Back to Library
BrainstormClaudeEditor's Pick

Creative Problem Solver

Apply multiple creative thinking frameworks to solve complex problems from different angles.

Prompt Template

You are a creative-thinking facilitator trained in design thinking, lateral thinking, and TRIZ. Apply multiple frameworks to one problem.

Problem statement: {{problem}}
Context (constraints, history, what's been tried): {{context}}
Who is affected: {{stakeholders}}
What "solved" looks like: {{success_criteria}}
Resources available: {{resources}}

Apply each framework and give 3 solution candidates from each:

1. **First-principles** — strip the problem to physical/economic basics and rebuild.
2. **Inversion** — what if we tried to make the problem worse? What does that reveal?
3. **Analogous fields** — how does a different industry solve a structurally similar problem?
4. **Constraint removal** — what if budget / time / regulation were not constraints?
5. **TRIZ contradictions** — what's the underlying contradiction? Resolve it.

After all candidates, deliver:
- The 3 most promising ideas across all frameworks
- For each: one-sentence pitch, biggest risk, smallest test you can run this week
- One "non-obvious" idea worth a second look even if it sounds weird

Fill in Your Details

ChatGPTClaudeGeminiDeepSeek

Example Output

Problem: Our support team is drowning in repetitive password-reset tickets.
FIRST-PRINCIPLES: 1) Passwords exist because we can't yet identify users another way. 2) Reset tickets exist because users forget. → Solution: replace passwords with magic links by default.
INVERSION: How do we make password resets worse? Hide the reset link, demand more info. → Reveals: every step in our current reset flow could be cut. → Solution: 1-click email reset.
ANALOGOUS: How do banks handle this? Step-up auth with device-bound credentials. → Solution: passkeys.
[...continues for all 5 frameworks...]
TOP 3 IDEAS:
1. Ship magic-link login as default. Risk: email deliverability. Test: enable for 5% of new signups this week.
2. Add passkeys. Risk: enterprise SSO conflict. Test: design spike, 2 days.
3. AI deflection in the support widget. Risk: bad answers hurt trust. Test: read-only suggestions to the agent, not the user.
NON-OBVIOUS: Give users a printed recovery code at signup. Old-school, weirdly effective.

Tips

  • The framework that feels weirdest for your problem usually produces the best ideas.
  • "Smallest test to run this week" is the highest-leverage section. Don't skip it.
  • Re-run with different success_criteria to see how the answer changes.