usa la skill `wp-content-generation`.
+**AI Content Generation** (v2.12.2): Pipeline completa 7 step per generare contenuti da zero: brief → keyword research (GSC se disponibile) → outline (4 pattern: standard, tutorial, listicle, FAQ) → draft (calibrato sulla voce del sito) → SEO optimize → structured data injection → publish as draft. Procedure-based, usa tool MCP esistenti (`wp/v2`, `gsc_*`, `sd_*`). Usa la skill `wp-content-generation`.
 
-**Structured Data Management** (v2.12.1): Gestione completa Schema.org/JSON-LD con 3 tool MCP dedicati: `sd_validate` (validazione markup via URL o inline), `sd_inject` (iniezione JSON-LD nei post), `sd_list_schemas` (audit sitewide). 8 tipi supportati: Article, Product, FAQ, HowTo, LocalBusiness, Event, Organization, BreadcrumbList. Usa la skill `wp-structured-data`.
+**Structured Data Management** (v2.12.2): Gestione completa Schema.org/JSON-LD con 3 tool MCP dedicati: `sd_validate` (validazione markup via URL o inline), `sd_inject` (iniezione JSON-LD nei post), `sd_list_schemas` (audit sitewide). 8 tipi supportati: Article, Product, FAQ, HowTo, LocalBusiness, Event, Organization, BreadcrumbList. Usa la skill `wp-structured-data`.
 
 **Skill correlata**: `wp-content`, `wp-i18n`, `wp-content-repurposing`, `wp-programmatic-seo`, `wp-search-console`, `wp-content-optimization`, `wp-content-generation`, `wp-structured-data`
 
@@ -965,7 +965,7 @@ Le skill di sviluppo provengono da due fonti:
 - **3 skill strategia + SEO internazionale** (MIT) aggiunte in v2.3.0: Programmatic SEO, Content-Commerce Attribution, Multi-Language Network.
 - **3 skill distribuzione + SEO avanzato** (MIT) aggiunte in v2.4.0-v2.6.0: Social/Email Distribution, Google Search Console, AI Content Optimization.
 - **3 skill analytics + alerting + automazione** (MIT) aggiunte in v2.7.0-v2.9.0: Analytics (GA4/Plausible/CWV), Smart Alerting (Slack/SendGrid severity routing), Automated Workflows (triggers + multi-channel actions).
-- **4 skill distribuzione diretta + content factory** (MIT) aggiunte in v2.10.0-v2.12.1: LinkedIn Direct (posting + analytics), Twitter/X Direct (tweet + thread + metriche), Structured Data (Schema.org/JSON-LD), AI Content Generation (pipeline 7-step).
+- **4 skill distribuzione diretta + content factory** (MIT) aggiunte in v2.10.0-v2.12.2: LinkedIn Direct (posting + analytics), Twitter/X Direct (tweet + thread + metriche), Structured Data (Schema.org/JSON-LD), AI Content Generation (pipeline 7-step).
 
 ### Il Router Unificato
 
@@ -1194,7 +1194,7 @@ Aggiunte in v2.7.0-v2.9.0, queste skill completano il layer Observability + Auto
 
 ### Panoramica Skills Distribuzione Diretta + Content Factory (4)
 
-Aggiunte in v2.10.0-v2.12.1, queste skill completano il WCOP portando il punteggio totale da 8.8/10 a 9.2/10 (Distribution 9/10, Content Factory 10/10).
+Aggiunte in v2.10.0-v2.12.2, queste skill completano il WCOP portando il punteggio totale da 8.8/10 a 9.2/10 (Distribution 9/10, Content Factory 10/10).
 
