Brainstorming Ideas Into Designs
When to Use
Use this skill before:
- building a new feature
- changing behavior in a meaningful way
- creating a new component or workflow
- turning a rough idea into an implementation-ready design
Usage
Use a short design discussion when consequential choices remain open.
If the user has already supplied a workable design and asked for implementation,
record the essential assumptions and proceed with the authorized work.
For small, reversible changes, a concise inline design is enough.
inspect current context
-> identify decisions that affect the outcome
-> resolve consequential gaps
-> state the chosen approach and acceptance checks
-> implement when authorized
Decision Boundaries
Ask for input when the answer changes the product goal, user experience,
public contract, data retention, or an irreversible action.
Use existing project patterns for routine implementation choices.
Continue independent research and preparation while a necessary answer is pending.
An explicit brainstorming-only request remains a discussion task.
Practical Design Checklist
- Read the relevant code, requirements, and previous decisions.
- State the problem, intended user, and observable success criteria.
- Identify only the uncertainties that materially affect the solution.
- Compare alternatives when there is a real tradeoff; do not manufacture options.
- Choose a small, coherent implementation and explain its main tradeoff.
- Record an enduring design document only when future work needs it.
- Continue into implementation if the user requested execution.
Example
Request: Add CSV export to the existing table.
Known: Current filters and columns define the exported data.
Decision: Reuse the existing authorized data-fetch path.
Check: Export filtered rows, escape CSV cells, preserve column order.
Clarify only if export scope or access requirements remain ambiguous.
Handoff
Use a written plan when dependencies or scope need one.
Do not require a second plan or approval merely to invoke another skill.
Keep original user constraints and acceptance checks with the handoff.
The Process
Understanding the idea:
- Check out the current project state first (files, docs, recent commits)
- Ask questions one at a time to refine the idea
- Prefer multiple choice questions when possible, but open-ended is fine too
- Ask concise questions only for unresolved material choices; group closely related gaps when useful
- Focus on understanding: purpose, constraints, success criteria
Exploring approaches:
- Propose 2-3 different approaches with trade-offs
- Present options conversationally with your recommendation and reasoning
- Lead with your recommended option and explain why
Presenting the design:
- Once you believe you understand what you're building, present the design
- Scale each section to its complexity: a few sentences if straightforward, up to 200-300 words if nuanced
- Invite corrections without blocking already authorized implementation
- Cover: architecture, components, data flow, error handling, testing
- Be ready to go back and clarify if something doesn't make sense
After the Design
Documentation:
- For substantial designs, save the durable decisions to
docs/plans/YYYY-MM-DD-<topic>-design.md
- Use elements-of-style:writing-clearly-and-concisely skill if available
- Include the design with the task changes when a commit is requested
Implementation:
- Continue the requested implementation once consequential decisions are resolved
- Use writing-plans when the task needs an enduring plan; avoid forced skill chains
Key Principles
- One question at a time - Don't overwhelm with multiple questions
- Multiple choice preferred - Easier to answer than open-ended when possible
- YAGNI - Remove unnecessary features from designs
- Explore alternatives - Compare alternatives when they expose a real tradeoff
- Incremental validation - Confirm unresolved consequential choices; reuse existing authorization
- Be flexible - Go back and clarify when something doesn't make sense
1---2name: brainstorming3description: Use before creative product or engineering work when the user wants to design a feature, component, workflow, behavior change, or other solution whose goals, constraints, and tradeoffs need exploration before implementation.4---5
6# Brainstorming Ideas Into Designs
7
8## When to Use
9
10Use this skill before:
11
12- building a new feature
13- changing behavior in a meaningful way
14- creating a new component or workflow
15- turning a rough idea into an implementation-ready design
16
17## Usage
18
19Use a short design discussion when consequential choices remain open.
20If the user has already supplied a workable design and asked for implementation,
21record the essential assumptions and proceed with the authorized work.
22For small, reversible changes, a concise inline design is enough.
23
24```text
25inspect current context
26-> identify decisions that affect the outcome
27-> resolve consequential gaps
28-> state the chosen approach and acceptance checks
29-> implement when authorized
30```
31
32## Decision Boundaries
33
34Ask for input when the answer changes the product goal, user experience,
35public contract, data retention, or an irreversible action.
36Use existing project patterns for routine implementation choices.
37Continue independent research and preparation while a necessary answer is pending.
38An explicit brainstorming-only request remains a discussion task.
39
40## Practical Design Checklist
41
421. Read the relevant code, requirements, and previous decisions.
432. State the problem, intended user, and observable success criteria.
443. Identify only the uncertainties that materially affect the solution.
454. Compare alternatives when there is a real tradeoff; do not manufacture options.
465. Choose a small, coherent implementation and explain its main tradeoff.
476. Record an enduring design document only when future work needs it.
487. Continue into implementation if the user requested execution.
49
50## Example
51
52```text
53Request: Add CSV export to the existing table.
54Known: Current filters and columns define the exported data.
55Decision: Reuse the existing authorized data-fetch path.
56Check: Export filtered rows, escape CSV cells, preserve column order.
57Clarify only if export scope or access requirements remain ambiguous.
58```
59
60## Handoff
61
62Use a written plan when dependencies or scope need one.
63Do not require a second plan or approval merely to invoke another skill.
64Keep original user constraints and acceptance checks with the handoff.
65
66## The Process
67
68**Understanding the idea:**
69- Check out the current project state first (files, docs, recent commits)
70- Ask questions one at a time to refine the idea
71- Prefer multiple choice questions when possible, but open-ended is fine too
72- Ask concise questions only for unresolved material choices; group closely related gaps when useful
73- Focus on understanding: purpose, constraints, success criteria
74
75**Exploring approaches:**
76- Propose 2-3 different approaches with trade-offs
77- Present options conversationally with your recommendation and reasoning
78- Lead with your recommended option and explain why
79
80**Presenting the design:**
81- Once you believe you understand what you're building, present the design
82- Scale each section to its complexity: a few sentences if straightforward, up to 200-300 words if nuanced
83- Invite corrections without blocking already authorized implementation
84- Cover: architecture, components, data flow, error handling, testing
85- Be ready to go back and clarify if something doesn't make sense
86
87## After the Design
88
89**Documentation:**
90- For substantial designs, save the durable decisions to `docs/plans/YYYY-MM-DD-<topic>-design.md`
91- Use elements-of-style:writing-clearly-and-concisely skill if available
92- Include the design with the task changes when a commit is requested
93
94**Implementation:**
95- Continue the requested implementation once consequential decisions are resolved
96- Use writing-plans when the task needs an enduring plan; avoid forced skill chains
97
98## Key Principles
99
100- **One question at a time** - Don't overwhelm with multiple questions
101- **Multiple choice preferred** - Easier to answer than open-ended when possible
102- **YAGNI** - Remove unnecessary features from designs
103- **Explore alternatives** - Compare alternatives when they expose a real tradeoff
104- **Incremental validation** - Confirm unresolved consequential choices; reuse existing authorization
105- **Be flexible** - Go back and clarify when something doesn't make sense
seaworld008