devkindhq
OrganizationClaude Code skill for Ideogram AI image generation — structured JSON caption prompting, color palette control, and MCP-based style extraction from reference images.
Categories
Indexed Skills (15)
brand-identity-sheet
Generates a single polished "brand identity sheet" image — a Dribbble-style board showing a UI card with the wordmark, a grid of textured app icons, pill-shaped buttons, a giant central wordmark treatment, and a secondary lockup with a brand glyph — and actually renders it via the ideogram MCP tools, not just drafts a prompt. Use whenever the user wants to see a brand's whole *locked* visual system in one image (not a single logo), asks for an "identity sheet," "brand system sheet," "brand board," or "style board" with real colors/type already decided, wants app-icon / UI-component exploration alongside the logo, or references a reference image they want matched or extended. Distinct from logo-prompting's single-mark prompts — trigger this when the ask is bigger than one logo and wants the full palette/type/icon/UI system shown together, and the brand's colors/type/style are already decided (a brand.md or design.md exists, or the user can state them directly). If the brand is still early-stage and unlocked —
character-model-sheet
Generates a single polished "character model sheet" image — a multi-panel character design board showing a full-body turnaround (front / 3-4 / side / back), a head & detail close-up sheet (face, hair/headwear, hands/props), a color palette, an overview text block (name, role, personality, wardrobe), and an optional accessories/interior technical diagram — and actually renders it via the ideogram MCP tools, not just drafts a prompt. Use whenever the user wants a mascot, game character, animation character, or brand character rendered as a full reference/turnaround sheet (not a single portrait), asks for a "character sheet," "turnaround sheet," "character reference sheet," or "model sheet," wants multiple angles/views of the same character shown together, or references a character-sheet image they want matched or extended. Distinct from logo-prompting (single mark) and brand-identity-sheet (brand system of icons/buttons/wordmark) — trigger this specifically when the subject is a character (a mascot, a person, a
custom-model-training
Trains a custom Ideogram model on a folder of reference images, then generates a new image with that trained model to prove it works — a real end-to-end pipeline (create_dataset, upload_dataset_assets, train_model, poll get_model, generate_image with custom_model_uri), not just an explanation of how training works. Use whenever the user wants to "train a custom model," "fine-tune" on their own images, get "consistent" output of a specific character, product, or brand across many future generations, lock in a look so it doesn't drift between generations, or turns a set of reference images (including a previously generated brand-identity-sheet or character-model-sheet) into a reusable model. Also trigger this when the user asks what a `custom_model_uri` is or how to get one — that URI only exists after running this pipeline. Distinct from `brand-identity-sheet` and `character-model-sheet`, which produce the single locked reference image in the first place; this skill is the next step, turning that locked image
ideogram-prompt
Prompting technique guide for Ideogram image generation via the connected Ideogram MCP (generate_image, describe_image, remix_image, edit_image). Use whenever the user wants to generate, remix, or edit an image with Ideogram — including moodboards, style-matching a reference image, extracting a visual "recipe" from existing images and reapplying it to a new subject, or getting precise control over color palette, composition, and in-image text. Trigger this even if the user doesn't say "Ideogram" by name but is clearly asking for image generation and the Ideogram MCP tools are the ones available.
logo-prompting
Living reference for writing logo and brand-mark image-generation prompts (Ideogram-first, via ideogram-pack:ideogram-core-workflow-a). Use whenever the user asks for a logo prompt, a wordmark/icon direction, or wants feedback on a logo image-gen prompt they've drafted — even if they just paste a prompt and say "is this good" or "make this better." Also use when a brand.md exists and the user wants its Visual layer turned into an actual generation prompt. For a 3×3 moodboard/brand-exploration board instead of a single mark, use `moodboard-generator` instead. This skill accumulates real examples over time in examples/ — check there for a matching style family before starting from scratch, and append new accepted prompts back into it after each real logo job.
marketing-ui-mockup-hero-generator
Generates finished raster marketing visuals — landing-page hero shots, device-in-context product mockups, and dashboard screenshots — and actually renders them via the ideogram MCP tools, not just drafts a prompt. Use whenever the user asks for a "hero image," "landing page hero," "device mockup," "product mockup," "dashboard screenshot," "App Store screenshot," "pitch deck visual," or "pitch deck screenshot." This skill produces finished images, not functional HTML/CSS/React — a request for working code, a live component, or a responsive layout should go to a code-output tool instead, never get pulled into this skill just because it also involves UI. Sibling to `brand-identity-sheet`, `character-model-sheet`, and `moodboard-generator` structurally (same prompt-composition pattern, same paragraph-prompt + compositional-deconstruction pair of artifacts), but targets marketing-visual deliverables — a single hero shot, a device-in-context product photo, a dashboard screen — rather than a whole brand system or a
moodboard-generator
Generates a single polished "moodboard" image — a 3×3 grid brand-exploration board (Color Palette, Typography, Logo Exploration, Iconography Style, Photography & Graphics, Material Samples, Abstract Patterns, Mood Imagery, Application Mockup) — and actually renders it via the ideogram MCP tools, not just drafts a prompt. Use whenever the user wants to see a brand's early visual direction *before* a logo or full identity system is locked, asks for a "moodboard," "mood board," "vibe board," "vision board," or "brand exploration board," wants to explore adjectives/palette/imagery together as a first pass, or references a moodboard image they want matched or extended. This is the tool for the earliest, fuzziest stage — before there's even a full brand.md — so use it even when the user only has a few loose words to go on ("something warm and analog," "no adjectives yet, just a name and an industry"). Distinct from `brand-identity-sheet` (which renders a *locked* brand system — UI card, app-icon grid, wordmark at s
remove-background-workflow
Turns Ideogram's `remove_background` call into a guided workflow — resolve the source image, strip its background to a transparent PNG, save the result with its identifiers, and point to common follow-ups. Use when the user explicitly asks to "remove the background from this," "make this transparent," "cut out the subject," "isolate this on transparent," "strip the background," or "give me this on transparent/white/none." Only triggers on that kind of explicit request — it does not auto-run after `generate_image`, `edit_image`, or any other skill to proactively offer background removal.
