summon-ceo-control-room

Ranks company tasks by limiting-factor analysis and proposes one decision card per CEO wake.

Updated Jun 16, 2026
One-click install
npx skills add https://github.com/adamtpang/summon.company --skill summon-ceo-control-room-adamtpang
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: summon-ceo-control-room
Source: https://github.com/adamtpang/summon.company/tree/main/skills/summon-ceo-control-room
Command: npx skills add https://github.com/adamtpang/summon.company --skill summon-ceo-control-room-adamtpang

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Running a team of AI agents creates constant noise: blocked tasks, unowned work, and budget drift. This Skill gives a CEO agent a disciplined loop to find the company's single limiting factor, rank work ruthlessly, and present one clear decision to the human board instead of acting on its own. ## Core Features & Use Cases - Surface: Reads dashboard, issue, and roadmap data from the company API to build a limiting-factor chain down to the deepest actionable constraint. - Triage: Scores candidate tasks with the Summon Score formula and applies the Thiel test to isolate the single S-tier task, explicitly deferring everything else. - Route: Proposes one decision card per wake naming the owning agent, an honest ETA from run history, and a cost estimate, then files it as a comment for the board to approve. - Use Case: On a scheduled wake in 24/7 mode, the CEO agent detects that a landing page review is blocking launch, scores open tasks, and files a card asking the board whether the design agent should start the review now. ## Quick Start Ask the CEO agent to run its triage loop and tell you what matters most right now.

Frequently Asked Questions about summon-ceo-control-room

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I prioritize tasks across a team of AI agents?

Score each candidate task with the Summon Score formula: (money + time + importance) times urgency divided by effort times humans, all rated 0-5. Then apply the Thiel test to pick the single task you would do if only one thing could happen today.

What is a limiting-factor chain in company operations?

A limiting-factor chain traces revenue backward through its blockers until reaching an actionable root cause, such as no revenue leading to no launch leading to an unreviewed landing page. The deepest actionable link is the constraint to fix first.

Can the CEO agent assign tasks to other agents directly?

No. The CEO agent only files and proposes decision cards; the human board assigns work. When the board approves, the assignment itself wakes the agent, so the CEO never dispatches, wakes, or unpauses agents.

How are ETAs estimated for agent tasks?

ETAs come from run history via the company API, using the median duration of that agent's recent completed runs of similar shape. If no history exists, the card states there is no track record rather than inventing a number.

When does the triage loop run in manual versus 24/7 mode?

In manual mode the loop runs only when the board explicitly asks, such as saying triage or what matters. In 24/7 mode the loop runs once per timer wake, producing at most one decision card before ending the run.