| name | develop-creative-concept |
| description | Develop a vague content request into a reviewable creative strategy, big idea, treatment, script spine, visual system, and acceptance criteria before image or video production. Use when the user has a goal, product, theme, or audience but no concrete image, scene, campaign concept, narrative, or production-ready brief; also use to compare creative directions before building a workflow. |
Develop a creative concept
Do not create generation nodes until the concept is concrete enough to evaluate. Use $operate-content-workflow to inspect existing project context and save the accepted direction in the workflow objective.
Development loop
- Normalize the problem: objective, audience, desired response, publishing surface, deliverables, constraints, references, and proof of success.
- Find the strategic tension: current belief or behavior versus the change the content must create.
- Produce 2-4 genuinely different concept territories, each with one governing promise and one visual mechanism.
- Compare them on strategic fit, distinctiveness, visual executability, extensibility, continuity risk, and production cost.
- Develop the selected territory into a treatment, beat spine, visual system, sound idea, final image, and production assumptions.
- Hand image work to
$direct-image-creation, video work to $direct-video-creation, or a multi-asset system to $create-content-campaign.
Read references/creative-brief-and-strategy.md for any request. Read references/concept-generation-and-selection.md when proposing or comparing directions. Read references/treatment-script-and-beats.md before handing a concept into video or campaign production.
Output contract
Return the normalized brief, concept territories, recommendation with tradeoffs, selected treatment if a direction is chosen, non-negotiable visual anchors, unresolved production decisions, and explicit next-stage acceptance criteria. Keep a concept distinct from a prompt: prompts are downstream execution artifacts.
This method is an original Vetta adaptation informed by visual-skills by Serge Shima (CC BY 4.0, https://github.com/smixs/visual-skills) and Generative-Media-Skills (MIT).