seamless-pattern-texture-generator
Writes, generates, visually verifies, and iterates on seamless/tileable surface-design patterns — florals, geometrics, stripes, damask, plaids, novelty prints — and tileable material textures — fabric weaves, paper grain, wood, stone, leather, packaging substrates — for wallpaper, packaging, textile print, and digital surface design. Trigger on explicit requests using phrases like "seamless pattern," "tileable," "repeating pattern," "wallpaper design," "fabric print/texture," "packaging surface pattern," "surface design," or "this needs to tile edge to edge." Triggers on explicit request only — it does not auto-wire from `brand-identity-sheet`, `moodboard-generator`, or any other sibling skill; those skills are unmodified and unaware of this one, the same non-auto-wiring precedent `collections-management` sets for its own triggers.
upscale-image-workflow
Raises the resolution of one specific Ideogram image — from this session's generations, `get_recent_generations`, a collection, or a local upload — via the connected `upscale_image` tool. Use whenever the user asks to "upscale this," "make this higher-res," "upscale that logo render 4x," or "increase the resolution on the image I just uploaded." Wraps exactly one tool (`upscale_image`) and does NOT wrap `remix_image`, `reframe_image`, `edit_image`, or `remove_background` — those are separate tools, out of scope here. It does not auto-trigger from `brand-identity-sheet`, `character-model-sheet`, `moodboard-generator`, or `collections-management`; those four are unmodified and unaware of this one, and this only runs on explicit request.
photographic-icon-set-generator
Generates consistent, locked-style sets of photographic, textured, or 3D-material-rendered icons via Ideogram — 3D-rendered, claymation, hand-painted, or isometric-with-material-texture icon packs that hold a shared style recipe across every icon in the batch. Use when the user asks for "3D icon set," "claymation icons," "textured/material icon pack," "isometric icon set with material texture," "photographic icon set," "match this icon style," or "add an icon to my [style] set." Out of scope — flat, line, glyph, minimal, or generic vector-style icon sets (these need deterministic SVG code, not diffusion output); trigger `ideogram-prompt` for single one-off "generate me a 3D camera icon" requests instead. This skill exists specifically for the multi-icon consistency problem, where the same locked style recipe carries across a batch of subject variations.
social-marketing-asset-generator
Produces a finished social/marketing image asset — a social graphic, banner, or display ad — with its headline, subhead, CTA, and brand mark baked directly into the pixels by Ideogram, not composited afterward. Use when the user asks something like "make me an Instagram ad for X," "banner ad for our launch," "social graphic announcing Y," or "give me this in Story + Feed + LinkedIn sizes." Triggers only on explicit request for finished ad/banner/social-graphic creative — it is not auto-wired from `brand-identity-sheet`, `logo-prompting`, `moodboard-generator`, or `character-model-sheet`, and does not fire just because one of those skills ran. It is a sibling to, not a replacement for, those four skills — they establish or explore brand truth (a moodboard, a locked identity system, a single mark, a character sheet); this skill spends an already-decided brand truth and literal copy on one rendered creative asset.
bulk-image-generation-workflow
Turns one locked structured JSON caption into an on-brief batch of N prompt variations, submits them as a single `generate_images_bulk` call, tracks the async job, and reviews/culls the results into a shortlist. Use whenever the user says "generate a batch of," "give me N variations," "generate variations of this," "bulk generate," "sticker pack," or "generate 20 options of X." Does not re-implement caption-writing (defers to `ideogram-prompt` for building or refining the base caption) or collection-filing (defers to `collections-management` for organizing the shortlist afterward), and is not for single-image generation — a one-off image stays with `ideogram-prompt`'s `generate_image`.
concept-grid-generator
Generates a single "concept grid" image — a labeled comparison sheet showing several DISTINCT motif/icon concepts as tiles side by side, one locked palette and one locked rendering technique held constant across every tile — and actually renders it via the ideogram MCP tools, not just drafts a prompt. Use whenever the user wants to compare multiple genuinely different icon/motif ideas cheaply before committing to a full board, asks for a "concept grid," "icon exploration grid," "concept comparison sheet," or "let's see a few directions before we build the full sheet," or is choosing between candidate motifs for a mark that will later go into `brand-identity-sheet` or `character-model-sheet`. Distinct from `bulk-image-generation-workflow` (N variations of ONE locked caption/idea — same concept, different pose/prop/angle) and from `moodboard-generator` (a whole-brand loose exploration board with an unlocked, still-open palette) — this skill locks the palette and technique first and varies only the *concept* acr
world-builder
Builds out a fictional "world" around an already-trained custom character model — a staged, multi-batch pipeline (no-mascot moodboard, paired with/without-model scene tests, society & culture batch, family/village/society deep pass, landmarks/maps/art pass, optional contamination-check redo) that goes from "I have a trained mascot" to "I have a whole visual world this character lives in," using `generate_image`, `generate_images_bulk`, `remix_image`, and `describe_image`. Use whenever the user wants to expand a locked mascot or character (already produced by `character-model-sheet` and trained by `custom-model-training`) into a broader brand world — community scenes, rituals, hierarchy, family/village life, landmarks, maps, art objects — or asks for "world-building," "brand world," "the world Ding-Bot lives in," "expand the universe around this character," or "society/culture pass." Distinct from `character-model-sheet` (locks the one character's own design) and `custom-model-training` (turns that locked desi
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.