Skip to main content
Glama
SEO Analytics MCP

SEO Analytics MCP

by jafforgehq

SEO Analytics MCP

MCP server for Google Search Console (GSC) + Google Analytics 4 (GA4).

How It Works

What .venv Means

  • .venv is this project's private Python environment.

  • It contains its own Python interpreter and dependencies for this repo only.

  • run-local.sh activates .venv so the MCP server runs with the correct package versions.

What stdio Means

  • stdio means the MCP client and server communicate through standard input/output streams.

  • The client starts run-local.sh, sends tool requests in, and reads JSON responses out.

  • No web server or HTTP endpoint is required for the default local setup.

  1. Your MCP client (Claude/Codex) launches run-local.sh as a stdio server process.

  2. run-local.sh prepares .venv, installs dependencies, loads config/.env, and validates config/service-account.json.

  3. The Python server (seo-analytics-mcp) receives tool calls, fetches data from GSC/GA4 using the service account, and computes merged analysis outputs.

  4. Results are returned to the MCP client as structured JSON.

Related MCP server: mcp-site-analyst

Deployment Model

  • This project is deployed as a local process on your machine via stdio.

  • It is not an HTTP API service by default and does not require cloud hosting.

  • To "deploy" it for another environment, clone the repo there, provide config/.env and config/service-account.json, then register that machine's run-local.sh path in the MCP client.

Prerequisites

  • Python 3.11+

Quick Start

  1. Create local config files from templates:

cp config/.env.example config/.env
cp config/service-account.example.json config/service-account.json
  1. Fill required values:

  • config/.env

    • DEFAULT_GSC_SITE_URL (required format: sc-domain:example.com)

    • DEFAULT_GA4_PROPERTY_ID

  • config/service-account.json

    • Replace all REPLACE_WITH_... fields with real Google service-account values.

  1. Run the server:

./run-local.sh

The script creates/uses .venv, installs dependencies, loads config/.env, validates credentials, and starts seo-analytics-mcp over stdio.

MCP Client Setup

Use the local launcher script as the MCP server command:

/absolute/path/to/analytics/run-local.sh

Codex App

Add server:

codex mcp add seo-analytics -- /absolute/path/to/analytics/run-local.sh

Verify:

codex mcp list

Remove:

codex mcp remove seo-analytics

Claude Code CLI

Add server:

claude mcp add -s project seo-analytics -- /absolute/path/to/analytics/run-local.sh

Verify:

claude mcp list

Remove:

claude mcp remove seo-analytics

Claude Desktop App (macOS)

  1. Quit Claude Desktop completely.

  2. Edit ~/Library/Application Support/Claude/claude_desktop_config.json.

  3. Add this server entry (merge into existing mcpServers if present):

{
  "mcpServers": {
    "seo-analytics": {
      "type": "stdio",
      "command": "/absolute/path/to/analytics/run-local.sh",
      "args": []
    }
  }
}
  1. Start Claude Desktop again.

  2. In a new chat, run tool capabilities to confirm connection.

Main Tools

  • capabilities

  • gsc_list_sites

  • gsc_top_pages

  • gsc_top_queries

  • ga4_landing_pages

  • analytics_merge_page_metrics

  • analytics_generate_action_items

  • analytics_trend_report

  • analytics_data_quality_report

Screenshots

These screenshots are example outputs produced by using Claude Code with this MCP server. The MCP provides the underlying GSC/GA4 data and analytics tools; dashboard rendering is not generated by default by the server itself. For privacy reasons, values shown in the screenshots are randomized; the actual workflow was executed against real data.

MCP Dashboard View 1 MCP Dashboard View 2 MCP Dashboard View 3

Troubleshooting

  • service-account.json still contains placeholder values

    • Replace all REPLACE_WITH_... fields in config/service-account.json.

  • No such file or directory: .../credentials/service-account.json

    • You have a stale credential env override from an older setup.

    • Re-add the MCP server so it runs the latest run-local.sh and uses config/service-account.json.

  • Operation not permitted when Claude Desktop starts MCP

    • macOS is blocking execution from protected folders (commonly ~/Documents).

    • Either grant Claude Desktop access in System Settings -> Privacy & Security (Files and Folders and/or Full Disk Access) or move the repo to a non-protected path like ~/dev/analytics.

    • After changing permissions/path, restart Claude Desktop and re-add the MCP entry if needed.

  • Missing property_id and DEFAULT_GA4_PROPERTY_ID is not set

    • Set DEFAULT_GA4_PROPERTY_ID in config/.env.

  • Missing site_url...

    • Set DEFAULT_GSC_SITE_URL in config/.env.

  • GSC permission error on https://... property

    • This server enforces GSC properties in sc-domain:<domain> format to avoid URL-prefix mismatch issues.

