Bmad Idea

BMAD Idea Router

Use this skill as the repository's pre-planning creative intake and concept-routing layer.

The job is not to dump dozens of ideation techniques or persona gimmicks. The job is to: 1. normalize a messy early-stage idea packet, 2. choose the single framing mode that reduces ambiguity fastest, 3. produce one concept artifact that downstream work can actually use, 4. route the next step to the right neighboring skill.

Read references/operating-modes.md for mode-selection heuristics, references/handoff-boundaries.md for adjacent-skill routing, and references/concept-packet-template.md for reusable output scaffolding.

When to use this skill

• The user has a rough idea and needs a clearer problem, audience, concept, or story before planning starts
• The request is still ambiguous enough that jumping straight to a PRD, sprint plan, or launch plan would be premature
• A product, service, content, or GTM idea needs one concept brief, positioning brief, or narrative packet first
• A game idea needs pillars, player fantasy, core loop, and first concept framing before bmad-gds or task decomposition
• The user explicitly mentions brainstorming, concept shaping, design thinking, positioning, JTBD-style framing, story packaging, or pre-planning clarification

When not to use this skill

• The main need is phase-based delivery routing or deciding the next BMAD artifact after project framing already exists → use bmad
• The main need is backlog slicing, story decomposition, estimation, or sprint preparation → use task-planning
• The main need is broad marketing execution, launch planning, funnel work, or KPI-aware campaign routing → use marketing-automation
• The main need is game-production coordination for an existing GDD, playtest packet, build issue, or milestone → use bmad-gds
• The main need is only rewriting copy or polishing prose with no framing decision to make

Operating modes

Choose exactly one primary mode per run.

1. Problem framing

• Use when the idea is fuzzy and the main gap is what problem actually matters
• Best for founders, PMs, designers, and consultants starting from raw opportunity language

2. Audience and value framing

• Use when the offer exists but the target user, JTBD, pains, gains, or differentiation are weak
• Best for product marketing, GTM, and concept-validation work

3. Concept shaping

• Use when the direction is promising but still needs a tighter concept, constraints, options, and core promise
• Best for product, feature, service, and offer-level idea shaping

4. Game concept framing

• Use when the packet is about game fantasy, pillars, audience, core loop, progression, or tone before production planning
• Best for solo developers and small game teams before bmad-gds

5. Story packaging

• Use when the main need is a narrative, pitch, concept story, or internal alignment packet
• Best when the problem is understanding and buy-in rather than raw ideation volume

Instructions

Step 1: Normalize the idea intake

Capture the packet in this format before choosing a mode:

idea_packet:
  domain: product | gtm | content | consulting | game | mixed | unknown
  stage: raw-idea | rough-concept | candidate-offer | concept-with-notes | narrative-draft | unknown
  audience:
    primary_user: "who this is for"
    pains_or_needs:
      - pain 1
      - pain 2
  problem_or_goal: "what change or outcome is this idea trying to create?"
  current_material:
    - notes
    - brainstorm bullets
    - positioning draft
    - customer research
    - game concept notes
    - moodboard / references
    - pitch / story draft
  strongest_unknown: problem | audience | value | concept-scope | game-pillars | narrative | unknown
  downstream_target: bmad | task-planning | marketing-automation | bmad-gds | deck | unknown
  constraint: time | scope | evidence | differentiation | alignment | unknown

If the packet is incomplete, continue with explicit assumptions instead of stalling.

Step 2: Choose one primary mode

Select the mode that reduces ambiguity fastest.

Do not mix multiple modes unless the user explicitly asks for a two-pass workflow.

Step 3: Produce one concept artifact

Return exactly one primary artifact:

• problem brief
• positioning brief
• concept brief
• game concept packet
• story / pitch packet

The artifact should be concise, reusable, and ready to hand to a downstream skill.

Step 4: Make route-outs explicit

When the concept artifact is done, recommend the next downstream lane if appropriate:

• bmad → when the next job is phase routing and selecting PRD / tech spec / architecture / sprint-plan depth
• task-planning → when the concept is approved and needs backlog slices, milestones, or execution packets
• marketing-automation → when the concept is sound and now needs launch, messaging operations, campaign structure, or KPI-aware marketing follow-through
• bmad-gds → when the game concept is strong enough to turn into milestones, GDD slices, playtest planning, or production coordination

Always state why that route-out is next and what packet should be passed forward.

