ChosenXu

Eagle Untagged Organizer — Design skill for Claude Code

Design community

Agent Skills-compatible skill: batch-rename, annotate & tag untagged design assets in Eagle via eagle-mcp.

How to install Eagle Untagged Organizer

This entry records only its repository, not the path inside it, so there is no exact command to give. Open ChosenXu/eagle-untagged-organizer and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Eagle Untagged Organizer does

Agent Skills-compatible skill: batch-rename, annotate & tag untagged design assets in Eagle via eagle-mcp. Works with Claude Code, Codex CLI, Gemini CLI, GitHub Copilot, Cursor & WorkBuddy.

Alternatives in Design

Od Contribute — Open a first-contribution PR (or bug issue) on nexu-io/open-design — works for non-coders too 91.8k ★
Figma — Fetch design context, screenshots, and assets from Figma for implementation 14.6k ★
Canvas — Open, create, or update a visual canvas — add images, text, PDFs, wiki pages, and banana-generated assets to O 3.3k ★

README

Eagle Untagged Organizer

An [Agent Skills](https://agentskills.io)-compatible skill that batch-organizes **untagged** design assets in [Eagle](https://eagle.cool/) via the `eagle-mcp` MCP server — renaming, annotating, and tagging UI/UX references and graphic-design works in one pass. Works with Claude Code, Codex CLI, Gemini CLI, GitHub Copilot, Cursor, and WorkBuddy.

What it does

For every selected untagged asset it produces three outputs, written back to Eagle in a single `item_update` call:

Name — a concise, title-style name (searchable in the Eagle grid); if an asset already has a good name, it is kept or proposed for overwrite rather than blindly renamed.
Annotation — a structured five-field block (设计类型 / 结构 / 视觉 / 用途 / 参考价值)
Tags — selected verbatim from a controlled three-dimension vocabulary (design domain / visual style / technique)

Output language is configurable: 简体中文, 繁體中文（港式）, English, 日本語, 한국어, Русский, Español, Deutsch. When the user's instruction is written in a supported language, that language is inherited as the output language; otherwise (unsupported or ambiguous) it falls back to English.

Highlights

Truly "reads" each asset before acting — it never guesses from filenames. It first runs a multimodal pre-flight check, reading each image to understand its subjects, colors, and layout, and only then produces names and annotations — quality is guaranteed.
Structured, reusable, and on-track — names are short titles, not analytical sentences; annotations follow a fixed five-field template; tags come from a controlled three-dimension vocabulary and are selected verbatim (no invented terms). The result is a consistent library whose tag taxonomy never drifts out of control. 3