FREE SKILL: Install the Roast Room and get seven specialist subagents debating any business decision you throw at them. Each scores your idea from their own angle. You get a weighted verdict and a final recommendation in one report. Setup takes under five minutes.
Most business decisions get evaluated by one person. A business partner, a coach, a mentor. One perspective and one blind spot.
The Roast Room runs your idea through seven in parallel. A Skeptic looks for what can go wrong. An Optimist looks for where the real upside is. A Financial Analyst looks at the unit economics. A Legal Observer checks the regulatory exposure. A Marketer checks the positioning. An Operations Specialist looks at the bottlenecks. A Growth Strategist looks at the acquisition and retention loops.
They each write a position and a score. Then they debate each other. Then you get the synthesis.
One report. Seven mental models.
The Problem Your Current Process Has
You are probably making important decisions with incomplete feedback.
Not because you lack smart advisors. Because getting seven people in a room to debate your pricing model, your hiring decision, or your go-to-market strategy takes weeks to arrange and most operators do not have that network on speed dial.
The result is decisions made from one angle, checked against one person's experience, with whatever blind spots that person carries. The Roast Room changes that in a single session.
The Kickoff Command
Copy the prompt below and paste it after installing the skill. Replace the bracketed sections with your idea.
Copy: Basic Kickoff
Copy this.
Use the roast room skill. Roast this idea: [YOUR IDEA OR DECISION IN ONE TO THREE SENTENCES]
For sharper output, use the detailed version:
Copy: Full Kickoff Template
Copy this.
Use the roast room skill. Roast this idea: Idea: [ONE SENTENCE DESCRIPTION OF YOUR PLAN] Target audience: [WHO THIS IS FOR AND WHY THEY WOULD PAY] Revenue model: [HOW MONEY IS MADE — PRICING, TERMS, UNIT ECONOMICS IF KNOWN] What is driving this decision: [THE SPECIFIC PRESSURE OR OPPORTUNITY MAKING YOU CONSIDER IT NOW]
Examples that get strong output:
"A monthly retainer package for small e-commerce brands at $2,500/month. Audience: Shopify store owners doing $50k to $300k/year. Revenue model: 6-month minimum retainer, billed upfront. Driving this: three inbound leads asking for this exact scope in the last 30 days."
"Hiring a part-time operations coordinator at $2,000/month. Audience: internal hire. Revenue model: cost centre, expected to reclaim 8 hours per week of founder time. Driving this: onboarding bottleneck is delaying client delivery by an average of 12 days."
One sentence is fine. More detail produces sharper debate.
How to Use It
- Download the skill folder from the link below.
- In Claude Code, move the unpacked folder into
~/.claude/skills/roast-room/on your computer. If the skills folder does not exist, create it. - Open a new Claude Code session and run the kickoff command with your idea.
If you use Claude on the web or the desktop app rather than Claude Code: open Settings, then Capabilities, and upload the zip as a skill. Then start a new chat and run the kickoff command.
That is the entire install.
What You Get Back
The skill runs the following workflow after you submit your idea:
01. Position papers from all seven advisors. Each one writes from their specific role. The Skeptic leads with what is being glossed over. The Financial Analyst surfaces cash flow risk. The Marketer checks whether the positioning would actually land with the audience you named. All seven write before any debate.
02. One round of debate. Each advisor reads the strongest opposing point from another and responds to it. This is where the report becomes useful. It is not seven independent opinions. It is seven opinions that have been tested against each other.
03. Weighted average score. A single number from 0 to 100 calculated from all seven advisor scores. The number alone is not the point. The distribution matters more: a 72 average with a 20 from the Financial Analyst and a 95 from the Optimist tells you something different from a 72 average where all seven landed between 65 and 80.
04. Three risks and three opportunities. Surfaced from the full debate, not just the loudest voice.
05. Final recommendation. A synthesis of the weighted vote. What the panel decided, what they want you to watch, and what would change the outcome.
If a PDF renderer is available in your environment, the skill formats this as a PDF. If not, it returns structured text you can copy into a document.
Save It and Run It Again
The skill stays in place once installed. You can run the same idea through multiple times with different framings to test how the debate shifts when you change the pricing, the audience, or the timeline.
That is one of the more useful applications. Take a decision you have already made and describe it as you initially framed it, then describe it as you currently understand it. The difference in the debate tells you what changed in your thinking.
What This Is Not
Not a replacement for real advisors with skin in the game. The advisors here have no personal stake in your outcome. They do not know your specific market, your specific relationships, or the specific people involved in your decision. Their value is framework coverage and pressure testing. Not judgment about your specific situation.
Not a substitute for financial modeling. The Financial Analyst will flag unit economics risks but it works from whatever numbers you provide. If you provide no numbers, the analysis is conceptual. If you provide wrong numbers, the analysis is confidently wrong.
Output depends on input specificity. Vague framing produces vague debate. "A new product idea" gives the skill almost nothing to work with. "A $49/month SaaS tool for solo consultants that automates their invoicing" gives it everything it needs.
PDF rendering requires a PDF tool in your Claude Code environment. If you see plain text output instead of a PDF, the skill is working correctly. Copy the text into Google Docs or Notion.
If You Only Do One Thing This Week
Take the next decision you are sitting on. The pricing change, the hire, the pivot. Write it as a four-sentence brief: the idea, the audience, the revenue model, and what's driving the decision. Run it through the Roast Room.
You do not have to act on what it says. Run it once to see what seven simultaneous frameworks look like on one decision.
Download the skill and the setup guide below.