From s-kit
Use before creative or behavior-changing work. Clarifies intent, explores options, presents a design for approval, then hands off to plan-feature.
How this skill is triggered — by the user, by Claude, or both
Slash command
/s-kit:brainstormingThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Use this skill to turn a rough idea into an approved design before implementation starts.
Use this skill to turn a rough idea into an approved design before implementation starts.
Do not invoke implementation skills, write code, scaffold projects, or make behavior changes until you have presented a design and the user has approved it.Complete these in order:
grill-me as an optional stress-test before writing the approved design. This is optional; do not block normal approval if the user declines.docs/design/YYYY-MM-DD-{feature-name}/design.md using the structure in references/design-template.md.plan-feature with that exact design path so it can expand the design into requirements, manifest, execution log, orchestration, and task files under the matching docs/specs/YYYY-MM-DD-{feature-name}/ folder.build-feature.Start with the current project state. If the request is too broad for one spec, decompose it and brainstorm the first independently shippable piece.
If the request matches a small lane in using-s-kit (bug fix, refactor/docs, or hotfix), say so and route to that lane instead of starting a design. Brainstorming is for work that needs design decisions.
For an unfamiliar or large repository, offer to dispatch the s-kit-codebase-mapper agent first. It returns an evidence-backed map (architecture, conventions, relevant files, risks) that grounds the clarifying questions and design. This is optional; skip it when the project is already well understood.
Ask one question per message. Prefer multiple choice when it makes the decision easier. Focus on purpose, constraints, success criteria, dependencies, and what should be explicitly out of scope.
When you understand the work, propose approaches. Lead with the recommended approach and explain why. Keep the trade-offs practical: implementation effort, risk, future flexibility, and how well the approach fits the repo.
Present the design after the approach is selected. Scale the depth to the work. Cover architecture, components, data flow, error handling, verification, and rollout where relevant.
After approval, write design.md first, then invoke plan-feature. Offer grill-me as an optional stress-test before writing the approved design if the user wants deeper review. brainstorming owns the dated folder name. plan-feature must reuse that same folder name for the spec.
docs/design/YYYY-MM-DD-{feature-name}/
└── design.md
docs/specs/YYYY-MM-DD-{feature-name}/
├── README.md
├── spec.json
├── requirements.md
├── action-required.md
├── implementation-log.md
└── tasks/
├── task-01-{name}.md
└── task-02-{name}.md
design.md is the direct output of brainstorming. Follow references/design-template.md. It should capture the approved solution shape: context, selected approach, alternatives considered, architecture, major decisions, open risks, and verification strategy. grill-me can be offered before the file is written to stress-test the proposed design, but it is optional. plan-feature uses the approved design as the source material for requirements.md, README.md, spec.json, implementation-log.md, and task files. After the spec is reviewed, build-feature executes it wave by wave.
If visual questions are likely, offer this exact message by itself:
Some of what we're working on might be easier to explain if I can show it to you in a web browser. I can put together mockups, diagrams, comparisons, and other visuals as we go. This feature is still new and can be token-intensive. Want to try it? (Requires opening a local URL)
Wait for the user's response. If accepted, use visuals only for questions that are genuinely clearer when seen rather than read.
npx claudepluginhub j03fr0st/s-kit --plugin s-kitGuides creation, editing, and verification of skills for AI coding agents using test-driven development with subagent scenarios. Use when authoring or debugging skills.