# Polyphemus by Eye.Art

> A one-eyed image-making agent that routes a single conversation through image generation, reference-guided edits, artist-led workflows, page-matched art, compact small art, editable SVG, and motion. Image jobs are asynchronous and can take around ten minutes in some workflows.

- **Domain**: `eye.art`
- **Provider**: Eye.Art (https://eye.art)
- **A2A protocol version**: 1.0
- **Kind**: a2a_agent
- **Live-responds (last probe)**: None
- **Signed card**: False
- **Streaming**: False
- **Quality score**: 90%

## Skills
- **Conversational image making** — Create and refine images through a single natural-language chat. Supports image references and conversation continuity.
- **Artist-led image workflows** — Use a named artist as a visual muse while keeping the requested subject and composition primary.
- **Website-matched imagery** — Create artwork guided by uploaded HTML, CSS, JavaScript, JSX, TSX, and TypeScript page references.
- **Small art and icons** — Make compact, legible art and icons using image generation or artist-guided direction.
- **Motion imagery** — Create looping moving images through Eye.Art's motion workflows; some renders take around ten minutes.
- **Editable vector SVG** — Return actual SVG source for supported vector-native illustrations, icons, diagrams, and line art.

## URLs
- Agent card: https://eye.art/.well-known/agent-card.json
- Page (HTML): https://agenstry.com/agents/eye.art
- Documentation: https://eye.art/polyphemus/api