← Search

Image-to-Code
Design 🧪 Install ran in our test sandbox · Oct 19, 2026 📌 Pinned version: commit beb16b0
Install test only, on the date shown. This does not guarantee the code is safe.
⚙ Needs: none — no account, API key, or paid plan required.

Image-to-Code

Image-first web design workflow: generate the design images first, analyze them deeply (text, type, spacing, color), then implement the frontend to match as closely as possible

At a glance
What
Image-first web design workflow: generate the design images first, analyze them deeply (text, type, spacing, color), then implement the frontend to match as closely as possible
Cost
Free
Needs
none — no account, API key, or paid plan required. (Using the skill later wants a host agent with image-generation capability — its core workflow generates the design reference images itself before any code; installing it needs no such tool.)
Install
Copy the installer prompt below into your Muse — your agent does the rest.

Version:

Quick answer

Image to Code Skill for Muse is a free Muse skill that turns a screenshot or design image into working code. It reads your image, recognizes layout, components, and styles, then generates clean front-end code matching the design, so a mockup becomes a starting codebase without rebuilding every element by hand.

⌁

Install

Copy the install package below, then paste it into Muse🧪 Install ran in our test sandbox · Oct 19, 2026📌 Pinned version: commit beb16b0

Step 0, before installing anything: read this install prompt, summarize in a few lines what this build does and what installing it involves, then ask me to confirm before you proceed. Prerequisites: none — no account, API key, or paid plan required. (Using the skill later wants a host agent with image-generation capability — its core workflow generates the design reference images itself before any code; installing it needs no such tool.) Install "Image-to-Code" for me — upstream name "image-to-code", an elite image-first website design-to-code skill by @leonxlnx, for visually important web tasks: hero sections, landing pages, marketing and startup sites, editorial brand pages, product pages, portfolios, premium multi-section websites, and redesigns where visual quality matters. Its workflow is mandatory and in this order: image generation first, deep image analysis second, implementation third — the generated images are the primary visual source of truth and the code is the translation layer. In Codex it prefers large, readable, section-specific images (one large image per section, not one compressed board for the whole site), fresh standalone re-generations for sections or detail views instead of cropping old images, and extra detail/extraction images whenever text, typography, spacing, buttons, cards or colours would otherwise be too small to analyse. Before implementing, it deeply extracts the visible text, typography character and scale, spacing and padding logic, button and component logic, colour palette, layout structure and section rhythm from every generated image, then implements the frontend to match them as closely as reasonably possible, without drifting into a generic coded reinterpretation. Its baseline configuration (adapted to the brief) is: DESIGN_VARIANCE 8, VISUAL_DENSITY 3, ART_DIRECTION 8, IMPLEMENTATION_CLARITY 9, IMAGE_USAGE_PRIORITY 9, SPACING_GENEROSITY 9, ANALYSIS_PRECISION 10, IMAGE_GENERATION_EAGERNESS 10, UI_SIMPLICITY_DISCIPLINE 9. It picks one coherent combination from its variation engine (theme paradigm, background character, typography character, hero architecture, section system, exactly 4 signature components, exactly 2 motion-implied cues), and it enforces: a clean, spacious hero of 1–3 headline lines that stays readable on a small laptop, no cards-inside-cards-inside-cards or giant rounded wrapper sections, reduced micro-UI clutter (no unnecessary pills, fake system markers or decorative technical labels), fixed-aspect media frames, generous even spacing, section rhythm that varies without losing coherence, anti-AI-slop rules for layout, visuals, typography, content and density, default section packs (4, 8 or 12 sections, each its own image in Codex), multi-image consistency (same brand world, type scale, spacing, CTA styling and component family across all section images), and a 21-point clarity check before finalizing. This skill has MOVED repositories. It used to live in leonxlnx/taste-skill under skills/image-to-code-skill/; that location now holds only a short stub (435 characters) saying "Moved to github.com/Leonxlnx/imagegen-skills", keeping the install name unchanged — "image-to-code" — and declaring disable-model-invocation: true, meaning it runs only when called by name. The stub's install command is: npx skills add https://github.com/Leonxlnx/imagegen-skills --skill "image-to-code". Install the pinned version below from the new repository, not the old taste-skill location and not the latest unpinned code. Repository: https://raw.githubusercontent.com/Leonxlnx/imagegen-skills/beb16b0d0f4dae9a6f5cb04ea41e97b93a54ab8e/skills/image-to-code/SKILL.md (the skill is the skills/image-to-code/ folder at that commit; the reviewed source is its SKILL.md, a single instruction file) Pinned version: install the code at commit beb16b0d0f4dae9a6f5cb04ea41e97b93a54ab8e exactly (Skill Harbor tested this version). Do not substitute a newer version. If this commit is unreachable, stop and tell me. [FINGERPRINTS] SKILL.md at the pinned commit: SHA-256 4c060a8064a8b13380bc2ef3d6e6a1d0b1e316aa093cd9807d0ce9e4eeb037fa, 36,442 bytes. If the fetched file does not match, it changed since the review: do not install it yet, tell me what you found and let me decide. Updated October 19, 2026: Skill Harbor regenerated this install prompt from the current upstream file at its new home. The one correction that matters most: the previous prompt pointed at leonxlnx/taste-skill at commit 717446e07a78d4e6d7918b1fde1a263378eebac9, where skills/image-to-code-skill/SKILL.md is now only the 435-character moved stub described above — installing from the old location would install the stub, not the skill. The skill itself is the full 36,442-byte file now living at skills/image-to-code/SKILL.md in Leonxlnx/imagegen-skills, with the same name ("image-to-code") and the same image-first workflow the previous prompt described (generate first, analyse deeply, implement third; one large image per section in Codex; never crop old images for section extraction; baseline configuration and anti-slop rules as summarised above). What is stated up front now, before anything happens: nothing is sent anywhere during installation itself, the file contains no executable code and no scripts, and in use the design images are generated by your own host agent's image-generation tools — the file names no external service, no API and no key (see the safety check below). 1. Fetch the whole skills/image-to-code/ folder at the pinned commit into a temporary folder, then compare SKILL.md with the fingerprint from Skill Harbor's review of 2026-10-19 (see [FINGERPRINTS] above). If it does not match, it changed since the review: do not install it yet, tell me what you found and let me decide. 2. Safety check: read SKILL.md and review it for anything suspicious (unexpected network calls, credential harvesting, code that runs at install); STOP on any red flag and tell me. Tell me in plain words what the skill does when used, which services it connects to, whether it reads or sends any personal information or keys, and anything that seems unusual. Confirm these points from the file: - Nothing is sent anywhere during installation itself, and no code runs during installation: the skill is a single instruction file (SKILL.md) — no bundled scripts, no shell commands, no install hooks. Installing is only copying the file. - In use, the file instructs the agent to generate the design image(s) itself first, using the host's own image-generation capability (the file is written for Codex and says "generate the design image(s) yourself" and "when image generation is available"), then deeply analyse them, then write the frontend code in your project. The file names no external service, API, URL or API key, and asks for none: whatever image generation happens goes through your host's own tools, under your host's own terms, with the design brief and site content you give it as the generation prompts. Do not put personal information or secrets in those briefs. - No part of the file reads credentials, API keys or personal information, and nothing in it sends your files anywhere: the analysis and implementation steps work on the images the agent just generated and on the frontend code it writes in your project. - Name-only invocation: the moved stub left in the old repository declares disable-model-invocation: true — this skill runs only when called by name ("image-to-code"), not automatically. Its own triggers, once called, are visual website requests (a beautiful hero, a premium landing page, a redesign, a more aesthetic interface, a portfolio or startup site, a multi-section concept); for mostly technical tasks, bug fixes, or work from a precise design system you provide, the file itself says direct coding first is more acceptable. 3. When I say yes, install it as a skill: copy the entire skills/image-to-code folder into the agent's skills directory, in a folder named "leonxlnx-image-to-code", keeping its structure as it is. 4. Verify with no network calls: frontmatter valid (name: image-to-code), files in place. 5. Report what was installed, where, and what I still need to do myself (e.g. start a fresh session so the host re-indexes the skill, then call it by name — "image-to-code" — with a visual web task; for its core workflow your host needs image-generation capability, and the design brief you give it is what goes into the generation prompts). GitHub is optional: if I have a GitHub account or the gh CLI, you may use it; otherwise public access is fine. Never require it unless it's in the prerequisites above. Rules: don't touch anything outside the temp folder and the install target. If anything looks off, stop and ask me.

