LinklyAI

Best Skills Runner — Data skill for Claude Code

Data community

The data pipeline behind LinklyAI/best-skills — a daily, cross-ecosystem ranking of AI agent skills.

How to install Best Skills Runner

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

What Best Skills Runner does

The data pipeline behind LinklyAI/best-skills — a daily, cross-ecosystem ranking of AI agent skills.

Alternatives in Data

LLM Redteam — LLM red-team corpus runner — fires categorized prompt-injection / jailbreak / system-prompt-leak / data-exfil 4.5k ★
Data Pipelines Pack — Data pipeline development skills 1.6k ★
Check Daily — Check the daily pipeline's state — last run, today's report files, Sonnet quota, recent failures 355 ★

README

best-skills-runner

The data pipeline behind [**LinklyAI/best-skills**](https://github.com/LinklyAI/best-skills) — a daily, cross-ecosystem ranking of AI agent skills.

[![daily](https://github.com/LinklyAI/best-skills/actions/workflows/daily.yml/badge.svg)](https://github.com/LinklyAI/best-skills/actions/workflows/daily.yml) [![probe](https://github.com/LinklyAI/best-skills-runner/actions/workflows/probe.yml/badge.svg)](https://github.com/LinklyAI/best-skills-runner/actions/workflows/probe.yml) [![license](https://img.shields.io/badge/license-MIT-blue.svg)](LICENSE)

Every skills registry only sees its own ecosystem: skills.sh counts Claude/Vercel CLI installs, ClawHub counts OpenClaw downloads, SkillHub counts installs from China — and none of them see social traction. This pipeline collects all of them once a day, merges them into a single entity graph, computes nine rankings, validates the output, and publishes it as open CSV.

Runs on GitHub Actions every day at 01:17 UTC — the scheduled workflow lives in the [data repo](https://github.com/LinklyAI/best-skills/blob/main/.github/workflows/daily.yml), which checks this one out and runs it. Everything the pipeline produces lives in that repo; there is no database and no server.

Pipeline

sources/     collect raw data — skills.sh · ClawHub · SkillHub · GitHub · social buzz
   ↓
rank/        entity resolution (fully-qualified keys, cross-platform links only on
             exact upstream_url evidence) → within-platform log percentiles →
             nine rankings + a five-dimension weighted index
   ↓
publish/     write CSVs to ../best-skills/data/YYYY-MM-DD/{raw,rankings}/
             render the ranking previews into the data repo's READMEs, one per language
   ↓
validate.ts  publication gate — row-count floors, sort monotonicity, rank continuity,
             directory closure, day-over-day drift, sanity anchors
   ↓          (only if everything passes)
publish/     refresh data/latest/ → commit → p