Help us improve
Share bugs, ideas, or general feedback.
Assembles stakeholder perspectives to deliberate complex architectural, technology, or design decisions before brainstorming, surfacing tensions and convergences without forcing choices.
npx claudepluginhub repozy/superpowers-optimizedHow this skill is triggered — by the user, by Claude, or both
Slash command
/superpowers-optimized:deliberationThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Surface genuine tension in a decision before committing to a direction.
Evaluates decisions via stance rotation (neutral, advocate, critic perspectives), synthesizes confidence-rated recommendation with next steps. For architectural choices, tech options, build-vs-buy, tradeoffs.
Brainstorms complex architectural decisions by launching parallel agents from diverse perspectives (pragmatist, security, performance) and synthesizes unified optimal solutions.
Explores trade-offs in design and architecture decisions as a thinking partner, helping users understand options and make informed choices without recommending solutions.
Share bugs, ideas, or general feedback.
Surface genuine tension in a decision before committing to a direction.
Some decisions get worse when you're forced to choose too early. Use deliberation when:
Do not use for: decisions that are already well-framed with clear options. Those go directly to brainstorming. Deliberation is for when you're not yet ready for options.
State the decision being deliberated in one sentence. Be precise — a vague decision produces vague perspectives.
Example: "Should we migrate the authentication layer from JWT to session cookies?"
Choose perspectives that have real, distinct stakes in the decision. Each perspective represents a legitimate set of constraints or values — not a person, but a role with a coherent viewpoint.
Good perspectives for technical decisions:
Choose only perspectives with genuine stakes in this specific decision. Three well-chosen perspectives beat five generic ones.
For each perspective in turn:
Ground rules:
After all perspectives have spoken, identify:
Convergence — where all (or most) perspectives agree, despite coming from different values. These points are load-bearing: a decision that violates them will cause problems regardless of which option is chosen.
Live tension — where perspectives genuinely disagree and no option fully satisfies all of them. These are the real trade-offs. Do not paper over them — surface them explicitly.
Reframes — where the act of hearing all perspectives reveals that the original decision was mis-stated, and the real question is something else entirely.
## Deliberation: [Decision statement]
### Perspectives
**[Perspective name]**
Values: [What this perspective cares about most]
Concern: [The specific risk or constraint it sees]
What it loses going left: [Downside of option A]
What it loses going right: [Downside of option B]
[Repeat for each perspective]
### Convergence
[Points where perspectives agree regardless of their different values]
### Live Tension
[Where perspectives genuinely disagree — the real trade-offs that cannot be avoided]
### Reframe (if applicable)
[If the deliberation revealed that the original question was wrong, state the better question here]
### Next step
[One of: "proceed to brainstorming with this framing", "return to user — the decision needs more information before proceeding", or "the reframe changes the scope — revisit premise-check before continuing"]
brainstorming instead.