Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.kyberis.ai/llms.txt

Use this file to discover all available pages before exploring further.

​
Install the skill

Install the Kyberis skill so Cursor has the investigation workflow and response handling rules. 
npx -y @kyberis-ai/agent-skills install cursor

​
Connect MCP

npx -y @kyberis-ai/mcp connect cursor --token kct_abc123
The connector updates ~/.cursor/mcp.json by default. Use --dry-run to print Cursor-compatible MCP JSON without changing local config: 
{
  "mcpServers": {
    "kyberis": {
      "type": "http",
      "url": "<mcp_url>",
      "headers": {
        "Authorization": "Bearer <token>"
      }
    }
  }
}
Add the JSON to Cursor’s MCP configuration.

​
Verify

Ask Cursor: 
Use Kyberis to investigate whether this domain has related actors, campaigns, malware, or evidence. Resolve the domain first.

​
Working style

Keep Kyberis investigation prompts explicit:
  • Ask for resolution before assessment.
  • Ask for focused evidence calls for each claim.
  • Ask the agent to separate facts, inference, caveats, and next actions.

​
Troubleshooting

  • Use --json if you need clean machine-readable setup output.
  • Regenerate the connect token if setup says it was used, revoked, expired, or created for another client.
  • Check that Cursor sends the Authorization header with the MCP server request.