WyattLee-nanami

WyattLee-nanami / weft

Public

Weft — a local Claude Code workbench. Manage skills, see token usage, browse history. Native macOS, fully offline, MIT.

github.comWyattLee-nanamiskillbox claude-code desktop-app developer-tools macos react
14
1
94% credibility
Found May 18, 2026 at 24 stars 4x -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Rust
AI Summary

Weft is a native macOS app that helps you browse, search, and manage your Claude Code skill library. It reads your local skill files, organizes them by category, shows which skills you use most often, and lets you quickly copy slash commands to paste into Claude Code. The app also tracks your conversation history and provides backup/restore features for your configuration.

How It Works

1
📚 Your skill library has grown

You've collected dozens of Claude Code skills and can't remember which one does what.

2
🚀 You open Weft

The app automatically scans your skill folder and shows everything organized in one place.

3
🔍 You search or browse

Type to find skills instantly, or browse by category like design, tools, or life scenarios.

4
📊 You see which skills you actually use

Usage stats reveal what's working great versus what's been gathering dust.

5
📋 You copy the slash command

One click copies the skill name to your clipboard, ready to paste into Claude Code.

✨ You find the right skill in seconds

No more scrolling through walls of skill names — you get exactly what you need.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 24 to 14 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is weft?

Weft is a native macOS app that turns your Claude Code skill library into a searchable, browsable workbench. If you have more than a handful of skills in `~/.claude/skills/`, you know the pain of scrolling past 50+ entries trying to remember which one does what. Weft reads your local skill files and renders them with fuzzy search, usage tracking, and one-click slash command copying. It also tracks your token consumption across sessions and lets you browse your full Claude Code conversation history. Built with Rust and React inside a Tauri shell, it ships as a ~10MB native app that runs fully offline.

Why is it gaining traction?

The hook is simple: your skill index is broken, not your skills. Weft shows you which skills you actually use versus which ones are gathering dust, calculated from your session history. The presentation mode (`⌘D`) is a thoughtful touch for screen-sharing or team demos. Batch operations let you disable unused skills en masse without hunting through folders. The backup and restore features protect your entire Claude config with one click. It's the kind of tool that feels obvious once you see it.

Who should use this?

Claude Code power users with 20+ skills will get the most value. If you have a cluttered skills folder and spend time scrolling instead of shipping, Weft fixes that. Teams that share skill repositories can use the presentation mode for onboarding. Developers curious about their actual Claude Code usage patterns will appreciate the token breakdown by model and time window. If you have five skills and remember all of them, this won't change your life.

Verdict

Weft solves a real problem with a clean, focused implementation. At 14 stars, it's early and unproven -- there's no pre-built release yet, so you'll need to build from source. The roadmap promises live re-indexing (v0.2) and inline editing (v0.5), which would significantly improve the workflow. With a credibility score of 0.949999988079071%, this is a small, personal project that happens to be genuinely useful. Worth installing if you fit the use case, but watch the releases for v0.1.0 before recommending it to your team.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.