Find And Evaluate Startup Ideas

1. Clarify whether the task is idea generation, idea evaluation, or a pivot search.
2. Start from founder-specific edges, not abstract brainstorming: expertise, work history, communities, weird access, recent problems, or things the team wishes existed.
3. Generate or restate the candidate idea in concrete terms: who has the problem, how acute it is, why now, and why this team.
4. Pressure-test each idea against the core filters: founder-market fit, market size or growth, acute pain, competition and new insight, personal conviction, why-now timing, distribution, and ease of getting the first version in users' hands.
5. Flag classic failure modes early: solution-in-search-of-problem, tar pit ideas, vague societal problems, and ideas the team is poorly positioned to execute.
6. Prefer ideas that feel hard, boring, or already competitive if they point to real pain and are left on the table by others.
7. If the team needs more ideas, walk through the source-backed recipes: founder expertise, problems personally seen, things they wish existed, recent changes in the world, successful-company variants, targeted interviews in a fertile space, or broken large industries.
8. Return a prioritized shortlist or evaluation memo plus the next interviews, experiments, or launches needed to resolve uncertainty.

Output

• A concise YC-style startup idea memo with:
• Idea summary
• Founder-market-fit assessment
• Problem quality and why now
• Key risks or tar-pit warnings
• Recommended next steps or interview plan
• If multiple ideas are involved, return a ranked shortlist with brief reasoning.

Quality Bar

• Prefer operational guidance over abstract advice.
• Keep the response grounded in the source thesis.
• Ask follow-up questions if key inputs are missing.
• Do not repeat transcript wording unless a short quote is necessary.
• Avoid fake precision. If evidence is weak, say what needs to be tested.
• Bias toward real problems and founder-specific advantage over cleverness.

References

• Read references/source-context.md for thesis, triggers, and non-goals.
• Read references/transcript.md for source evidence and timestamps.
• Read references/patterns.md for concrete anti-patterns, fixes, and implementation details.