Ad-Ready: AI Advertising Image Generator

Generate professional advertising images from product URLs using a 4-phase AI pipeline on ComfyDeploy.

Source: github.com/PauldeLavallaz/ads_SV

---

Pipeline Architecture

The pipeline runs as a ComfyUI custom node deployed on ComfyDeploy. A single ProductToAds_Manual node executes 4 phases internally:

CODEBLOCK0

Phase 2: Campaign Brief (The Creative Brain)

The Brief Generator is the most critical intermediate step. It acts as a "Senior Art Director" that translates raw data into actionable creative direction using a 10-point framework:

1. 1. Strategic Objective — Why this campaign exists (awareness/positioning/launch)
2. Central Message — One idea perceivable without text
3. Visual Tone of Voice — Register: calm/energetic/intimate/monumental
4. Product Role — Hero vs co-protagonist vs implicit presence
5. Visual Language & Brand Coherence — Non-negotiable brand codes
6. Photographer & Equipment — Photography as concept, not execution
7. Extended Art Direction — Styling, casting, poses, hair/makeup, layout
8. Environment & Context — Where and why (conceptual, never decorative)
9. Texture, Material & Product Render — How surfaces are perceived
10. Final Image Signature — Finish, grain, temporal positioning

Without the brief, the Master Prompt must guess creative strategy. With it, the Master Prompt only executes.

The brief prompt template is included at {baseDir}/configs/Brief_Generator/brief_prompt.json.

Phase 3: Master Prompts (8 Funnel Stages)

Each funnel stage has a specialized Master Prompt that generates a production-ready JSON Blueprint. All share the same internal simulation:

• - ROUND -1: Brand Identity Forensics (stages 03+) — Unified Brand Style Manifest
• ROUND 0: Fidelity Lock — Product geometry & talent identity are IMMUTABLE
• ROUND 1: Stage Strategy — Strategic approach specific to funnel position
• ROUND 2: Graphic Design — UI, typography, CTA engineering

The Blueprint JSON covers: scene production, talent lock, camera perspective, subject action/pose/wardrobe, lighting, product constraints, layout architecture, typography, CTA engineering, and brand asset placement.

Master prompt files are included at {baseDir}/configs/Product_to_Ads/.

Reference Analyzer

Reference images (referencia) are optional and off by default. The pipeline generates creative direction internally from Brand Identity + Campaign Brief. Only use a reference when the user explicitly asks to clone a specific ad's style.

When used, the reference is analyzed for pose, photographic style, and location cues.

---

⚠️ CRITICAL: Required Inputs Checklist

Before running ANY ad generation, ensure these are provided:

Input	Required?	How to Get It
INLINECODE4	✅ ALWAYS	User provides the product page URL
INLINECODE5

✅ ALWAYS | Download from the product page, or user provides |
| --brand-profile | ✅ NEVER EMPTY | Pick from catalog or run brand-analyzer first. NEVER leave as "No Brand" if a brand is known |
| --prompt-profile | ✅ ALWAYS | Choose based on campaign objective |
| --aspect-ratio | Default: 4:5 | Change if needed for platform |
| --model | 🔶 OPTIONAL | Model/talent face. Ads with talent perform much better. Empty = product-only ad (no person). When used, pick from ~/clawd/models-catalog/catalog/images/ (114 models available) |
| --logo | 🔶 OPTIONAL | Try to find it. Use if good quality & easy to get. Skip if low-res or hard to find. Empty = bypassed |
| --reference | 🔶 OPTIONAL (off) | Only when user explicitly asks to clone a reference ad. Empty = bypassed |
| --creative-brief | 🔶 ON-DEMAND | Only when user gives explicit creative direction. Omit to let pipeline auto-generate from brand profile |
| --language | 🔶 ON-DEMAND | Only when user requests a specific language. Omit to use default (es) |

🚨 NEVER Skip These Steps:

1. 1. Product image — Download the main product photo from the product URL. The scraper is fragile; always provide a product image explicitly.
2. Brand profile — If the brand doesn't exist in the catalog, run brand-analyzer skill FIRST to generate one. Never submit with "No Brand" when a brand is known.
3. Brand logo — TRY to find it (Clearbit, logo.dev, brand website). Use if good quality. If not found or low-res, skip it — the variable accepts empty string (bypassed server-side).
4. Reference — Do NOT search for references by default. Only provide when the user explicitly asks to clone a specific ad or says "find a good ad to clone".