Step 5: Use this output structure

# Idea Routing Brief

## Scope
- Domain: ...
- Stage: ...
- Strongest unknown: ...
- Downstream target: ...
- Confidence: high | medium | low

## Chosen mode
- problem-framing | audience-and-value-framing | concept-shaping | game-concept-framing | story-packaging

## What matters most now
- 2-4 bullets grounded in the packet

## Recommended artifact
- problem brief | positioning brief | concept brief | game concept packet | story / pitch packet

## Artifact contents
| Section | Decision | Why it matters now |
|---------|----------|--------------------|
| ... | ... | ... |

## Immediate next steps
1. ...
2. ...
3. ...

## Route-outs
- Skill / workflow: ...
- Why next: ...
- What to pass forward: ...

## What not to do yet
- 1-3 bullets preventing premature planning, launch detail, or execution drift

Step 6: Keep the artifact lean and decision-oriented

Good outputs are short enough to reuse.

Guardrails:

• prefer a few strong decisions over a giant brainstorm transcript
• convert ideas into constraints, assumptions, and next artifacts
• preserve uncertainty explicitly instead of pretending the concept is resolved
• stop before roadmap depth, sprint decomposition, or campaign execution unless the user asks for the handoff skill next

Compatibility commands and legacy naming

Older BMAD-CIS usage may mention these commands or personas:

• bmad-cis-brainstorming
• bmad-cis-design-thinking
• bmad-cis-innovation-strategy
• bmad-cis-problem-solving
• bmad-cis-storytelling
• Carson / Maya / Victor / Dr. Quinn / Sophia

Treat them as mode hints, not separate first-class workflows. Map them into the modern modes above:

• brainstorming / design thinking / problem solving → usually problem framing or concept shaping
• innovation strategy → usually audience and value framing or concept shaping
• storytelling → usually story packaging

Output format

Always return a short operator-style idea routing brief.

Required qualities:

• one chosen mode, not a kitchen sink of creativity tactics
• one reusable artifact, not a pile of disconnected ideas
• explicit route-outs to neighboring skills when the next lane is clear
• visible assumptions and unresolved questions
• keep the result under roughly 400-700 words unless the user asks for a larger concept packet

Examples

Example 1: product concept before planning

Input

Use bmad-idea. We think there is an opportunity around AI-assisted team handoffs, but we do not know the sharp problem or who it is really for.

Output sketch

• Chosen mode: problem framing
• Recommended artifact: problem brief
• Route-out after approval: bmad

Example 2: GTM / positioning ambiguity

Input

We built a workflow tool, but our messaging is fuzzy and every pitch sounds generic.

Output sketch

• Chosen mode: audience and value framing
• Recommended artifact: positioning brief
• Route-out after approval: marketing-automation

Example 3: game concept shaping

Input

Help us shape a cozy automation game idea before we start a real GDD.

Output sketch

• Chosen mode: game concept framing
• Recommended artifact: game concept packet
• Route-out after approval: bmad-gds

Best practices

1. Act like a concept router, not a prompt zoo — choose one framing mode that fits the ambiguity. 2. Prefer one artifact that travels well — the point is to make downstream planning easier. 3. Use neighboring skills on purpose — bmad, task-planning, marketing-automation, and bmad-gds each own later-stage work. 4. Keep uncertainty visible — a concept brief should surface assumptions, not bury them. 5. Bridge domains without flattening them — product, GTM, and game ideas share structure, but their outputs differ. 6. Treat legacy commands as compatibility language — the modern source of truth is the routing model.

References

• references/operating-modes.md
• references/handoff-boundaries.md
• references/concept-packet-template.md
• ../bmad/SKILL.md
• ../task-planning/SKILL.md
• ../marketing-automation/SKILL.md
• ../bmad-gds/SKILL.md

bmad-idea Reference

bmad-idea is now the repository's pre-planning creative router. Its source of truth is SKILL.md; this file exists as a quick compatibility reference for older BMAD-CIS naming.

Modern source-of-truth modes

Legacy command / persona compatibility

Older BMAD-CIS references may mention the commands or personas below. Treat them as mode hints, not as separate first-class workflows:

Downstream handoffs

Reference links

• references/operating-modes.md
• references/handoff-boundaries.md
• references/concept-packet-template.md

Concept Packet Template

Use this as the default skeleton after choosing a mode.

# Idea Routing Brief

## Scope
- Domain:
- Stage:
- Strongest unknown:
- Downstream target:
- Confidence:

## Chosen mode
- problem-framing | audience-and-value-framing | concept-shaping | game-concept-framing | story-packaging

## What matters most now
- Bullet 1
- Bullet 2
- Bullet 3

## Recommended artifact
- problem brief | positioning brief | concept brief | game concept packet | story / pitch packet

## Artifact contents
| Section | Decision | Why it matters now |
|---------|----------|--------------------|
| Problem / audience / fantasy | ... | ... |
| Value / hook / transformation | ... | ... |
| Constraints / scope / evidence | ... | ... |
| Open questions | ... | ... |

