Fills the Template, Doesn't Design the Deal. — The LBO-Model Skill (Anthropic Skills)
The lbo-model skill has Claude fill in an LBO template's formulas, validate the calculations against each other, and check formatting against a professional standard, working on whatever template it's given rather than designing the deal itself.
When Claude runs an Anthropic Skill, it does not design the task's output the way an analyst would. Running the lbo-model skill does not mean Claude is designing a leveraged buyout deal from scratch. It fills in a template's formulas and checks them, working within whatever structure the template provides.
What a skill is, mechanically
A skill is a folder Claude reads before it works. This one is lbo-model, and the SKILL.md file inside it is the full instruction set, written in plain language with no hidden logic. Claude reads the file, then acts on it; the file is the program. The actual pipeline lives in the file's Steps section, and Claude reads each step and executes it one at a time, in the order written. The process is linear and does not branch unless a specific step says to branch.
The three concrete things it does
Concretely, this means three actions. It fills in the template's formulas. It validates the calculations against each other, checking that the numbers are internally consistent. And it checks the formatting against a professional standard. Because the skill reads the template's own structure first, before doing any of this, it can operate on whatever LBO template you hand it, not just one fixed layout it was built around. That template independence is a direct consequence of the skill reading structure before acting on it.
Fills and validates, but doesn't design
Inside the lbo-model skill, Claude fills in a template's formulas and validates them against each other. It does not design the deal itself, meaning it does not decide the deal structure, the assumptions driving the model, or the strategic logic behind the leveraged buyout. That is why it works on whatever template you hand it, and never goes beyond what that template specifies. The skill is bounded entirely by the structure it is given; it does not introduce new modeling decisions on its own.
Why this distinction matters
Confusing "fills and validates a template" with "designs the deal" would mean trusting the skill for a kind of judgment it never performs. The value of the skill is in mechanical reliability: consistent formula-filling, cross-checked calculations, and formatting held to a standard, applied to any template structure it is given. The actual deal design, choosing the assumptions, the capital structure, the exit strategy, remains the analyst's job entirely.
Seeing the steps before they run
Before relying on the output, it helps to ask Claude to read the lbo-model skill and walk through exactly what it will do before it does it. That last clause matters: asking Claude to explain first, before running anything, is what actually reveals the steps the file wrote for it, rather than leaving the process opaque.
Key takeaways
- The lbo-model skill fills a template's formulas, validates calculations against each other, and checks formatting against a professional standard.
- It does not design the deal itself; deal structure and assumptions remain the analyst's responsibility.
- Because it reads the template's own structure first, it works on whatever LBO template it's given, not one fixed layout.
- The process is linear: read each step in the SKILL.md file's Steps section and execute it in order, with no branching unless specified.
- Asking Claude to explain what it will do before running the skill reveals the actual steps it will follow.
Who this is for
Financial analysts and anyone building leveraged buyout models with Claude who wants to understand the boundary between mechanical template execution and the deal design judgment that still requires a human analyst.
Chapters
Full transcript(auto-generated, with timestamps)
The naive framing: "does Claude design the model?"
[0:00]Someone assumed Claude designs an LBO model from scratch. It doesn't. It fills in a template's formulas and checks them. So when Claude runs the LBO model skill, does it fill the model? A skill
A skill is a folder
[0:11]Is a folder Claude reads before it works. This one is LBO model. The skill MD file inside it is the full instruction set. Plain language, no hidden logic. Claude reads it then acts. The file is the program. The actual
Steps, in order
[0:24]Pipeline lives in the file step section. Claude reads each step and executes it one at a time in the order written. Linear, it doesn't branch unless a step tells it to. Concretely, that means
Fills, validates, formats
[0:35]Three things. It fills in the templates formulas. It validates the calculations against each other. And it checks the formatting against a professional standard. And because it reads the template's own structure first, it can do that on whatever LBO template you hand it, not just one fixed layout. Inside the LBO model skill, Claude fills
Carry-out
[0:54]In a template's formulas and validates them. It doesn't design the deal itself. That's why it works on whatever template you hand it and never beyond what that template specifies. Your turn. Paste
Your turn
[1:05]This into Claude. I want to fill in an LBO model template for a leveraged buyout deal. Read the LBO model skill and walk me through what you will do before you do it. That last clause matters. Asking Claude to explain first before it runs is what actually shows you the steps the file wrote for it. Phil's the template doesn't design the
Outro
[1:22]Deal. Lay them in for bear.





