48Nauts-Operator

Skill Dash — Data skill for Claude Code

Data community

Skill Dash uses Jev to judge Claude Code and Codex skills and plugins: usefulness, redundancy, clarity, duplicates, safety.

How to install Skill Dash

This entry records only its repository, not the path inside it, so there is no exact command to give. Open 48Nauts-Operator/skill-dash and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Skill Dash does

Skill Dash uses Jev to judge Claude Code and Codex skills and plugins: usefulness, redundancy, clarity, duplicates, safety. Local dashboard plus the corpus pipeline behind whichskills.dev. MIT.

Alternatives in Data

Context Mode — Benchmark Results — Benchmarked against real outputs from popular Claude Code MCP servers, Skills, and dev tools 5.6k ★
LLM Redteam — LLM red-team corpus runner — fires categorized prompt-injection / jailbreak / system-prompt-leak / data-exfil 4.5k ★
Debug Plus — Enhanced debug with swarm integration and prevention pipeline 508 ★

README

skill-dash

Judge a tree of Claude Code and Codex skills with [Jev](https://typesafe.ai), TypeSafe's typed-judgment model. One row per skill, one question per row: keep, rewrite, merge or delete. Also the pipeline behind [whichskills.dev](https://whichskills.dev), a public census of 18,041 skills from the 200 most-starred repos.

Python standard library and SQLite, no dependencies, no build step. Runs on 127.0.0.1 only. Nothing leaves your machine except the Jev calls you choose to make.

You need

A Jev API key. Sign in at console.typesafe.ai, create a key, export it as TYPESAFE_API_KEY (on macOS the Keychain entry service xnaut, account plugin/typesafe/TYPESAFE_API_KEY also works). Docs: docs.typesafe.ai. Jev was in early access when this was built; hello@typesafe.ai is the listed contact if you have no access yet. The loader, the evidence scan and the static safety pre-scan run without a key; judgments and the overlap audit need one.
Python 3. Standard library only.

Cost, for orientation: TypeSafe listed $42 per billion input tokens on 2026-09-20. A content-only pass over 78 skills is well under a cent; the full-text safety read of 545 bodies for whichskills.dev was about 18 cents. The dashboard shows the token count of every run.

Quick start

git clone https://github.com/48Nauts-Operator/skill-dash && cd skill-dash
export TYPESAFE_API_KEY=...
python3 server.py --port 3345       # your own ~/.claude/skills tree plus enabled plugins
python3 server.py --port 3346 --roots ~/some/skills-repo   # any repo of SKILL.md files

Open the port in a browser. Settings runs the overlap audit; Judges edits the questions; the drawer on each row holds the evidence, the Jev distributions, a radar against the tree average and your decision.