Pressure-test a SaaS idea with a free Claude Code skill that scores evidence and returns a dated go or no-go verdict

Solo founders tend to treat their own excitement, compliments and waitlist sign-ups as proof of demand, then spend months building something nobody pays for. They need a repeatable way to name the biggest gap in their evidence and test it cheaply before building.

1. Install the Solo Founder Skills plugin in Claude Code from its marketplace and call the validate skill (or let it load when you ask whether an idea is worth building). 2. The skill picks a mode: find a problem (no idea yet), full validation, pressure-test only, design a test, or verdict on results you already have. 3. With no idea yet, it asks five one-line questions about groups you belong to and recurring complaints, drops problems only you have or that happen rarely, and keeps one where you can name a real person who has it. 4. Pressure-test: it asks three stage-specific questions, records your answers word for word and marks each as evidenced, partial or none. Any gap blocks a verdict. 5. Design one test (conversations, landing page, fake door, pre-sale or manual delivery) and fix the metric, pass bar and deadline before running it. If no stranger has shown interest yet, the first test is conversations, using the bundled interview guide. 6. After running the test yourself, score six dimensions from 1 to 5 (frequency, intensity, willingness to pay, market size, unique advantage, current solutions) and run the included Python script, which adds them up and applies two override rules: a 1 on willingness to pay caps the verdict at one more experiment, and a high advantage score cannot rescue a low total. Waitlist sign-ups count as interest, never as willingness to pay. 7. It writes a dated brief (verdict, evidence table, next test, revisit date) and points to the pricing and first-customers skills to sell the first ten by hand before writing code.

Not reported. The repository describes the method and ships the prompts, reference files and scoring script, but gives no example of a founder's idea it validated or rejected and no customer or revenue numbers. It had about 259 stars and 44 forks on GitHub when read.

No results, case studies or user feedback are reported, so it is unknown whether the verdicts are well calibrated. The founder still has to run the interviews or tests; the skill only structures and scores them. Scores are self-assessed by the founder, so the script can only be as honest as its inputs. The market-size method is a rough bottom-up estimate. Requires Claude Code. The publish date of the skill is not stated (the repository was created in January 2026).

Claude Code with plugin support; Python 3 to run the scoring script; willingness to talk to real potential customers.

In Claude Code: /plugin marketplace add whawkinsiv/solo-founder-skills, then /plugin install solo-founder@solo-founder-skills-marketplace, then call /solo-founder:validate.

• GitHub (Will Hawkins, whawkinsiv) ↗ Code repository, publication date unknown

Source published: unknown. Last reviewed here: October 22, 2026. Spot a mistake? Tell us.