Kairova Mission Control

Every button above is a skill — a small instruction file (SKILL.md) in the AI brain that tells Claude exactly what to do, which specialist agents to use, and where the result goes. Adding one is easy.

The easy way

Just tell Claude in plain English what you want, e.g. "Add a command /mn-weekly-report — every Friday, pull the week's posted content and engagement, summarise wins and misses, and post it to the dashboard and a draft." That one sentence is enough — Claude writes the skill, wires it to this dashboard, and it appears here automatically.

What makes a command work well

Include these when you describe it (Claude fills gaps, but the more you give, the sharper it is):

1. Trigger — what you'd say / when you'd use it.
2. Outcome — the finished thing it should produce.
3. Steps — the sequence, if you have one in mind.
4. Agents / data — which specialists or files it should use.
5. Inputs — what it needs from you each time (a topic, a date range).
6. Guardrails — anything it must never do; where approval is required.
7. Output — where the result goes (this dashboard, a draft, a file).

How it works underneath

• The skill file lives in the brain (kairova-brain\mission-control\skills\) and syncs to Channi.
• Clicking the button here queues the command; the runner on the machine executes it via Claude and writes the result back to this board.
• Approval-gated actions (publishing, sending, spending) always stop for Craig — a command can draft, never send on its own.

Current commands are the ten buttons above. Ask Claude any time to add, rename, or remove one.