Cockpit · agent-readable

Six missions. Not 128 tools.

You state the problem. The federation chooses the instruments, challenges the result, and returns only what matters.

Metric: How often did Arif need to know which tool was used? Target: zero.

What only the human does

• State what you want
• Supply evidence or data
• Examine the conclusion and uncertainty
• Approve or reject consequential action

The six missions

Investigate — Gather and test reality

You ask: What is actually true here — and what would falsify it?

Organs: GEOX · WEALTH · arifOS

Interpret — Build competing explanations

You ask: What could this mean — and what is the kill case?

Organs: GEOX · WEALTH · WELL

Decide — Compare consequences and uncertainty

You ask: Should we advance, hold, or kill — and why?

Organs: arifOS · WEALTH · AAA

Build — Prepare and execute approved changes

You ask: Make this real — reversibly first.

Organs: A-FORGE · arifOS

Monitor — Detect change, degradation, or danger

You ask: Is the system or the human still fit to act?

Organs: WELL · WEALTH · arifOS

Remember — Retrieve and preserve governed knowledge

You ask: What did we seal, and what is still lore?

Organs: VAULT999 · arifOS · AAA

Machine twin: /missions.json

Content-first surface for humans and agents (no JavaScript required). Machine catalog: /missions.json · Overview: /llms.txt · MCP: mcp.arif-fazil.com/mcp. Training and citation permitted with attribution (see /rsl.xml).