Agent Skills
Brainstorming
AIPOCH
Creative exploration and brainstorming; use it when you need to clarify goals/constraints, explore multiple options, and converge on a direction before implementation.
3
0
FILES
86100Total Score
View Evaluation ReportCore Capability
85 / 100
Functional Suitability
11 / 12
Reliability
9 / 12
Performance & Context
7 / 8
Agent Usability
14 / 16
Human Usability
8 / 8
Security
11 / 12
Maintainability
9 / 12
Agent-Specific
16 / 20
Medical Task
20 / 20 Passed
90You have a vague idea and need to clarify goals, scope, and constraints before committing to a plan
4/4
86You want to generate multiple solution directions for a problem and compare trade-offs
4/4
85Goal and boundary clarification (objectives, constraints, success criteria)
4/4
85Structured ideation to produce multiple distinct options
4/4
85End-to-end case for Goal and boundary clarification (objectives, constraints, success criteria)
4/4
SKILL.md
When to Use
- You have a vague idea and need to clarify goals, scope, and constraints before committing to a plan.
- You want to generate multiple solution directions for a problem and compare trade-offs.
- You need to break down a complex problem into smaller, actionable parts.
- You’re stuck and need alternative approaches, assumptions to challenge, or fresh angles.
- You want to converge from many ideas into a prioritized shortlist and define next steps.
Key Features
- Goal and boundary clarification (objectives, constraints, success criteria).
- Structured ideation to produce multiple distinct options.
- Problem decomposition into components, subproblems, and decision points.
- Evaluation and convergence (trade-offs, risks, feasibility, impact).
- Actionable next steps (experiments, prototypes, questions to validate).
- Reference workflow support via
references/brainstorming-workflow.md.
Dependencies
- None
Example Usage
Prompt
I want to build a feature that helps users learn faster in our app, but I’m not sure what exactly to build.
Constraints: 2-week implementation, no new backend services, must work on mobile.
Goal: increase weekly retention.
Please brainstorm options, compare them, and recommend a direction with next steps.
Expected Output Structure (Runnable Template)
## 1) Clarify Goals & Boundaries
- Primary goal: increase weekly retention
- Constraints: 2-week build, no new backend services, mobile-first
- Success metrics: retention uplift, feature adoption, session frequency
## 2) Generate Options (Diverge)
1. **Spaced repetition reminders**
- What it is: ...
- Why it helps: ...
2. **Daily micro-challenges**
- What it is: ...
- Why it helps: ...
3. **Progress-based streaks + milestones**
- What it is: ...
- Why it helps: ...
4. **Personalized review queue (client-side)**
- What it is: ...
- Why it helps: ...
## 3) Compare & Evaluate
| Option | Effort | Risk | Expected impact | Notes |
|---|---:|---:|---:|---|
| Spaced repetition reminders | Medium | Low | Medium | ... |
| Daily micro-challenges | Medium | Medium | High | ... |
| Streaks + milestones | Low | Low | Medium | ... |
| Client-side review queue | Medium | Medium | Medium | ... |
## 4) Converge on a Recommendation
- Recommended direction: **Daily micro-challenges**
- Rationale: ...
- Key assumptions to validate: ...
## 5) Next Steps
- Prototype: ...
- Instrumentation: ...
- A/B test plan: ...
- Open questions: ...
Implementation Details
-
Workflow: Follow the structured process described in
references/brainstorming-workflow.md:- Clarify goals, constraints, and success criteria.
- Diverge: generate multiple distinct options (avoid premature convergence).
- Decompose: break options into components and required decisions.
- Evaluate: compare trade-offs (impact, effort, risk, feasibility).
- Converge: select a direction and define concrete next steps.
-
Core parameters to capture
- Goal: what outcome to optimize (e.g., retention, cost, latency, satisfaction).
- Constraints: time, budget, platform, compliance, dependencies.
- Evaluation criteria: impact, effort, risk, reversibility, time-to-value.
- Assumptions: what must be true for the chosen direction to work.
- Next-step artifacts: shortlist, decision rationale, experiment plan, open questions.