Multimodal Hero Banner Prompt Audit Checklist
Audit and refine text-to-image prompt specifications for high-converting e-commerce promotional hero banners.
Use this template when preparing creative prompt matrices for flash sales and holiday banner campaigns. It ensures multimodal image prompts contain exact lighting, copy space, and composition controls before pipeline deployment.
Role: Senior Multimodal Creative Technologist specializing in algorithmic promotional asset generation.
Context
- Merchant Brand: {{brand_name}}
- Promotional Campaign: {{promotional_event}}
- Output Display Channels: {{target_channels}}
- Aesthetic Style System: {{visual_style_guide}}
- Primary Promotional Message: {{key_offer_element}}
- Target Image Model Engine: {{model_engine}}
Task
Generate a comprehensive pre-generation audit checklist to evaluate, sanitize, and optimize text-to-image prompt blueprints for the upcoming {{promotional_event}} banner series, ensuring visual consistency and maximum conversion readiness.
Method
- Analyze the core parameters of {{model_engine}} regarding token weighting, keyword syntax, and aspect ratio handling.
- Cross-reference {{visual_style_guide}} tokens with promotional urgency cues required for {{key_offer_element}}.
- Establish negative prompt safety nets to prevent brand-degrading artifacts, distorted limbs, and illegible pseudo-text.
- Define negative space verification checks across {{target_channels}} to safeguard layout room for dynamic typography overlays.
- Calibrate composition depth and lighting prompts to harmonize with {{brand_name}} color palettes.
- Formulate item-separation checks ensuring the hero product remains the unambiguous focal anchor.
- Package all criteria into an actionable gate-by-gate review checklist containing binary pass/fail verification steps.
Constraints
- MUST specify model-specific parameter flags (e.g., aspect ratio, seed, style weight) appropriate for {{model_engine}}.
- MUST NOT allow ambiguous aesthetic adjectives such as "photorealistic" or "hyper-quality" without technical lighting and lens qualifiers.
- Every checklist item must include a clear remediation instruction for failed items.
- Maximum checklist items: 15 targeted checks across three distinct verification phases.
Output format
1. Model Configuration & Syntax Gate (4-5 checks)
2. Composition & Negative Space Gate (4-5 checks)
3. Brand Consistency & Artifact Prevention Gate (4-5 checks)
4. Prompt Remediation Reference Guide (Markdown table with Flagged Issue, Root Cause, Remediation Snippet)
Self-review
- Ensure every checklist item contains a binary verification condition [ ] Pass / Fail.
- Confirm negative text-in-image protections are explicitly detailed for {{target_channels}}.
- Verify all variables ({{brand_name}}, {{promotional_event}}, {{model_engine}}, etc.) are actively utilized.
Explicit role, a named task, and discrete steps the model can follow.
Background, inputs and variables the model needs before it starts.
Hard boundaries — what the model must and must not do.
A named, field-level shape for the response.
Ordered work items that force analysis before an answer.
Length and structure that travel across frontier models.
Signal density — instruction weight without padding.
Documented variables so the scaffold adapts to new inputs.
Quality bar, assumptions and behaviour when inputs are thin.
How much real usage the template has behind it.