Mode d’emploi
- Copiez la commande et examinez le paquet avant de l’installer dans votre environnement agent.
- Consultez le document original pour les outils, identifiants et environnements requis.
- Décrivez les entrées et le livrable souhaité ; comparez la sortie aux sources.
npx skills add https://github.com/aj-geddes/claude-code-bmad-skills --skill bmad-brainstormAvant de commencer
- L’installation n’inclut ni abonnement au modèle, ni quota API, ni accès aux fichiers.
- Étoiles et installations reflètent l’activité, pas un test indépendant de qualité ou de sécurité.
Document Skill original
Rendu depuis le dépôt source en conservant titres, exemples, code, tableaux, liens et images.
BMAD Brainstorm
Structured ideation harness for the BMAD planning lifecycle. Produces a brainstorming-report.md of organized ideas and insights ready to feed downstream planning skills (product-brief, PRD, architecture).
Three Intents
Ask the user which intent applies — or infer from context:
| Intent | When to use |
|---|---|
| Create | New brainstorming session on a topic or problem |
| Update | Add techniques, deepen coverage, or incorporate new constraints |
| Validate | Confirm generated insights align with project goals or decision-log |
Technique Menu
| Technique | Best For | Approx Time |
|---|---|---|
| 5 Whys | Root cause analysis | 10-15 min |
| SCAMPER | Feature ideation & creative variations | 20-30 min |
| Mind Mapping | Idea organization & hierarchy | 15-20 min |
| Reverse Brainstorming | Risk and failure-mode identification | 15-20 min |
| Six Thinking Hats | Multi-perspective evaluation | 30-45 min |
| Starbursting | Question exploration (Who/What/When/Where/Why/How) | 20-30 min |
| SWOT Analysis | Strategic positioning | 30-45 min |
| Brainwriting | Silent parallel idea generation | 15-20 min |
Technique Selection Guide
- Exploring a problem? → 5 Whys, then Starbursting
- Generating features or solutions? → SCAMPER, Mind Mapping
- Assessing risk or failure modes? → Reverse Brainstorming, Six Thinking Hats (Black Hat)
- Strategic planning or positioning? → SWOT, Six Thinking Hats (full cycle)
- Need parallel idea volume? → Brainwriting
Session Workflow
Use TodoWrite to track each step.
- Gather context — Ask: topic/problem, relevant constraints, which BMAD track is active
- Read project context — Load
bmad-output/project-context.mdandbmad-output/decision-log.mdif present - Select techniques — Choose 2-4 complementary techniques; confirm with user
- Execute sessions — Apply each technique systematically (diverge first, do not filter)
- Organize ideas — Categorize all generated ideas; label Impact (H/M/L) and Feasibility (H/M/L)
- Extract insights — Identify top 3-5 actionable insights across all techniques
- Identify risks — Capture risks surfaced during ideation
- Produce report — Fill
templates/brainstorm-session.template.md; save to output folder - Update decision-log — Record key decisions or directions confirmed during the session
- Recommend next steps — Which BMAD skill to engage next
Script Utilities
Generate SCAMPER prompts for a specific topic:
bash
bash "${CLAUDE_PLUGIN_ROOT}/skills/bmad-brainstorm/scripts/scamper-prompts.sh" "your topic"Generate a SWOT analysis scaffold:
bash
bash "${CLAUDE_PLUGIN_ROOT}/skills/bmad-brainstorm/scripts/swot-template.sh" "your subject" > bmad-output/swot-draft.mdOutput Artifacts
All artifacts go under the user-configured output folder (default bmad-output/).
| File | Description |
|---|---|
bmad-output/brainstorming-report.md | Primary output — organized ideas + insights |
bmad-output/decision-log.md | Append key decisions made during session |
bmad-output/swot-draft.md | Optional SWOT scaffold (from script) |
Use templates/brainstorm-session.template.md for the report structure.
Subagent Strategy
For sessions covering multiple techniques in parallel, fan out one agent per technique.
Pattern: Fan-Out Ideation Agents: 2-6 parallel (one per selected technique)
| Agent | Task | Output |
|---|---|---|
| Agent 1 | Apply SCAMPER to generate feature variations | bmad-output/brainstorm-scamper.md |
| Agent 2 | Build Mind Map to organize ideas hierarchically | bmad-output/brainstorm-mindmap.md |
| Agent 3 | Reverse Brainstorming — identify failure modes | bmad-output/brainstorm-risks.md |
| Agent 4 | Six Thinking Hats — multi-perspective analysis | bmad-output/brainstorm-hats.md |
| Agent 5 | Starbursting — Who/What/When/Where/Why/How | bmad-output/brainstorm-questions.md |
| Agent 6 | SWOT Analysis — strategic positioning | bmad-output/brainstorm-swot.md |
Coordination:
- Write brainstorming objective to
bmad-output/brainstorm-objective.md - Select 2-6 techniques; launch one agent per technique with context path
- Each agent generates 10-20 ideas/insights using its assigned technique
- Main context synthesizes all outputs into
bmad-output/brainstorming-report.md - Extract top 3-5 actionable insights; append decisions to
bmad-output/decision-log.md
Example subagent prompt:
Task: Apply SCAMPER technique
Context: Read bmad-output/brainstorm-objective.md for topic and constraints
Output: Write 15-20 creative variations to bmad-output/brainstorm-scamper.md
Apply each SCAMPER letter systematically. Generate 2-4 ideas per letter.
For each idea: one-sentence description, potential value, innovation level
(incremental / breakthrough). Conclude with your top 3 most promising ideas.
Do not implement anything — ideation output only.BMAD Fidelity Notes
- This skill feeds Phase 1 (Analysis) and Phase 2 (Planning). Insights flow into
the product-brief and PRD skills.
- Insights that become commitments must be logged in
bmad-output/decision-log.md. - This skill does not produce stories. It produces raw insight material that
upstream skills (product-manager, system-architect) refine into epics and stories.
- Scale-adaptive tracks apply: Quick Flow sessions may use 1-2 techniques; Enterprise
sessions should apply 4-6 for comprehensive coverage.
Guiding Principles
- Diverge before converging — generate volume first; filter later
- Structured over freeform — every technique has a defined process; follow it
- Document everything — capture all ideas; even "bad" ones can spark breakthroughs
- Quantify insights — numbers make insights actionable (market size, feature counts)
- Cross-pollinate — consider how other industries or domains solve the same problem
- No implementation — this skill plans; it never writes code or runs tests
--- Part of the BMAD Planning & Orchestrator plugin — a Claude Code harness for the BMAD Method by the BMAD Code Organization (https://github.com/bmad-code-org/BMAD-METHOD). Implements the spirit of bmad-brainstorming. All methodology credit belongs to the BMAD Code Organization.
