seo-gsc-cannibalization

GSC Cannibalization — Keyword Cannibalization Detection

@skills/seo/references/mcp-degradation.md @skills/seo/references/gsc-api-reference.md

Finds queries where multiple pages compete for the same ranking, causing them to cannibalize each other's traffic. Identifies the "winning" page and "losing" pages for each cannibalized keyword.

MCP Check

Before calling any GSC tool, verify the MCP is connected:

1. Use ToolSearch with query +google-search-console
2. If tools returned — note the actual tool name prefix, proceed to Inputs
3. If no tools returned — display the GSC MCP error template from references/mcp-degradation.md and stop:

## Google Search Console MCP Not Available

The `/seo gsc cannibalization` command requires the GSC MCP, which is not currently connected.

**What you can do:**
- Use `/seo technical <url>` for crawlability and indexability analysis (no live data)
- Use `/seo audit <url>` for a full static SEO audit

**To connect GSC MCP:**
- Install and configure a Google Search Console MCP server (see README for setup)
- Add it to ~/.claude/mcp.json at user scope (NOT project scope)
- Verify GSC property access before running commands (domain vs URL prefix format)
- See references/gsc-api-reference.md for property format details

Inputs

• site: The GSC property URL. Accept both formats:
  • Domain property: sc-domain:example.com
  • URL prefix: https://example.com or https://www.example.com
  • If user provides a bare domain (no prefix), call list_sites to identify the correct property format registered in GSC.

Date Calculation

Use Bash to calculate dates (GSC has ~3 day delay):

endDate=$(date -v-3d +%Y-%m-%d)
startDate=$(date -v-31d +%Y-%m-%d)
echo "endDate: $endDate | startDate: $startDate"

Execution

Step 1 — Pull query+page data: Call query_search_analytics with:

• siteUrl: the site property
• startDate: calculated startDate
• endDate: calculated endDate
• dimensions: ["query", "page"]
• rowLimit: 1000