Image: GitHub
Why it mattersA founder gets back a sized market, a price band buyers accept and a break-even number in one afternoon, so the decision to build or kill a product arrives ahead of the first real engineering spend.
A new free skill pack lets you pitch 100 simulated buyers before you spend a day writing code. Jake Schincariol published founder-skill on GitHub as a set of eleven Claude skills under the MIT licence, and the repository reached 108 stars and 15 forks within one day of its first commit.
The eleven skills share one shape. Each is a Markdown file with a brief and a numbered procedure. Installing a skill is copying a folder, which keeps the pack usable in Claude Code and in the standard Claude chat without any Python package setup.
The eleven skills a founder runs in order
Board scores an idea with three sub-agents that read like $100M Offers by Alex Hormozi, Zero to One by Peter Thiel and the Steve Jobs biography by Walter Isaacson. Competitors maps direct and indirect rivals with pricing and customer complaints, each with a source. Consumer runs a panel of 100 buyer agents with different incomes and objections and reports how many would buy at a given price. Pricing turns the panel answers into an acceptable range using Van Westendorp's price-sensitivity analysis from 1976, which asks four questions about what feels too cheap, too expensive and worth paying.
Offer builds a stack of bonuses, guarantees and urgency against the objections the panel raised, then re-tests. CFO calculates contribution per sale, margins, break-even and the cash a founder needs through year one. Marketing writes positioning and a 30-day campaign with ten hooks and a channel budget. Brand produces names with trademark and domain checks. Ops maps suppliers, staffing, permits and a risk register and feeds the costs back to CFO. Launch writes a cheap real-world test, a dated countdown and a 30-day plan. Plan compiles the business case and prints a one-word verdict: Profitable, Not Yet, or Incomplete.
Honest about what the panel is
The repository README treats the buyer-panel results as an upper bound on how many real buyers would say yes, and the pricing answers as a list of questions to take to real buyers next. The brief says in plain words that these outputs are a guide, never a measurement. The caveat matters, because treating a hundred synthetic buyers as a market size is where tools like this fall over. The skill ships with the warning written into its own brief.
A founder gets back a sized market, a price band and a break-even number in one afternoon that would otherwise take a week of calls and a spreadsheet. The build-or-kill decision then arrives with numbers behind it rather than a hunch.
Source
Jakeschincariol/founder-skill on GitHub, by Jake Schincariol at opusjake.ai.
This item was written by an AI system from the linked source. Reveneau is responsible for what it publishes.
Get AI News in your inbox
New developer tools, model and agent releases, and how teams are actually using them to release software. Short, and only when there is something worth reading.