| name | brainstorm |
| description | Explore ambiguous ideas and solution directions before evidence-backed research or implementation planning. Use for ideation, framing, assumption checks, option generation, product or architecture trade-offs, and deciding whether research or a plan is needed. |
| license | MIT |
Brainstorm
Use this skill to shape a fuzzy problem before committing to research, planning, or implementation.
Brainstorming is for option discovery and judgment. It is not source-backed research, implementation planning, or coding.
Role Boundary
- Use
brainstorm when the problem is unclear, the user wants ideas, or the solution space needs exploration.
- Use
research when the answer depends on current facts, external sources, standards, benchmarks, vendor behavior, or source-backed recommendations.
- Use
plan after a direction is selected and the implementation is Standard or Complex.
- Use
scout, debug, or code-review when the answer depends primarily on existing code behavior.
- If brainstorming reveals factual uncertainty, hand off to
research instead of presenting guesses as evidence.
Workflow
- Frame the problem, desired outcome, constraints, and non-goals.
- Ask only high-value clarifying questions. If reasonable assumptions are safe, state them and continue.
- Surface hidden constraints, failure modes, and over-engineering risks.
- Present 2-3 viable options unless only one practical option exists.
- Recommend a default direction and explain why it fits the user's context.
- Identify when the next step should be
research, plan, implementation, or no action.
Evaluation Lens
Compare options by:
- User value and problem fit.
- Simplicity and reversibility.
- Implementation effort.
- Operational burden.
- Data, security, and compliance risk.
- Long-term maintainability.
- Dependencies and lock-in.
Rules
- Be direct about weak ideas, hidden costs, and likely failure modes.
- Mark assumptions clearly.
- Do not claim external facts without evidence; route to
research when evidence matters.
- Do not write code, scaffold files, or change the repository while brainstorming.
- Do not create a long report unless the user asks for one.
Output
Problem: <one sentence>
Assumptions:
- ...
Options:
| Option | Best when | Trade-offs | Risk |
|---|---|---|---|
Recommendation: <recommended option and why>
Next step: <research | plan | implement | no action>
Open questions:
- ...
