COG/RITHM

Error Message

COG/RITHM

Orchestration Portal v2.16

Or

SUBSCRIBE

An active subscription is required to access the platform.

New Platform Key

Copy this key now. You will not be able to see it again.

Revoke this key?

All systems using it will instantly lose access. This action cannot be undone.

Cog/rithm Help

Stateless by Design

Standard LLMs charge API costs per-token. Feeding an entire chat history back into every subsequent prompt causes costs to grow exponentially. To enforce maximum efficiency, Cog/rithm defaults to a stateless architecture—each orchestration is processed independently.

If you require historical context to be analyzed, either manually paste previous text into your query, or enable Maintain Context in the Configuration panel to automatically append your several immediate prior exchanges.

Orchestration Tiers

Standard: A high-density, single-shot synthesis. Fast and efficient.

Ultimate: A full Cognitive 'rithm. Reasoning, introspection then a synthesized, structurally sound directive.

Configuration & Access

Keys: To enable Cog/rithm AI optimization, select your login name (lower left-side menu panel) to open the Settings menu and enter your unique Platform Key and BYOK Provider Key. Click 'Done'. BYOK LLM keys are stored with a Zero-Knowledge Architecture locally on the device and are never written to or stored on our servers.

Session Preview

COG/RITHM

You're In. Let's Wire Up Your Engine.

Generate a key below, then connect it to Claude, ChatGPT, or Gemini in under a minute.

Authentication

Generate a platform key to authenticate your MCP connection to Claude, ChatGPT, Gemini, Cursor, or VS Code.

MCP Integration

Connect Haakon's MCP tools to the AI assistant you already use. Your platform key from above goes directly in the connection URL below — no separate provider key required.

Claude (claude.ai, Claude Desktop, Claude Code)

Settings → Connectors → Add custom connector, then paste this URL:

https://api.haakon.ai/mcp/haakon?key=YOUR_HAAKON_KEY

ChatGPT (also works in Codex — same connector list)

In the app: Plugins → MCPs → Add server, then paste this URL. No separate token or headers needed — your key travels in the URL itself.

https://api.haakon.ai/mcp/haakon?key=YOUR_HAAKON_KEY

Using ChatGPT in a browser instead of the app? Custom connectors there require a Plus, Pro, Business, or Enterprise plan with Developer Mode enabled first (Settings → Connectors → Advanced, or Settings → Security and login) before you'll see the same Add connector option.

Gemini

Custom MCP connectors are currently part of Gemini Spark (personal Google accounts, 18+, US only, as this feature continues rolling out). Go to Settings & help → Connected Apps, and enter the URL below — no separate credentials needed, since your key travels in the URL itself:

https://api.haakon.ai/mcp/haakon?key=YOUR_HAAKON_KEY

Other IDEs & Agent Frameworks (Cursor, VS Code, etc.)

These connect straight to the same remote server — no local install needed.

Cursor — .cursor/mcp.json

{
  "mcpServers": {
    "haakon": {
      "url": "https://api.haakon.ai/mcp/haakon?key=YOUR_HAAKON_KEY"
    }
  }
}

VS Code — .vscode/mcp.json (note the root key is servers, not mcpServers)

{
  "servers": {
    "haakon": {
      "type": "http",
      "url": "https://api.haakon.ai/mcp/haakon?key=YOUR_HAAKON_KEY"
    }
  }
}

Available Tools: haakon, haakon_help

Cognitive Lenses

Pass these configuration flags into your IDE requests or MCP tool payloads to forcefully alter the LLM's reasoning framework.