What MCP Tools Are Available in OpenSEO for AI Agents: Complete Reference Guide

OpenSEO provides 15 MCP (Machine-Controlled Platform) tools that AI agents can invoke via JSON-RPC to automate SEO workflows including project management, keyword research, SERP analysis, backlink profiling, and technical site audits.

OpenSEO is an open-source SEO platform that exposes its functionality through a standardized MCP interface located in src/server/mcp/. These MCP tools in OpenSEO for AI agents allow automated systems to perform complex SEO operations programmatically, with each tool implemented as a TypeScript module under src/server/mcp/tools/ and validated through Zod schemas defined in src/server/mcp/output-schemas.ts.

MCP Tool Categories

The available MCP tools in OpenSEO span seven functional categories, enabling comprehensive SEO automation for AI agents.

Project Management Tools

Manage SEO projects and authentication context through three core tools located in src/server/mcp/tools/:

• create-project.ts – Creates a new SEO project with specified parameters.
• list-projects.ts – Returns all projects accessible to the authenticated user.
• whoami.ts – Retrieves identity information about the authenticated MCP caller using the auth context from src/server/mcp/context.ts.

Keyword Research and Management

Execute keyword research and persist findings using DataForSEO integration:

• research-keywords.ts – Executes keyword research queries via the DataForSEO API.
• save-keywords.ts – Persists keyword sets to the user's saved keyword store.
• list-saved-keywords.ts – Retrieves the user's saved keyword collections.

These tools handle the complete keyword lifecycle from discovery to storage, with input validation enforced by src/server/mcp/output-schemas.ts.

SERP and Rank Tracking

Monitor search engine results and ranking positions:

• get-serp-results.ts – Fetches SERP data for specific queries and geographic locations.
• get-rank-tracker.ts – Accesses rank tracking information for previously saved keywords.

Domain Intelligence

Analyze domain performance and generate content ideas:

• get-domain-overview.ts – Returns high-level metrics including traffic estimates, authority scores, and organic keyword counts.
• get-domain-keyword-suggestions.ts – Generates keyword ideas based on a domain's existing content profile.

Backlink Analysis

Evaluate link profiles and referring domain authority:

• get-backlinks-profile.ts – Delivers detailed backlink profiles with individual link data.
• get-backlinks-overview.ts – Summarizes aggregate metrics including total backlink counts and referring domains.

Site Audit and Search Console Integration

Perform technical SEO analysis and access Google Search Console data:

• site-audit-tools.ts – Executes comprehensive site audits and returns technical SEO findings.
• search-console-tools.ts – Interfaces with Google Search Console to fetch queries, clicks, and performance metrics.

DataForSEO Bridge

Low-level API integration support:

• dataforseo-research-tools.ts – Provides lower-level DataForSEO API helpers used by higher-level research tools.

Tool Architecture and Transport Layer

The MCP tools in OpenSEO rely on a robust transport and validation system. The src/server/mcp/transport.ts file handles incoming JSON-RPC requests to the /mcp endpoint, validates OAuth scopes (MCP_SCOPE), and routes calls to the appropriate tool implementations. Each tool receives an auth context from src/server/mcp/context.ts containing the openSeoAuth payload.

Output formatting is standardized through src/server/mcp/table.ts, which renders results as plain-text tables for text-only clients. Structured data follows the Zod schemas in src/server/mcp/output-schemas.ts, ensuring type-safe responses. Authentication is managed via src/server/mcp/oauth-provider.ts, securing access control for external AI agents.

Calling MCP Tools from AI Agents

AI agents interact with OpenSEO MCP tools through JSON-RPC requests. Each request must include the method name matching the tool implementation (using snake_case) and parameters validated against the Zod schemas.

Example: Domain Overview Request

To retrieve domain metrics, call get-domain-overview.ts using the get_domain_overview method:

{
  "jsonrpc": "2.0",
  "id": "1",
  "method": "get_domain_overview",
  "params": {
    "domain": "example.com",
    "language_code": "en"
  }
}

The MCP server validates the request against output-schemas.ts, executes the tool, and returns structured data:

{
  "jsonrpc": "2.0",
  "id": "1",
  "result": {
    "domain": "example.com",
    "traffic": 12500,
    "authority_score": 72,
    "backlinks": 3421,
    "organic_keywords": 87
  }
}

Example: Saving Keywords

To persist keywords to a project, invoke save-keywords.ts using the save_keywords method:

{
  "jsonrpc": "2.0",
  "id": "2",
  "method": "save_keywords",
  "params": {
    "project_id": "proj_123",
    "keywords": ["cloud hosting", "serverless SEO"]
  }
}

Response confirmation:

{
  "jsonrpc": "2.0",
  "id": "2",
  "result": {
    "status": "saved",
    "count": 2
  }
}

Example: Listing Saved Keywords

Retrieve stored keywords using list-saved-keywords.ts via the list_saved_keywords method:

{
  "jsonrpc": "2.0",
  "id": "3",
  "method": "list_saved_keywords",
  "params": { "project_id": "proj_123" }
}

The response includes timestamped keyword data:

{
  "jsonrpc": "2.0",
  "id": "3",
  "result": [
    { "keyword": "cloud hosting", "saved_at": "2024-09-01T12:34:00Z" },
    { "keyword": "serverless SEO", "saved_at": "2024-09-01T12:35:00Z" }
  ]
}

Summary

• OpenSEO exposes 15 MCP tools organized across seven functional categories: project management, keyword handling, SERP tracking, domain insights, backlink analysis, site audits, and DataForSEO integration.
• All tools are implemented in src/server/mcp/tools/ and accessed via the /mcp JSON-RPC endpoint handled by src/server/mcp/transport.ts.
• The architecture uses Zod schemas in src/server/mcp/output-schemas.ts for type-safe validation and OAuth scope enforcement via src/server/mcp/oauth-provider.ts.
• AI agents invoke tools using JSON-RPC with method names matching the TypeScript file implementations (e.g., get_domain_overview calls get-domain-overview.ts).

Frequently Asked Questions

What does MCP stand for in OpenSEO?

MCP stands for Machine-Controlled Platform. It is the protocol layer that allows AI agents to programmatically invoke SEO tools within the OpenSEO ecosystem through standardized JSON-RPC requests to the /mcp endpoint.

How do AI agents authenticate with OpenSEO MCP tools?

AI agents authenticate using OAuth tokens with the MCP_SCOPE scope. The src/server/mcp/oauth-provider.ts module handles token issuance, while src/server/mcp/context.ts provides the openSeoAuth context to each tool invocation, ensuring secure access to project data and SEO APIs.

Can AI agents save and retrieve keyword research data?

Yes. AI agents can use research-keywords.ts to discover keywords via DataForSEO, save-keywords.ts to persist them to a project, and list-saved-keywords.ts to retrieve previously stored keyword collections. All operations are project-scoped and validated through the Zod schemas in src/server/mcp/output-schemas.ts.

What is the difference between get-backlinks-profile.ts and get-backlinks-overview.ts?

get-backlinks-overview.ts provides aggregate metrics such as total backlink counts and referring domain statistics, while get-backlinks-profile.ts delivers granular, detailed information about individual backlinks. Agents typically use the overview for high-level competitive analysis and the profile for deep-dive link audits.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}