I asked Claude if my startup idea was good. It said "decent". The plugin I built with Claude Code found a competitor making $32 a month.
retained text
Same idea, two answers: an AI app that writes wedding speeches, at $19 each.
**Plain Claude:** "It's a decent small business." No sources. Suggested spending ~$200 on ads to test.
([full answer, unedited](https://github.com/RemiPelloux/kill-my-idea/blob/main/examples/plain-claude-wedding.md))
**kill-my-idea:** NO-GO, 48/100.
- 18 sources.
- 14 competitors priced $0–34; one live competitor made **$32.75 in 30 days**.
- Each paid customer loses money (LTV ~$18 vs CAC ~$35).
**What it is:** a free, MIT-licensed Claude Code plugin. Instead of asking Claude "is my idea good?", it makes Claude look for reasons the idea will fail, then:
- scores it /100 across 7 criteria and **21 sub-scores**, each with a source or marked as an estimate
- applies fixed rules (a free alternative with no edge caps the score; no distribution channel means no GO)
- shows **how the idea could be saved** (e.g. PIVOT 55 → 63 with a done-for-you setup)
- writes a consultant-style HTML report that prints to PDF
Live report: https://remipelloux.github.io/kill-my-idea/reports/notion-real-estate/
**How Claude helped:** I built it with Claude Code end to end:
- the skills (research method, scoring rules, writing standard);
- the JSON schemas;
- the HTML report template;
- the test runs on real ideas I was about to spend money on.
One of those was a coloring book I built *before* checking. A 5-minute search found the same book for free.
/plugin marketplace add RemiPelloux/kill-my-idea
/plugin install kill-my-idea@kill-my-idea
Repo: https://github.com/RemiPelloux/kill-my-idea
Drop an idea (or a "$10k/month" tweet) in the comments and I'll run it.document fetched 2026-11-10 00:22 · canonical https://reddit.com/r/ClaudeAI/comments/1xrwb6f/i_asked_claude_if_my_startup_idea_was_good_it