seo-intel 1.5.26 → 1.5.27

package/CHANGELOG.md

@@ -1,5 +1,18 @@
 # Changelog
 
+## 1.5.27 (2026-05-16)
+
+### MCP — three new free-tier read tools
+The MCP server (`seo-intel-mcp`) now exposes individual records, not just summaries. AI agents can drill from inventory into actual pages, keywords, and heading structures without leaving the agent chat.
+
+- **`get_pages(project, role?, limit?, offset?)`** — paginated page list with url, title, word count, status, click depth, and domain role. Filterable by role (target / owned / competitor). Returns total count for pagination math.
+- **`list_keywords(project, domain?, limit?)`** — top extracted keywords grouped by domain + location (title / h1 / h2 / meta / body). Use to surface what each site is targeting before running gap analysis.
+- **`get_headings(project, url, limit?)`** — heading structure (H1–H6) for a specific page. Returns ordered `{ level, text }` list. Useful for content-architecture comparisons between target and competitor pages.
+
+All three are **free tier** — no license required. Pairs naturally with the existing `list_projects` and `get_intel(raw)` to give AI agents a complete free-tier read surface: list projects → inspect inventory → drill into pages → read headings → analyze with the agent's own flagship LLM.
+
+Errors are returned as proper MCP `isError: true` responses with helpful guidance (e.g. `get_headings` on an unknown URL points the agent at `get_pages`).
+
 ## 1.5.26 (2026-05-16)
 
 ### New — MCP server (`seo-intel-mcp`)

package/mcp/server.js

@@ -116,11 +116,128 @@ server.registerTool(
   }
 );
 
+// ── Tool: get_pages (free) ────────────────────────────────────────────────
+server.registerTool(
+  'get_pages',
+  {
+    description: 'Paginated list of crawled pages for a project, with url, title, word count, status, and domain role. Use this to drill into individual pages after seeing the inventory summary from get_intel. Free tier.',
+    inputSchema: {
+      project: z.string().describe('Project slug'),
+      role: z.enum(['target', 'owned', 'competitor']).optional().describe('Filter by domain role'),
+      limit: z.number().int().positive().max(500).optional().describe('Max pages to return (default 50, max 500)'),
+      offset: z.number().int().nonnegative().optional().describe('Offset for pagination (default 0)'),
+    },
+  },
+  async ({ project, role, limit = 50, offset = 0 }) => {
+    try {
+      const db = getDb();
+      const whereParams = role ? [project, role] : [project];
+      const where = role ? 'd.project = ? AND d.role = ?' : 'd.project = ?';
+      const rows = db.prepare(
+        `SELECT p.url, p.title, p.word_count, p.status_code, p.click_depth,
+                d.domain, d.role
+         FROM pages p JOIN domains d ON d.id = p.domain_id
+         WHERE ${where}
+         ORDER BY d.role, d.domain, p.url
+         LIMIT ? OFFSET ?`
+      ).all(...whereParams, limit, offset);
+      const total = db.prepare(
+        `SELECT COUNT(*) AS n FROM pages p JOIN domains d ON d.id = p.domain_id WHERE ${where}`
+      ).get(...whereParams)?.n || 0;
+      const out = { project, role: role || 'any', total, returned: rows.length, offset, pages: rows };
+      return {
+        content: [{ type: 'text', text: JSON.stringify(out, null, 2) }],
+        structuredContent: out,
+      };
+    } catch (err) {
+      return { content: [{ type: 'text', text: `seo-intel error: ${err.message}` }], isError: true };
+    }
+  }
+);
+
+// ── Tool: list_keywords (free) ────────────────────────────────────────────
+server.registerTool(
+  'list_keywords',
+  {
+    description: 'Top extracted keywords for a project, grouped by domain. Each keyword has frequency, location (title/h1/h2/meta/body), and source domain. Use this to surface what each site is targeting before running gap analysis. Free tier.',
+    inputSchema: {
+      project: z.string().describe('Project slug'),
+      domain: z.string().optional().describe('Optional: filter to a single domain'),
+      limit: z.number().int().positive().max(1000).optional().describe('Max keywords to return (default 100, max 1000)'),
+    },
+  },
+  async ({ project, domain, limit = 100 }) => {
+    try {
+      const db = getDb();
+      const params = [project];
+      let where = 'd.project = ?';
+      if (domain) { where += ' AND d.domain = ?'; params.push(domain); }
+      params.push(limit);
+      const rows = db.prepare(
+        `SELECT k.keyword, k.location, d.domain, d.role, COUNT(*) AS freq
+         FROM keywords k
+           JOIN pages p ON p.id = k.page_id
+           JOIN domains d ON d.id = p.domain_id
+         WHERE ${where}
+         GROUP BY k.keyword, k.location, d.domain
+         ORDER BY freq DESC
+         LIMIT ?`
+      ).all(...params);
+      const out = { project, domain: domain || 'all', returned: rows.length, keywords: rows };
+      return {
+        content: [{ type: 'text', text: JSON.stringify(out, null, 2) }],
+        structuredContent: out,
+      };
+    } catch (err) {
+      return { content: [{ type: 'text', text: `seo-intel error: ${err.message}` }], isError: true };
+    }
+  }
+);
+
+// ── Tool: get_headings (free) ─────────────────────────────────────────────
+server.registerTool(
+  'get_headings',
+  {
+    description: 'Heading structure (H1–H6) for a specific page. Returns ordered list of { level, text }. Useful for content architecture comparisons between target and competitor pages. Free tier.',
+    inputSchema: {
+      project: z.string().describe('Project slug'),
+      url: z.string().describe('Exact page URL (as crawled). Get URLs from get_pages.'),
+      limit: z.number().int().positive().max(200).optional().describe('Max headings (default 50)'),
+    },
+  },
+  async ({ project, url, limit = 50 }) => {
+    try {
+      const db = getDb();
+      const page = db.prepare(
+        `SELECT p.id, p.title, p.word_count, d.domain, d.role
+         FROM pages p JOIN domains d ON d.id = p.domain_id
+         WHERE d.project = ? AND p.url = ?`
+      ).get(project, url);
+      if (!page) {
+        return {
+          content: [{ type: 'text', text: `No crawled page found for url="${url}" in project "${project}". Use get_pages to discover URLs.` }],
+          isError: true,
+        };
+      }
+      const headings = db.prepare(
+        `SELECT level, text FROM headings WHERE page_id = ? ORDER BY id LIMIT ?`
+      ).all(page.id, limit);
+      const out = { project, url, page_title: page.title, domain: page.domain, role: page.role, word_count: page.word_count, headings };
+      return {
+        content: [{ type: 'text', text: JSON.stringify(out, null, 2) }],
+        structuredContent: out,
+      };
+    } catch (err) {
+      return { content: [{ type: 'text', text: `seo-intel error: ${err.message}` }], isError: true };
+    }
+  }
+);
+
 async function main() {
   const transport = new StdioServerTransport();
   await server.connect(transport);
   // stderr is fine; the host typically surfaces this in its MCP logs panel.
-  console.error(`[seo-intel-mcp] v${VERSION} ready on stdio. Tools: list_projects, get_intel.`);
+  console.error(`[seo-intel-mcp] v${VERSION} ready on stdio. Tools: list_projects, get_intel, get_pages, list_keywords, get_headings.`);
 }
 
 main().catch(err => {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "seo-intel",
-  "version": "1.5.26",
+  "version": "1.5.27",
   "description": "Local Ahrefs-style SEO competitor intelligence. Crawl → SQLite → cloud analysis.",
   "type": "module",
   "license": "SEE LICENSE IN LICENSE",