## Immediate next steps
1. ...
2. ...
3. ...

## Route-outs
- Skill / workflow:
- Why next:
- What to pass forward:

## What not to do yet
- ...
- ...

Mode-specific emphasis

• Problem framing → bias toward pain, stakes, and assumptions
• Audience and value framing → bias toward user, JTBD, differentiation, and proof
• Concept shaping → bias toward the shortest crisp summary plus scope constraints
• Game concept framing → bias toward fantasy, pillars, loop, and first-playable proof
• Story packaging → bias toward narrative tension, promise, proof, and ask

bmad-idea Handoff Boundaries

bmad-idea owns pre-planning framing. It should stop once the concept packet is clear enough for a later-stage skill to do its job.

Neighboring skills

bmad

Route to bmad when the idea is framed enough that the next question is:

• what BMAD phase are we in,
• what artifact comes next,
• whether we need a product brief, PRD, tech spec, architecture, or sprint plan.

What to pass forward:

• chosen concept / positioning / problem brief
• main assumptions and constraints
• project type and risk notes

task-planning

Route to task-planning when the concept is approved and the next job is execution slicing.

What to pass forward:

• concept summary
• scope boundaries
• acceptance questions
• milestone or first-version constraint

marketing-automation

Route to marketing-automation when the concept is clear and the next job is launch, growth, onboarding, funnel, or messaging execution.

What to pass forward:

• audience and value framing
• story / pitch packet
• proof points and unresolved message risks
• intended KPI or success condition

bmad-gds

Route to bmad-gds when a game concept is strong enough to turn into production coordination.

What to pass forward:

• player fantasy
• pillars
• core loop
• target platform / audience
• first milestone or first-playable questions

Routing rules

• If the user still says "we are not sure what the problem is" → stay in bmad-idea
• If the user says "we know what this is, now what do we write next?" → bmad
• If the user says "break this into work" → task-planning
• If the user says "take this to market" → marketing-automation
• If the user says "turn this game concept into a production plan" → bmad-gds

Boundary anti-patterns

• Do not let bmad-idea become a giant prompt catalog again.
• Do not absorb roadmap, sprint, or campaign execution detail into the ideation step.
• Do not treat generic copy polishing as concept routing.
• Do not send a route-out without describing what packet should be handed forward.

bmad-idea Operating Modes

Use bmad-idea to choose the smallest framing move that meaningfully reduces ambiguity.

Mode selection heuristics

1. Problem framing

Choose this when the user keeps describing features, solutions, or ambition but cannot yet state:

• what problem matters,
• for whom it matters,
• what evidence or constraint is shaping the choice.

Good prompts:

• "Help me sharpen this idea before we plan it"
• "I have a bunch of notes but not a real problem statement"
• "What should we clarify before writing a PRD?"

Primary artifact: problem brief

Recommended sections:

• target user or situation
• current pain / friction
• desired outcome
• why this matters now
• assumptions to test next

2. Audience and value framing

Choose this when the concept exists but the market/user/value angle is weak.

Good prompts:

• "Who is this really for?"
• "Our positioning is generic"
• "What job are we helping with?"
• "How do we explain why this is better?"

Primary artifact: positioning brief

Recommended sections:

• primary audience
• JTBD / pains / gains
• differentiated promise
• proof points or missing proof
• message risks / confusion traps

3. Concept shaping

Choose this when the user has multiple partial ideas and needs one tighter direction.

Good prompts:

• "Brainstorm and narrow this"
• "Turn these notes into one concept"
• "What should the first version actually be?"

Primary artifact: concept brief

Recommended sections:

• concept summary
• problem / user / value
• must-have constraints
• scope boundaries
• open questions

4. Game concept framing

Choose this when the packet is still pre-production and centered on player fantasy or core loop.

Good prompts:

• "Shape this game idea"
• "What are the pillars for this concept?"
• "How do we define the hook and loop before a GDD?"

Primary artifact: game concept packet

Recommended sections:

• player fantasy
• target audience / platform
• pillars
• core loop
• progression / tone / differentiator
• what the first playable should prove

5. Story packaging

Choose this when the concept exists but people do not yet understand it or support it.

Good prompts:

• "Turn this into a pitch"
• "Make the story clearer"
• "We need a concept narrative for stakeholders"

Primary artifact: story / pitch packet

Recommended sections:

• audience
• current problem or tension
• desired transformation
• promise
• proof / evidence / unresolved risk
• call to action or next decision

Anti-patterns

• Do not mix all five modes in one response.
• Do not produce a giant brainstorm transcript as the main artifact.
• Do not jump into roadmap, sprint, or campaign execution detail when the current job is still framing.
• Do not use story packaging when the real gap is still basic problem or audience clarity.