I Built an Agentic SEO System with Claude Code (Google Search Console + Google Analytics)

This video demonstrates a multi-agent system that runs automated, data-driven SEO audits and recommendations using Google Search Console, Google Analytics, and external APIs. It shows how agents fetch real data, compute metrics with deterministic methods (e.g., pandas), save reports to Google Drive, and publish content to WordPress. The system supports per-site configuration, changelogs, CSV exports, prioritized action lists, and combined insights for conversion analysis.

• Architecture: Multi-agent design with site-specific configuration and skills invoked via cloud code; agents include system prompts to improve reliability.
• Data & methods: Pulls real data from Google Search Console, Google Analytics, and external APIs; calculations use deterministic methods (e.g., pandas) rather than relying on LLM arithmetic.
• Outputs: KPI overviews, keyword analysis, combined GA/GSC insights, prioritized action items, CSV exports saved to Google Drive, and optional publishing to WordPress.
• Reliability features: Per-site changelogs prevent repeated recommendations and agents call explicit API/methods for reproducible results.

Quotes:

Type a single command to get a complete SEO analysis for any website.

Agents use real Google Search Console and Analytics data — not random AI guesses.

Every site gets its own changelog so recommendations don’t repeat.