gsc — Claude Code MCP server: what it does, how to install
gsc — an MCP server for Google Search Console API and Indexing API: search analytics, index status checks, sitemap submission, and structured data notifications.
MIT ★ 8 updated 2026-10-19
claude mcp add --env GSC_CLIENT_ID=<GSC_CLIENT_ID> --env GSC_CLIENT_SECRET=<GSC_CLIENT_SECRET> --env GSC_REFRESH_TOKEN=<GSC_REFRESH_TOKEN> --env GSC_SERVICE_ACCOUNT_KEY_PATH=<GSC_SERVICE_ACCOUNT_KEY_PATH> --transport stdio gsc -- npx -y @mikusnuz/gsc-mcpWhat gsc does
The server exposes all Search Console and Indexing API endpoints through 13 tools. You can query search performance data (clicks, impressions, CTR, position) grouped by device or date, inspect URL index status, submit sitemaps, and add properties.
Indexing API accepts notifications only for pages with JobPosting structured data or livestream pages with BroadcastEvent embedded in a VideoObject. Google confirms receipt but does not guarantee indexing. Authentication requires an OAuth2 refresh token or a service account JSON key. The API has a daily limit of 200 publish requests per project; a batch of 100 URLs counts as 100 requests.
For Indexing API tools, the service account must be added as a delegated owner in Search Console. User-level permission is insufficient.
What it helps with
- query search performance for the last 30 days
- notify Google about an updated job posting or livestream
- diagnose why a page is not indexed
- compare click-through rates between mobile and desktop
- submit a sitemap to Search Console
Example prompts for the agent
- Show me the top 10 queries from the last 30 days
- Check the index status of
/careers/developer - Send a notification for the job posting at
/job/12345
How to install gsc
Platform CLI
claude mcp add --env GSC_CLIENT_ID=<GSC_CLIENT_ID> --env GSC_CLIENT_SECRET=<GSC_CLIENT_SECRET> --env GSC_REFRESH_TOKEN=<GSC_REFRESH_TOKEN> --env GSC_SERVICE_ACCOUNT_KEY_PATH=<GSC_SERVICE_ACCOUNT_KEY_PATH> --transport stdio gsc -- npx -y @mikusnuz/gsc-mcpReplace <…> with your own values
Platform CLI
codex mcp add gsc --env GSC_CLIENT_ID=<GSC_CLIENT_ID> --env GSC_CLIENT_SECRET=<GSC_CLIENT_SECRET> --env GSC_REFRESH_TOKEN=<GSC_REFRESH_TOKEN> --env GSC_SERVICE_ACCOUNT_KEY_PATH=<GSC_SERVICE_ACCOUNT_KEY_PATH> -- npx -y @mikusnuz/gsc-mcpReplace <…> with your own values
Config file
{
"mcpServers": {
"gsc": {
"command": "npx",
"args": [
"-y",
"@mikusnuz/gsc-mcp"
],
"env": {
"GSC_CLIENT_ID": "<GSC_CLIENT_ID>",
"GSC_CLIENT_SECRET": "<GSC_CLIENT_SECRET>",
"GSC_REFRESH_TOKEN": "<GSC_REFRESH_TOKEN>",
"GSC_SERVICE_ACCOUNT_KEY_PATH": "<GSC_SERVICE_ACCOUNT_KEY_PATH>"
}
}
}
}Replace <…> with your own values
Config file
{
"servers": {
"gsc": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"@mikusnuz/gsc-mcp"
],
"env": {
"GSC_CLIENT_ID": "<GSC_CLIENT_ID>",
"GSC_CLIENT_SECRET": "<GSC_CLIENT_SECRET>",
"GSC_REFRESH_TOKEN": "<GSC_REFRESH_TOKEN>",
"GSC_SERVICE_ACCOUNT_KEY_PATH": "<GSC_SERVICE_ACCOUNT_KEY_PATH>"
}
}
}
}Replace <…> with your own values
Platform CLI
gemini mcp add -e GSC_CLIENT_ID=<GSC_CLIENT_ID> -e GSC_CLIENT_SECRET=<GSC_CLIENT_SECRET> -e GSC_REFRESH_TOKEN=<GSC_REFRESH_TOKEN> -e GSC_SERVICE_ACCOUNT_KEY_PATH=<GSC_SERVICE_ACCOUNT_KEY_PATH> gsc npx -y @mikusnuz/gsc-mcpReplace <…> with your own values
Step by step, with commands for each platform: How to add an MCP server to Claude Code, Codex, Cursor, VS Code, Gemini CLI and more
Questions about gsc
How do I install gsc in Claude Code?
Run `claude mcp add gsc -e GSC_CLIENT_ID=<GSC_CLIENT_ID> -e GSC_CLIENT_SECRET=<GSC_CLIENT_SECRET> -e GSC_REFRESH_TOKEN=<GSC_REFRESH_TOKEN> -- npx -y @mikusnuz/gsc-mcp. You can use a service account JSON key instead by setting GSC_SERVICE_ACCOUNT_KEY_PATH`.
What does gsc do?
The server connects Google Search Console API and Indexing API to your AI agent. You can query search analytics, inspect URL index status, and send notifications for job postings and livestreams.
Which pages can use the Indexing API?
Google accepts notifications only for pages with JobPosting structured data or livestream pages with BroadcastEvent embedded in a VideoObject. Other content types are not supported.