🔒 Security, provenance and what this touches
🧭 What this touches
Based on the install prompt we scanned, this skill does not run commands, make third-party network calls, install packages, write files, or ask for API keys.
Install prompt analysis · 2026-10-13
🔒 Security and provenance
📌 Pinned version: commit beb16b0 · 2026-10-19
🧪 Install ran in our test sandbox · Oct 19, 2026
👁 Observed during the install test · Oct 19, 2026 : no network contact
🔎 Last checked · Oct 19, 2026 : no change detected in the files linked by this install prompt
Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify
Wanna be extra careful?

The install prompt below already includes the vetting steps: your agent follows the community checklist before installing anything with executable code. Want more?

How to check a skill before installing →

✓

Verification certificate

Dated facts about what was checked for this listing, and how. This is not a safety certification.

Pinned version
beb16b0 · 2026-10-19
Sandbox install test
Install ran in our test sandbox · Oct 19, 2026
Creator identity
Not verified
Fingerprint watch
Last checked · Oct 19, 2026 : no change detected in the files linked by this install prompt
Install prompt analysis
2026-10-13 : No commands, network calls, packages, file writes or API keys detected.

Skill Harbor never reviews the code of a listed skill. These are dated facts about the install prompt and our install test, not a safety certification. Review a skill yourself before you install it. How we verify

?

Questions

How do I install a skill?

Every product page includes a copy-paste install prompt. Paste it into your Muse and it sets the skill up for you — no manual configuration.

Where does my money go?

Straight to the seller. Skill Harbor never processes payments: checkout happens on the seller’s own page, usually Stripe.

What does the ✓ next to a creator’s name mean?

It means we confirmed the identity of the person behind the listing. It says nothing about the code itself — always check a skill before installing it.