brainstorming

Brainstorming Ideas Into Designs

Help turn ideas into fully formed designs and specs through natural collaborative dialogue.

Start by understanding the current project context, then ask questions one at a time to refine the idea. Once you understand what you're building, present the design and get user approval.

<HARD-GATE> Do NOT invoke any implementation skill, write any code, scaffold any project, or take any implementation action until you have presented a design and the user has approved it. This applies to EVERY project regardless of perceived simplicity. </HARD-GATE>

Production-Grade Doctrine applies with full force here — trade-offs are first chosen in brainstorming: you MUST NOT simplify a design by quietly cutting a required behavior; surface every material trade-off and let the user decide. Never weaken, bypass, or remove a security control — a security regression is never acceptable.

Anti-Pattern: "This Is Too Simple To Need A Design"

Every project goes through this process. A todo list, a single-function utility, a config change — all of them. "Simple" projects are where unexamined assumptions cause the most wasted work. The design can be short (a few sentences for truly simple projects), but you MUST present it and get approval.

Checklist

You MUST create the session bead + step children up front via bd import: first bd create "Brainstorming: <topic>" -t task (the permanent session bead — the audit trail; note its id), then author each checklist step as JSONL (one issue per line, "issue_type":"chore", an explicit "priority", id omitted, and a parent-child dep to the session id) and pipe to bd import -. Read bd import --help on first use; confirm the output shows no Skipped dependency. Cleanup: close each step bead as you complete it; if you abandon the brainstorm, close or bd purge the open steps (import children are permanent, not self-cleaning). Then complete the steps in order:

1. Explore project context — check files, docs, recent commits, and query the KB for prior decisions/research on the topic
2. Offer the visual companion just-in-time — NOT upfront. The first time a question would genuinely be clearer shown than described, offer it then (its own message); on approval its browser tab opens for you. If no visual question ever arises, never offer it. See the Visual Companion section below.
3. Ask clarifying questions — one at a time, understand purpose/constraints/success criteria
4. Propose 2-3 approaches — with trade-offs and your recommendation
5. Present design — in sections scaled to their complexity, get user approval after each section
6. Write design doc — save to .internal/specs/YYYY-MM-DD-<topic>-design.md and commit
7. Spec self-review — quick inline check for placeholders, contradictions, ambiguity, scope (see below)
8. User reviews written spec — ask user to review the spec file before proceeding
9. Spec-review gate offers stress-test — the spec-review gate (step 8) includes an "Approved + stress-test" option, offered every time; if selected, invoke stress-test before writing-plans
10. Transition to implementation — invoke writing-plans skill to create implementation plan

Process Flow

digraph brainstorming {
    "Explore project context" [shape=box];
    "Ask clarifying questions" [shape=box];
    "Propose 2-3 approaches" [shape=box];
    "Present design sections" [shape=box];
    "User approves design?" [shape=diamond];
    "Write design doc" [shape=box];
    "Spec self-review\n(fix inline)" [shape=box];
    "User reviews spec?" [shape=diamond];
    "Stress-test selected\nat gate?" [shape=diamond];
    "Invoke stress-test skill" [shape=box];
    "Invoke writing-plans skill" [shape=doublecircle];

    "Explore project context" -> "Ask clarifying questions";
    "Ask clarifying questions" -> "Propose 2-3 approaches";
    "Propose 2-3 approaches" -> "Present design sections";
    "Present design sections" -> "User approves design?";
    "User approves design?" -> "Present design sections" [label="no, revise"];
    "User approves design?" -> "Write design doc" [label="yes"];
    "Write design doc" -> "Spec self-review\n(fix inline)";
    "Spec self-review\n(fix inline)" -> "User reviews spec?";
    "User reviews spec?" -> "Write design doc" [label="changes requested"];
    "User reviews spec?" -> "Stress-test selected\nat gate?" [label="approved"];
    "Stress-test selected\nat gate?" -> "Invoke stress-test skill" [label="selected"];
    "Stress-test selected\nat gate?" -> "Invoke writing-plans skill" [label="skipped"];
    "Invoke stress-test skill" -> "Invoke writing-plans skill";
}

The terminal state is writing-plans. The only other skill brainstorming may invoke is stress-test (optional, between spec approval and writing-plans). Do NOT invoke frontend-design, mcp-builder, or any other implementation skill.

The Process

Understanding the idea:

• Check out the current project state first (files, docs, recent commits)
• Before proposing any design, query the knowledge store: bd list --label <topic> --status all + bd search "<keywords>" --status all + bd memories <keyword> (the memory half — lessons, patterns, root-causes; knowledge-beads alone miss it entirely) (body terms: add --desc-contains "<term>" — search matches titles only; >10 hits: narrow the query, never triage truncated titles). Then read — hits are pointers, not knowledge: bd show <id1> <id2> ... every hit that plausibly bears on this design (full bodies, never truncated below 50 lines; distilled <=2.5KB by design — sanctioned spend). 0 relevant does not mean none exist — re-angle the query once before emitting KB check: none. Emit KB check: N bead hits, M memory hits, K read (or KB check: none) plus a one-line disposition per read hit — folded in (what it changed) or ruled out (why). The check is complete when every plausibly-relevant hit is dispositioned. If a prior decision already covers this, surface it rather than re-litigating.
• Before asking detailed questions, assess scope: if the request describes multiple independent subsystems (e.g., "build a platform with chat, file storage, billing, and analytics"), flag this immediately. Don't spend questions refining details of a project that needs to be decomposed first.
• If the project is too large for a single spec, help the user decompose into sub-projects: what are the independent pieces, how do they relate, what order should they be built? Then brainstorm the first sub-project through the normal design flow. Each sub-project gets its own spec → plan → implementation cycle.
• For appropriately-scoped projects, ask questions one at a time to refine the idea
• Prefer multiple choice questions when possible — use your structured question tool for these (structured options are faster to answer than reading text and typing a response). Open-ended questions that don't have clear discrete options can remain as text.
• Only one question per message - if a topic needs more exploration, break it into multiple questions
• Focus on understanding: purpose, constraints, success criteria

Exploring approaches:

• Propose 2-3 different approaches with trade-offs
• Use your structured question tool to present the approaches as structured options. Put your recommended option first with "(Recommended)" in the label. Use the description field for trade-offs and reasoning. This is more efficient than text blocks that require the user to read and type a response.
• If approaches need detailed explanation beyond what fits in option descriptions, present the analysis as text first, THEN follow up with a structured question for the actual selection
• YAGNI ruthlessly - Remove unnecessary speculative features from all designs (never drop a required behavior, edge case, or security control — that is descoping, not YAGNI)

Presenting the design:

• Once you believe