B — Some risk signals Graded October 16, 2026Minor concerns worth a quick check.Based on public adoption and transparency signals as of October 16, 2026. Our opinion, not a code audit or security guarantee.
AI agent skill · Data & databases, Web & browsing
open-seo-mcp-skills
Free SEO MCP server + open-source SEO and GEO skills for Claude: keyword research, rank tracking, audits, backlinks, AI visibility on your real GSC/GA4/ads data. claude mcp add ryze --transport http https://connector.get-ryze.ai/mcp
Bottom line: Signal-based grade (4/7 points): very widely starred on GitHub, public source code. This reflects adoption and transparency, not a code audit.
- Popularity score
- 42/100
- Rank
- #259 of 500
- GitHub stars
- 5.8k
What it does
Connects Claude to your real SEO and marketing data by integrating with Google Search Console, Google Analytics 4, and Google Ads. Enables AI agents to perform keyword research, track search rankings, audit site performance, analyze backlinks, and assess AI visibility—all against your live account data.
Who it’s for
Marketing teams, SEO professionals, and builders who want Claude to analyze and act on authentic search and advertising metrics. Best suited for organizations already using GSC, GA4, and Ads and comfortable connecting those accounts to an MCP server.
Copy-paste setup
Check these steps against the project’s own instructions before running anything.
Follow the setup steps in the project’s official README: https://github.com/Ryze-AI-Adgent/open-seo-mcp-skills (opens external site)
Risks
- Credential scope creep: OAuth tokens to GSC, GA4, and Ads grant broad read/write access; confirm the MCP server requests minimal necessary scopes only.
- Data exfiltration: SEO and traffic data sent to Claude's servers; review Anthropic's data retention policy if you handle sensitive competitive or client data.
- Prompt injection: Malicious queries could manipulate the MCP to extract or modify Google Ads campaigns or bidding strategies; validate input sanitization.
- Third-party hosting: The connector runs on
connector.get-ryze.ai(external server); any compromise could expose keys or read your accounts. - Supply chain: Dependency on Ryze's maintenance and security posture; 0 weekly npm downloads suggests limited production adoption for auditing.
How to stay secure
Security Hardening Guide: open-seo-mcp-skills
Least-privilege permissions
- Use a dedicated Google account for this MCP with only GSC (Google Search Console), GA4 (Google Analytics), and Ads access—not your main business account. If the account is compromised, damage stays limited.
- Review OAuth scopes before connecting: When prompted to grant access, verify Ryze only requests "read" permissions for reports and data, never "write" or "delete" for campaigns or budgets. Revoke access immediately if it asks for broader rights.
- Rotate credentials quarterly: Regenerate API keys and revoke old tokens in Google Cloud Console to limit exposure if they're accidentally leaked.
- Disable unused integrations: Only connect GSC, GA4, or Ads if you actually need them; don't authorize all three if you only track rankings.
Network isolation
- Run on a separate device or isolated network if handling competitor or highly sensitive client data; don't use the same machine for public web browsing or email.
- Use a VPN when accessing the MCP to encrypt traffic between your device and
connector.get-ryze.ai, especially on public WiFi. - Don't share the MCP connection URL or API credentials via email, Slack, or chat—treat them like passwords.
Monitoring
- Check Google's "Connected apps & sites" (myaccount.google.com/permissions) monthly; look for Ryze and revoke if you stop using it.
- Review Claude conversation history: Export or delete chats containing sensitive campaign data, bidding strategies, or competitor insights after use (Anthropic may retain training data otherwise).
- Set up alerts in Google Analytics and Ads for unusual activity—sudden budget changes, new keywords, or account access from unknown locations.
Red flags to watch for
- Unexpected permission requests: If the MCP suddenly asks to "write" to Ads or modify campaigns, stop and contact Ryze support immediately.
- Data slowness or failures: If the connector frequently times out or returns incomplete reports, don't assume it's safe—it could be under attack or leaking data silently.
- Ryze security incidents: Monitor Ryze's status page and GitHub for breaches; this tool has minimal community adoption (near-zero npm usage), so security patches may be slow.
- Claude leaking campaign details: If Claude-generated reports accidentally include exact bid amounts, budgets, or client names in public chats, treat it as a breach and rotate credentials.
Alternatives
- Semrush API (opens external site) – Commercial SEO platform with official API; higher barrier to entry but mature security model.
- Ahrefs API (opens external site) – Enterprise backlink and rank tracking; premium pricing, well-established trust.
- Google Search Console API (native) (opens external site) – Direct, first-party integration; requires custom implementation.
- SEO.com Agent Skills (opens external site) – Commercial platform with AI agent connectors; vendor-managed security.