    • Use a Domain property in Search Console and grant your service account access there.

Security

  • Never commit config/service-account.json or config/.env.

  • Commit only template files:

    • config/service-account.example.json

    • config/.env.example

Available Tools

16 tools
analytics_data_quality_reportC

Show merge coverage and top URL mismatches between GSC and GA4.

Inputs: site_url, property_id, start_date, end_date, +2 more. Returns: object

analytics_generate_action_itemsC

Generate prioritized SEO and content action items from merged data.

Inputs: site_url, property_id, start_date, end_date, +4 more. Returns: object

analytics_merge_page_metricsC

Merge normalized GSC + GA4 page metrics into one dataset.

Inputs: site_url, property_id, start_date, end_date, +2 more. Returns: object

analytics_popularity_snapshotC

Get top pages by clicks, impressions, sessions, and conversions.

Inputs: site_url, property_id, start_date, end_date, +2 more. Returns: object

analytics_query_page_opportunitiesC

Find high-impression query/page pairs with weak CTR and weak on-page outcomes.

Inputs: site_url, property_id, start_date, end_date, +3 more. Returns: object

analytics_topic_clustersC

Extract high-impact query token clusters to guide content focus.

Inputs: site_url, start_date, end_date, min_query_impressions, +2 more. Returns: object

analytics_trend_reportC

Compare current vs previous period to surface gainers and decliners.

Inputs: site_url, property_id, start_date, end_date, +2 more. Returns: object

capabilitiesB

Show enabled connectors, defaults, and key analysis thresholds.

Inputs: no named parameters. Returns: object

ga4_channel_reportC

Break down GA4 landing pages by default channel group.

Inputs: property_id, start_date, end_date, top_n. Returns: object

ga4_landing_pagesC

Return GA4 landing pages with engagement and conversion metrics.

Inputs: property_id, start_date, end_date, top_n, +1 more. Returns: object

ga4_run_report_rawC

Run raw GA4 report with optional dimension/metric filters and ordering.

Inputs: property_id, start_date, end_date, dimensions, +8 more. Returns: object

gsc_list_sitesB

List Search Console properties available to the authenticated account.

Inputs: no named parameters. Returns: object

gsc_query_page_pairsC

Return query+page combinations from GSC for intent/page matching analysis.

Inputs: site_url, start_date, end_date, search_type, +2 more. Returns: object

gsc_search_analytics_rawC

Run a raw Search Console Search Analytics query with full options.

Inputs: site_url, start_date, end_date, dimensions, +6 more. Returns: object

gsc_top_pagesC

Return top pages from GSC with clicks, impressions, CTR, and position.

Inputs: site_url, start_date, end_date, search_type, +2 more. Returns: object

gsc_top_queriesC

Return top queries from GSC by clicks.

Inputs: site_url, start_date, end_date, search_type, +2 more. Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    B3.2/5.0

    Scored across 16 tools

    Disambiguation4/5

    Most tools have distinct purposes, such as merging data, generating reports, or querying specific platforms. However, some overlap exists between analytics tools like 'analytics_popularity_snapshot' and 'gsc_top_pages', which could cause confusion in selection for similar tasks.

    Naming Consistency3/5

    The naming is mixed, with prefixes like 'analytics_', 'ga4_', and 'gsc_' used consistently within groups, but the overall pattern is not uniform. Tools like 'capabilities' break the verb_noun convention, and there is no single naming style across all tools.

    Tool Count4/5

    With 16 tools, the count is slightly high but reasonable for an SEO analytics domain that covers data merging, reporting, and platform-specific queries. It provides comprehensive coverage without being overly bloated.

    Completeness5/5

    The tool set offers complete coverage for SEO analytics, including data integration from GSC and GA4, trend analysis, opportunity identification, and action item generation. There are no obvious gaps, supporting end-to-end workflows effectively.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      B
      quality
      D
      maintenance
      An MCP server integrating Google Analytics 4, Search Console, and Indexing API, enabling AI agents to run reports, inspect URLs, manage properties, and request indexing.
      16
      429 npm
      2
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      An MCP server that provides Google Analytics and Search Console data as tools for AI assistants, enabling natural language queries for web analytics, SEO performance, and site insights.
      13
      MIT
    • A
      license
      Not graded
      quality
      D
      maintenance
      MCP server that exposes the Google Search Console API, allowing LLMs to query SEO data, inspect URLs, manage sitemaps, and analyze search performance via natural language.
      47 npm
      MIT
    • A
      license
      C
      quality
      B
      maintenance
      MCP server providing 32 tools for Google Search Console and Google Analytics 4, enabling search analytics, URL inspection, indexing, and cross-platform analysis via natural language.
      81
      59 PyPI
      24
      MIT