---

Auto-Preparation Workflow

When the user asks to generate an ad:

CODEBLOCK1

---

Usage

Full command (recommended):

CODEBLOCK2

With reference (only when explicitly requested):

CODEBLOCK3

Auto-fetch mode (downloads product image and logo automatically):

CODEBLOCK4

List available brands:

uv run {baseDir}/scripts/generate.py --list-brands

---

API Details

Endpoint: https://api.comfydeploy.com/api/run/deployment/queue
Deployment ID: INLINECODE17

ComfyDeploy Input Variables

Variable	Type	Description
INLINECODE18	string	Product page URL to scrape
INLINECODE19

image URL | Product image (uploaded to ComfyDeploy) | | model | image URL | Model/talent face reference. OPTIONAL — empty = product-only ad without a person. When used, select from models catalog (~/clawd/models-catalog/catalog/images/model_XX.jpg, 114 available) | | referencia | image URL | Style reference ad — OPTIONAL, empty = bypassed. Only when user asks to clone a reference | | marca | image URL | Brand logo — OPTIONAL, empty = bypassed. Use if found easily in good quality | | brand_profile | enum | Brand name from catalog (70+ brands) | | prompt_profile | enum | Funnel stage master prompt | | aspect_ratio | enum | Output format (1:1, 4:5, 5:4, 9:16, etc.) | | language | string | ON-DEMAND ONLY. Output language for ad copy/CTA. Default: es. Only send when the user explicitly requests a different language. Otherwise, DO NOT include this parameter — let the pipeline use its default. | | creative_brief | string | ON-DEMAND ONLY. Free-text creative direction override. Only use when the user explicitly asks for a specific creative direction, scene, mood, or concept. Otherwise, DO NOT include this parameter — let the pipeline generate its own brief from the Brand Identity profile automatically. |

---

Funnel Stages — Strategic Detail

01 — Awareness

Goal: Scroll-stop, curiosity, brand introduction Reject: Generic "product on table" concepts Strategy: Dynamic camera angles, world-building environments, high-concept creativity CTA: Soft or optional Visual Hierarchy: Talent → Product → Optional CTA

02 — Interest

Goal: Sustained attention, introduce value proposition Reject: Abstract visuals that hide the product Strategy: One clear visual idea, believable micro-world hinting at use-case CTA: Learn More, Discover, See Details Visual Hierarchy: Talent → Product → Headline → CTA

03 — Consideration

Goal: Informed evaluation, reduce uncertainty Reject: Pure mood storytelling, vague emotional content Strategy: Communicate WHAT product does, ONE primary differentiator, ONE proof cue CTA: Compare, See Details, Explore Visual Hierarchy: Talent → Product → Key Benefit → Proof Cue → CTA New: Adds Brand Identity Manifest to Blueprint JSON

04 — Evaluation

Goal: Validate purchase decision, proof & trust Reject: Pure mood, unsupportable claims, visual clutter Strategy: One trust anchor (quality/legitimacy/authority), one proof cue (reviews/certification) CTA: See Reviews, Verified Quality, Learn More Visual Hierarchy: Trust Anchor → Proof Cue → Product → Talent → CTA

05 — Conversion

Goal: Trigger decisive action, remove friction Reject: New hesitation-inducing info, complex compositions Strategy: One hero (product), one action, optional micro-reassurance CTA: Buy Now, Get Yours, Complete Order (PRIMARY visual element) Visual Hierarchy: Product → CTA → Optional Reassurance → Brand → Talent

06 — Retention

Goal: Post-purchase confidence, reduce churn Reject: Hard-sell, urgency, price talk Strategy: "You made the right choice" + "Here is the next step" CTA: Start, Set Up, Learn, Track (guidance, not purchase) Visual Hierarchy: Confirmation → Next Step → Product → Talent

07 — Loyalty

Goal: Strengthen emotional bond over time Reject: Sales layouts, instructional tone, aggressive CTAs Strategy: "This brand is part of who you are" — habitual engagement CTA: Optional: Explore, Be Part Of, Continue Visual Hierarchy: Brand World/Mood → Talent (identity mirror) → Product → Brand

08 — Advocacy

Goal: Turn customers into voluntary brand ambassadors Reject: Sales language, instructional tone, forced testimonials Strategy: Signal belonging, create share-worthy imagery, enable organic sharing CTA: Optional or absent: Join the Movement, Part of Us Visual Hierarchy: Mood → Talent (identity proxy) → Product (symbol) → Brand

09 — Morfeo Creative 🎨 (DEFAULT)

Goal: Maximum visual impact, narrative-rich, cinematic quality Reject: White backgrounds, studio shots, "product on table", generic poses, sterile compositions Strategy: Build immersive WORLDS, not backgrounds. Talent is a CHARACTER with emotion and action. Subtle surreal/magical elements elevate the mundane. Think movie stills + magical realism + high fashion. CTA: Present but integrated into scene aesthetics Visual Hierarchy: Scene → Talent (as character) → Product (organic in scene) → CTA Creative Philosophy:

• - NEVER a white background or studio
• Every image has depth (foreground/midground/background layers)
• Lighting is narrative (golden hour, practicals, colored atmosphere)
• One subtle surreal element per scene (impossible beauty, dream-logic detail)
• Wardrobe is costume design, not "simple clothes"
• Camera has personality (specific film stocks, intentional imperfections)