 | Skill | Si attiva quando... | Risorse | Agent dedicato |
 |-------|---------------------|---------|----------------|
@@ -1238,8 +1238,8 @@ Le skill includono 31 script Node.js (`.mjs`) che eseguono analisi automatica de
 | `workflow_inspect.mjs` | Action channel config, automation plugins, custom REST endpoints, WP-Cron, webhook config (v2.9.0) |
 | `linkedin_inspect.mjs` | LinkedIn access token, profile info, API connectivity (v2.10.0) |
 | `twitter_inspect.mjs` | Twitter Bearer token, OAuth tokens, API v2 connectivity (v2.10.0) |
-| `schema_inspect.mjs` | SEO plugins (Yoast, Rank Math), existing JSON-LD in theme, Schema Pro (v2.12.1) |
-| `content_gen_inspect.mjs` | REST access, GSC credentials, pipeline step availability (v2.12.1) |
+| `schema_inspect.mjs` | SEO plugins (Yoast, Rank Math), existing JSON-LD in theme, Schema Pro (v2.12.2) |
+| `content_gen_inspect.mjs` | REST access, GSC credentials, pipeline step availability (v2.12.2) |
 
 ### WordPress Playground — Ambienti Disposable
 
@@ -2363,6 +2363,6 @@ bash ~/.claude/plugins/local/wordpress-manager/scripts/validate-wp-operation.sh
 
 ---
 
-*Guida v2.12.1 — WordPress Manager Plugin per Claude Code*
+*Guida v2.12.2 — WordPress Manager Plugin per Claude Code*
 *Ultimo aggiornamento: 2026-03-01*
 *WCOP Score: 8.8/10 (Tier 4+5 complete)*

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "claude-plugin-wordpress-manager",
-  "version": "2.12.1",
+  "version": "2.12.2",
   "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 43 skills, 12 agents, and security hooks. v2.12.0 completes WCOP Tier 6+7 with content generation + structured data.",
   "author": {
     "name": "vinmor",

package/servers/wp-rest-bridge/build/tools/schema.js

@@ -72,29 +72,35 @@ export const schemaHandlers = {
                     return { toolResult: { isError: true, content: [{ type: "text", text: "Invalid JSON in markup parameter." }] } };
                 }
             } else {
-                // Fetch URL and extract JSON-LD
+                // Fetch URL and extract ALL JSON-LD blocks
                 const response = await axios.get(url, { timeout: 15000 });
                 const html = response.data;
-                const jsonLdMatch = html.match(/<script[^>]*type="application\/ld\+json"[^>]*>([\s\S]*?)<\/script>/i);
-                if (!jsonLdMatch) {
-                    return { toolResult: { content: [{ type: "text", text: JSON.stringify({ valid: false, error: "No JSON-LD found on page", url }, null, 2) }] } };
+                const jsonLdRegex = /<script[^>]*type="application\/ld\+json"[^>]*>([\s\S]*?)<\/script>/gi;
+                const allJsonLd = [];
+                let match;
+                while ((match = jsonLdRegex.exec(html)) !== null) {
+                    try {
+                        const parsed = JSON.parse(match[1]);
+                        if (Array.isArray(parsed)) allJsonLd.push(...parsed);
+                        else allJsonLd.push(parsed);
+                    } catch { /* skip invalid JSON-LD blocks */ }
                 }
-                try {
-                    jsonLd = JSON.parse(jsonLdMatch[1]);
-                } catch {
-                    return { toolResult: { content: [{ type: "text", text: JSON.stringify({ valid: false, error: "Invalid JSON-LD on page", url }, null, 2) }] } };
+                if (allJsonLd.length === 0) {
+                    return { toolResult: { content: [{ type: "text", text: JSON.stringify({ valid: false, error: "No valid JSON-LD found on page", url }, null, 2) }] } };
                 }
+                jsonLd = allJsonLd;
             }
 
             // Basic Schema.org validation
             const issues = [];
             const schemas = Array.isArray(jsonLd) ? jsonLd : [jsonLd];
             for (const schema of schemas) {
-                if (!schema['@context'] || !schema['@context'].includes('schema.org')) {
-                    issues.push('Missing or invalid @context (should include schema.org)');
+                const schemaLabel = schema['@type'] || '(unnamed)';
+                if (!schema['@context'] || !String(schema['@context']).includes('schema.org')) {
+                    issues.push(`${schemaLabel}: Missing or invalid @context (should include schema.org)`);
                 }
                 if (!schema['@type']) {
-                    issues.push('Missing @type');
+                    issues.push(`${schemaLabel}: Missing @type`);
                 }
             }
 
