Atomic AGI

What Atomic AGI MCP server does

Atomic AGI connects compatible AI agents with Atomic, an SEO and AI search optimization platform from Omnius. The platform covers strategy, AI visibility tracking, brand reputation, content workflows, entity strategy, technical audits, and Agent Experience (AX).

The repository packages the connection rather than implementing the Atomic application or MCP server. It includes public plugin metadata, a remote MCP connection definition, and a reusable atomic-keyword-research skill.

How it works

The integration uses Atomic’s remote MCP endpoint:

https://app.atomicagi.com/api/mcp

An MCP-compatible client connects to that endpoint and operates within the permissions of the Atomic account and projects being used. Clients with an OAuth connection flow can use that authentication method. Other clients can use a personal token created in Atomic’s MCP settings and pass it privately through the client configuration.

The included keyword research workflow is designed to inspect project context, review existing content and available search data, suggest topic clusters based on intent, separate measured evidence from hypotheses, avoid duplicate content, and verify authorized Grid changes.

Setup and configuration

An Atomic account and access to the relevant projects are required. Configure the remote endpoint in the chosen client, then complete OAuth or add a personal MCP token according to that client’s authentication model. Tokens should remain in local or private configuration and must not be committed to the repository.

For Cursor, the MCP configuration uses the endpoint and an authorization header:

{
  "mcpServers": {
    "atomic-agi": {
      "url": "https://app.atomicagi.com/api/mcp",
      "headers": {
        "Authorization": "Bearer atmcp_YOUR_TOKEN"
      }
    }
  }
}

The portable mcp.json describes the remote server; authentication is configured separately by the client. Plugin and skill installation behavior can vary between clients.

Tools and capabilities

• Connect AI agents to Atomic projects over MCP.
• Support SEO and AI search optimization workflows.
• Track AI visibility and manage brand reputation workflows.
• Work with content, entity strategy, technical audits, and AX reviews.
• Research keyword opportunities using project context, existing content, and available search data.
• Propose intent-based topic clusters while checking for duplicate content and distinguishing evidence from hypotheses.

Limitations and notes

The package is an integration layer, not the Atomic application or the MCP server implementation. Access depends on the Atomic account, project permissions, and the capabilities of the selected client. Keep credentials out of source control, and verify authorization before making Grid changes through a workflow.