agentleFS
maccydeemaccydee/claude-planning-skills/brainstorming/SKILL.md

brainstorming

Skill3 starsChanged 3 months agoMITView on GitHub

Collaborative, one-question-at-a-time design brainstorming that walks through a system or feature design section by section, always leading with a recommendation, and produces a DESIGN.md at the end. Use when the solution is unclear and needs design refinement first, when asked to "brainstorm", "help me design", "work through a design", "think through the architecture", or before writing an implementation plan. Precedes writing-plans.

190 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use chinese-code-review, chinese-commit-conventions and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • Leodorareluctant259Skill

    chinese-code-review

    中文代码审查规范——在保持专业严谨的同时,用符合国内团队文化的方式给出有效反馈

    Also in 8 projects that use this

  • Leodorareluctant259Skill

    chinese-commit-conventions

    中文 Git 提交规范 — 适配国内团队的 commit message 规范和 changelog 自动化

    Also in 8 projects that use this

  • ryanportfolioSkill

    addskill

    Create, import, update, or install repo or personal skills: runtime ownership, resources, discovery checks.

    Also in 6 projects that use this

  • ryanportfolioSkill

    apply-firmware

    Bring current folder (plain folder, existing repo, old or partial firmware copy) up to full Harness Firmware in place. Mirror external repo → adopt-repo.

    Also in 6 projects that use this

  • +2 more from Leodorareluctant259
  • +6 more from ryanportfolio

The file

We split this skill into the rules it gives an agent

8 rules. 0 are common, 5 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 170 edited versions we read keep these.

  • 5

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    One question at a time: Don't overwhelm. Ask, wait, incorporate the answer, then ask the next.

    line 28
  2. Distinctive:

    Explore 2-3 alternatives: before settling on any choice, and name the trade-off explicitly.

    line 42
  3. Distinctive:

    Be flexible: If something doesn't make sense, go back and clarify.

    line 49
  4. Distinctive:

    Ask a single validating question.

    line 60
  5. Distinctive:

    Wait for the answer. Incorporate it.

    line 61
  6. Very specific:

    If --output-dir provided: write to {OUTPUT_DIR}/DESIGN.md

    line 67
  7. Very specific:

    Otherwise: write to docs/plans/YYYY-MM-DD-DESIGN-<topic>.md

    line 68
  8. Very specific:

    After the design is validated and DESIGN.md is saved, offer the next step: optionally produce a PLAN.md via writing-plans, which turns the …

    line 77

“Common” is measured: kept by at least 6 in 10 of the 170 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Brainstorming

Overview

Refine a design collaboratively before any code or plan exists. Walk through the design one section at a time, always leading with a concrete recommendation, and validate each section with the partner before moving on. The goal is a design the partner fully understands and can defend, not a design handed to them fully formed.

Announce at start: "I'm using the brainstorming skill to work through this design with you."

Core Rules

These are the heart of the skill, follow them on every turn:

  1. Every question MUST include a recommendation. Lead with "I'd lean toward X because Y. Does that match your thinking?", never an open-ended "what do you think?". The partner is validating or redirecting a proposal, not staring at a blank page. This is what keeps momentum and teaches the reasoning as you go.
  2. One question at a time. Don't overwhelm. Ask, wait, incorporate the answer, then ask the next. This means one question per turn, not one decision question with a second "and also…" tacked on the end. If you need context before you can recommend, that context question IS the turn: ask it alone and wait. Resist the pull to gather background and validate a decision in the same message; splitting them is what keeps the partner in control and each answer unambiguous.
  3. Multiple choice preferred where it fits. Offer A / B / C options, each with its trade-off named, so the choice is concrete and the partner can weigh it.
  4. Present the design in small sections (200-300 words each). Validate after each section before drafting the next. Small increments keep the partner in control and surface disagreement early, while it's cheap.
  5. Explore 2-3 alternatives before settling on any choice, and name the trade-off explicitly. A recommendation with no alternative is an assertion; a recommendation with the runner-up and the reason is a decision the partner can own.
  6. YAGNI ruthlessly. Cut features not required for the actual goal. If it isn't needed for the stated objective, it doesn't go in the design.
  7. Be flexible. If something doesn't make sense, go back and clarify. The section-by-section structure is a guide, not a cage, revisit earlier decisions when a later one exposes a problem.

Working Pattern

For each section of the design:

  1. State the recommendation up front, with a one-line reason.
  2. Give 2-3 alternatives with their trade-offs (multiple-choice where possible).
  3. Explain the chosen tech in enough depth that the partner can defend it live, not just what, but why, and how it compares to the obvious textbook alternative.
  4. Ask a single validating question.
  5. Wait for the answer. Incorporate it. Move to the next section only once validated.

Deliverable

At the end of the session, produce a DESIGN.md:

  • If --output-dir provided: write to {OUTPUT_DIR}/DESIGN.md
  • Otherwise: write to docs/plans/YYYY-MM-DD-DESIGN-<topic>.md

The DESIGN.md is the artifact the partner keeps (and may convert to PDF via Cmd-P for sharing). It should capture, per section: the decision made, the alternatives considered, and the trade-off that justified the choice, so the reasoning survives, not just the conclusion.

Handoff

After the design is validated and DESIGN.md is saved, offer the next step: optionally produce a PLAN.md via writing-plans, which turns the design into a task-by-task implementation plan.

Related Skills

SkillWhen to Use
writing-plansAfter the design is settled, turn DESIGN.md into an implementation plan
executing-plansAfter a plan exists, implement it in batches
skepticTo pressure-test the design's necessity and scope before committing

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.