@@ -119,12 +125,22 @@ export const schemaHandlers = {
                 '@context': 'https://schema.org',
                 '@type': schema_type,
                 ...schema_data,
-            });
-            // Store JSON-LD in post meta via WordPress REST API
-            const response = await makeWordPressRequest('POST', `posts/${post_id}`, {
-                meta: { _schema_json_ld: jsonLd },
-            });
-            return { toolResult: { content: [{ type: "text", text: JSON.stringify({ success: true, post_id, schema_type, stored: true }, null, 2) }] } };
+            }, null, 2);
+            const scriptBlock = `\n<!-- wp:html -->\n<script type="application/ld+json">\n${jsonLd}\n</script>\n<!-- /wp:html -->`;
+            // Fetch current post content
+            const post = await makeWordPressRequest('GET', `posts/${post_id}`, { _fields: 'content', context: 'edit' });
+            let content = post.content?.raw || post.content?.rendered || '';
+            // Remove existing JSON-LD block for this schema type if present
+            const existingPattern = new RegExp(
+                `\\n?<!-- wp:html -->\\n<script type="application/ld\\+json">\\n[\\s\\S]*?"@type":\\s*"${schema_type}"[\\s\\S]*?</script>\\n<!-- /wp:html -->`,
+                'g'
+            );
+            content = content.replace(existingPattern, '');
+            // Append new JSON-LD block
+            content = content.trimEnd() + scriptBlock;
+            // Update post
+            await makeWordPressRequest('POST', `posts/${post_id}`, { content });
+            return { toolResult: { content: [{ type: "text", text: JSON.stringify({ success: true, post_id, schema_type, method: 'content_block', note: 'JSON-LD injected as wp:html block in post content' }, null, 2) }] } };
         } catch (error) {
             const errorMessage = error.response?.data?.message || error.message;
             return { toolResult: { isError: true, content: [{ type: "text", text: `Error injecting schema: ${errorMessage}` }] } };
@@ -134,19 +150,24 @@ export const schemaHandlers = {
     sd_list_schemas: async (params) => {
         try {
             const { schema_type } = params;
-            // Fetch recent posts and check for JSON-LD in meta
-            const posts = await makeWordPressRequest('GET', 'posts', { per_page: 100, _fields: 'id,title,meta' });
+            // Fetch posts and scan content for JSON-LD script blocks
+            const posts = await makeWordPressRequest('GET', 'posts', { per_page: 100, _fields: 'id,title,content' });
             const schemas = {};
+            const jsonLdRegex = /<script[^>]*type="application\/ld\+json"[^>]*>([\s\S]*?)<\/script>/gi;
             for (const post of posts) {
-                const meta = post.meta?._schema_json_ld;
-                if (meta) {
+                const content = post.content?.rendered || '';
+                let match;
+                while ((match = jsonLdRegex.exec(content)) !== null) {
                     try {
-                        const parsed = JSON.parse(meta);
-                        const type = parsed['@type'] || 'Unknown';
-                        if (schema_type && type !== schema_type) continue;
-                        if (!schemas[type]) schemas[type] = { count: 0, posts: [] };
-                        schemas[type].count++;
-                        schemas[type].posts.push({ id: post.id, title: post.title?.rendered });
+                        const parsed = JSON.parse(match[1]);
+                        const items = Array.isArray(parsed) ? parsed : [parsed];
+                        for (const item of items) {
+                            const type = item['@type'] || 'Unknown';
+                            if (schema_type && type !== schema_type) continue;
+                            if (!schemas[type]) schemas[type] = { count: 0, posts: [] };
+                            schemas[type].count++;
+                            schemas[type].posts.push({ id: post.id, title: post.title?.rendered });
+                        }
                     } catch { /* skip invalid JSON */ }
                 }
             }