convez.ai
SkillDesign & UIIntermediário
kovishin-hdi

Hillarys Image Tone Skill

kovishin-hdi/hillarys-image-tone-skill

Padronize tom e cor de imagens para a identidade Hillarys

Ver no GitHub
InicialPluginScore · de 0 a 100

Skill para uniformizar a cor e o tom de imagens conforme a identidade visual de Hillarys Curtains & Blinds. Transforma fotos brutas (de telefone, WhatsApp, instaladores) em ativos visuais prontos para marca, aplicando uma paleta visual quente, morna e premium através de instruções AI reutilizáveis. Pode ser usado diretamente com Claude, ChatGPT, Gemini, Sharp.js ou editores manuais como Lightroom.

Instalação

$npx convez-ai kovishin-hdi/hillarys-image-tone-skill

1 comando no terminal: detecta o formato do repositório (marketplace ou plugin avulso) e instala via Claude Code. Requer Node 18+ e o Claude Code instalado.

Métricas

Estrelas
1
Forks
0
Acompanhando
0
Saúde
28%
Último commit
há 5 meses
Versão
—
Issues abertas
0
Contribuidores
1

Casos de uso

  • Harmonizar conteúdo visual de múltiplas fontes (clientes, instaladores, redes sociais)
  • Converter imagens telefônicas em ativos de marca sem re-fotografar
  • Manter consistência visual em campanhas de marketing e feeds sociais

Veredito

Solução sólida e bem documentada para padronização de tom visual em escala. Útil especialmente para marcas com alto volume de conteúdo visual de múltiplas fontes. A abordagem de guia portável (markdown) é inteligente, mas o valor depende de integração efetiva com fluxos de produção reais.

README

A tradução para o português deste README ainda está na fila.

Hillarys Image Tone AI

Turning usable images into Hillarys-ready visuals.


What it is

A reusable AI editing instruction system that packages the Hillarys visual identity into a portable, consistent set of rules — so any AI tool, editor, or developer can take a raw image and make it look like it belongs on the Hillarys feed.

The core file is Hillarys-Image-Tone-Guide.md — a standalone markdown specification you can upload to ChatGPT, Claude, Gemini, or any AI agent. It contains everything needed to colour-grade and tone-match images to the Hillarys Curtains & Blinds brand.


Why it exists

Hillarys has images. Not all are brand-ready.

  • Many images come from phones, WhatsApp groups, installers, consultants, or customer homes.
  • The content may be useful, but the tone is often too cold, harsh, dark, saturated, or inconsistent.
  • Rejecting these images reduces content volume and creates unnecessary dependency on professional shoots.

This skill solves that. It upgrades what already exists — applying Hillarys visual identity quickly and consistently, converting average usable images into stronger organic assets, and increasing the usable content bank without increasing shoot cost.


What the tone looks like

Warm. Soft. Premium. Homely.

The skill moves images toward the Hillarys look:

  • Creamy, golden highlights
  • Lifted shadows (never true black — warm charcoal)
  • Muted, understated colour (warm tones preserved, cool tones quietened)
  • Faded blacks for a soft, film-like matte finish
  • Peachy, healthy skin tones
  • A natural "lived-in luxury" feel

What's inside

hillarys-image-tone-skill/
│
├── Hillarys-Image-Tone-Guide.md    ← The portable tone spec (use anywhere)
├── CLAUDE.md                        ← Project instructions for Claude Code
├── README.md                        ← You are here
│
└── .claude/
    └── skills/
        └── hillarys-image-tone.md   ← Same spec, loaded as a Claude skill

The Guide covers:

SectionWhat it gives you
Brand Tone OverviewMood, visual principles, and the "why" behind every adjustment
Colour Profile ParametersDetailed specs: warmth, saturation, contrast, shadow lift, skin tones, black point
Sharp.js ParametersCopy-paste Node.js functions for programmatic colour grading
AI Prompt TemplatesFull and short prompts for ChatGPT, Claude, Gemini, or any image-editing AI
Lightroom / Manual EditingExact slider values, tone curves, HSL channels, and split toning settings
Batch Processing NotesPer-category adjustments and source lighting compensation tables

How to use it

With any AI agent (ChatGPT, Claude, Gemini, etc.)

  1. Download Hillarys-Image-Tone-Guide.md from this repo
  2. Upload it as context/instructions alongside your raw image
  3. Ask the AI to edit the image following the guide
  4. The AI Prompt Templates section has ready-to-use prompts

With Claude Code (as a skill)

The .claude/skills/hillarys-image-tone.md file is automatically loaded when Claude Code works in this repo — no upload needed.

With Sharp.js (programmatic)

Copy the functions from Section 3 of the guide. Run them against any image directory for automated batch processing.

With Lightroom / manual editing

Section 5 has every slider value, tone curve point, and HSL adjustment you need to replicate the look by hand.


How the workflow runs

  1. Collect — Bring in raw images from installs, customer homes, consultants, site visits, BTS captures, or WhatsApp drops.
  2. Select — Choose images with a usable moment, angle, product, room, or story. The source image must still have content value.
  3. Apply Skill — Run the Hillarys Image Tone instruction set to warm, soften, lift shadows, mute colours, and align the visual mood.
  4. Review — Check whether the edited image feels warm, natural, premium, soft, and usable — not fake, over-filtered, or campaign-heavy.
  5. Deploy — Use the output for organic posts, Reel covers, Stories, project highlights, before-after content, or visual content banks.

What it improves

  • Visual consistency across all organic content
  • Content volume — more images become usable
  • Speed — seconds per image instead of hours
  • Brand alignment — every image carries the Hillarys tone
  • Cost efficiency — reduces dependency on professional shoots for everyday content

What it does not do

BoundaryDetail
Not a shoot replacementCampaign visuals, catalogues, print assets, website hero images, and major brand films still need higher production control.
Not a miracle fixIt can improve tone and brand fit, but cannot fully rescue poor framing, low resolution, bad composition, or irrelevant content.
Not 10/10 outputThis is a 5-to-7 system. The purpose is practical uplift at scale — not turning every image into a final campaign masterpiece.

Built for

Hillarys Curtains & Blinds — India market, organic content pipeline.

Part of the Hillarys AI Content Engine.


Maintained by the Hillarys marketing team.

Ver o original no GitHub

Bom para

Relacionados