---

Creating New Ad Types

To create a new funnel stage or specialized ad type:

1. 1. Copy the closest existing Master Prompt from INLINECODE30
2. Redefine ROUND 1 with the new strategic objective
3. Adjust ROUND 2 UI hierarchy accordingly
4. Shift talent/product narrative roles
5. Modify CTA philosophy and copy voice
6. Keep the JSON output structure identical for pipeline compatibility
7. Maintain the Fidelity Lock (ROUND 0) — product and talent are always immutable
8. Save as Master_prompt_XX_NewStage.json — the node auto-discovers new profiles

Key Evolution Pattern Across Stages:

Aspect	Early (01-02)	Mid (03-05)	Late (06-08)	Morfeo (09)
Talent role	Attention anchor	Credibility anchor	Identity mirror	Character in story
Product role

Secondary hero | Evaluation hero | Familiar symbol | Organic in world | | CTA | Soft/exploratory | Proof-led → Decisive | Guidance → Optional | Integrated/aesthetic | | Copy voice | Intriguing | Clarity, proof, action | Supportive → Proud | Evocative/poetic | | Visual density | High-concept | Structured, scannable | Editorial, spacious | Cinematic/layered | | Environment | World-building | Context-rich | Lifestyle | Immersive + surreal | | Environment | World-building | Context-rich | Lifestyle, intimate |

---

Image Input Types

Binding Images (strict fidelity — immutable)

• - talent: Face/body locked, no deviation in facial structure, ethnicity, proportions
• product1-4: Shape, label text, material, proportions preserved 1:1
• brandlogo: UI/button style derived from logo geometry

Soft References (optional, off by default)

Reference image input (referencia) is optional. When provided, it's analyzed for:

• - POSEREF → Body position, limbs, weight, gaze, micro-gestures
• PHOTOSTYLEREF → Camera, lens, lighting, grading, grain
• LOCATIONREF → Setting, materials, colors, mood

When empty (default), creative direction comes from Brand Identity + Campaign Brief alone.

---

Brand Profiles

Catalog (70+ brands):

CODEBLOCK6

Creating new brand profiles:

Use the brand-analyzer skill: CODEBLOCK7

The Brand Analyzer uses a 3-phase methodology:

1. 1. Phase 1: Official research via Google Search (canonical data: name, founding, positioning, vision, mission, tagline)
2. Phase 1.1: Independent campaign research (10+ distinct campaigns via Google Images/Pinterest)
3. Phase 2-3: Visual analysis → JSON profile following the standard template

Output covers: brandinfo, brandvalues, targetaudience, toneofvoice, visualidentity, photography, campaignguidelines, brandbehavior, channel_expression, compliance.

---

Aspect Ratios

Ratio	Use Case
INLINECODE34	Default. Instagram feed, Facebook
INLINECODE35

Stories, Reels, TikTok |
| 1:1 | Square posts |
| 16:9 | YouTube, landscape banners |
| 5:4 | Alternative landscape |
| 2:3 | Pinterest |
| 3:4 | Portrait |

---

Config Files Reference

The skill includes reference copies of all pipeline configuration files:

CODEBLOCK8

These configs are the canonical reference for the pipeline's behavior. The actual live configs are stored in the ComfyUI deployment at ads_SV/configs/.

---

Known Limitations

1. 1. Product image scraping is fragile — always provide product images manually
2. Some websites block scraping — provide product data manually when scraping fails
3. Gemini hallucinations — occasional issues in complex reasoning steps
4. No brief editing — brief is generated automatically; manual override not yet supported
5. Logo & reference are optional — both use server-side bypass; empty string = not used. Logo: use if good quality. Reference: only on explicit request

---

Ad-Ready vs Morpheus

Feature	Ad-Ready	Morpheus
Input	Product URL (auto-scrapes)	Manual product image
Brand intelligence

70+ brand profiles | None | | Funnel targeting | 8 funnel stages | None | | Brief generation | Auto (10-point creative direction) | None | | Creative direction | Objective-driven (brief → blueprint) | Pack-based (camera, lens, lighting) | | Best for | Product advertising campaigns | Fashion/lifestyle editorial photography | | Control level | High-level (strategy-first) | Granular (every visual parameter) |

---

API Key

Uses ComfyDeploy API key. Set via COMFY_DEPLOY_API_KEY environment variable.

Source Repository

• - GitHub: PauldeLavallaz/ads_SV
• Architecture: ComfyUI custom node package with 3 nodes:

- ProductToAds_Manual — Full manual control, single format - ProductToAds_Auto — Auto-downloads images, generates 4 formats - BrandIdentityAnalyzer — Analyzes brands via Gemini + Google Search