styleref.io

Search StyleRef's public gallery of reusable creative styles (visual + voice).

Fetch a StyleRef style compiled into a ready-to-use spec.

Wrap a user request in a StyleRef style and return a single ready-to-paste prompt for the target tool (ChatGPT, Claude, Gemini, Midjourney, FLUX, DALL·E, Stable Diffusion, or generic).

Return a complete STYLE.md file body (the stylemd format, with `generator: styleref.io` frontmatter) for a StyleRef style.

List the styles in the connected user's own StyleRef library (name, visibility, last update, link).

Save (like) a public StyleRef gallery style to the connected user's library so it shows under My Styles → Saved.

Fork (remix) a public StyleRef gallery style into the connected user's library as a private editable copy with lineage back to the original.

Create a new style in the connected user's library.

Edit one of the connected user's own styles: rename it and/or add or replace blocks (a supplied block replaces the existing block of the same type, otherwise it is added).

Lint STYLE.md text against the StyleRef spec: frontmatter (name, generator, sections), section ids, body structure, and hex-color validity.

Run StyleRef's AI extraction to derive a structured style (colors, typography, mood, lighting, voice…) from any combination of: a text description, a named reference (e.g.

Poll an extraction started by extract_style_from_image.

Publish one of the connected user's own styles through StyleRef's moderation pipeline: visibility "public" lists it in the gallery (requires at least 10 style items), "unlisted" activates a shareable link only.

Lightweight visibility control for one of the connected user's own styles: "private" revokes the share link, "unlisted" removes a listed style from the gallery while keeping its link (only allowed once the style has already been moderated — to first publish or activate a link, use publish_style).