Explore the live, interactive rankings →

Every skills registry only sees its own ecosystem. skills.sh counts Claude/Vercel CLI installs, ClawHub counts OpenClaw downloads, Tencent SkillHub counts installs from China — and none of them see social buzz. Best Skills merges all of these views into one cross-ecosystem picture: for each skill you can see global installs, China installs, and social mentions side by side. No other ranking does this.

• 9 rankings, refreshed daily
• Raw numbers preserved — every CSV keeps per-platform original counts so anyone can verify or re-rank
• Never adds apples to oranges — cross-platform numbers are shown side by side, ranked by within-platform percentile composite (see methodology)
• Judged by jev, not keyword counts — TypeSafe's decision model checks whether each social post is really about the skill, whether a listing is a real, maintained skill, and which category it belongs to; placeholders, deprecated and harmful listings are left out of the rankings, and every probability is published (see methodology)

Reading it yourself. Each list answers one question (see What each list means); best-100 is the place to start. The tables below show each list's Top 10, the CSVs hold the full Top 100, and linkly.ai/skills renders them for browsing. Cov tells you how many registries a score rests on (A = all three, C = one), and every CSV keeps the per-platform counts next to the score, so any number can be checked against its source.

Giving it to an AI agent. Paste one line into Claude Code, Codex, Cursor, OpenClaw or any agent that can fetch a URL:

Read https://linkly.ai/skills/llms.txt

llms.txt tells the agent where the CSVs are, what every column means, how to go from a question to the right file, and the rules for using the numbers. To make the rankings part of an agent's standing knowledge instead, install this repository as a skill:

npx skills add https://github.com/LinklyAI/best-skills --skill best-skills

Last updated: 2026-10-26 (UTC) · Top 10 preview per list — full Top 100 in the CSVs.

data/
├── YYYY-MM-DD/
│   ├── raw/                # per-platform original counts, untouched
│   │   ├── skills-sh.csv · clawhub.csv · skillhub.csv · github-repos.csv
│   │   └── buzz.csv · x-posts.csv · judgments.csv · …
│   └── rankings/           # the 9 ranking lists computed from raw/
│       ├── best-100.csv · top-installs.csv · trending-7d.csv
│       └── social-buzz.csv · … · rising-stars.csv
├── latest/                 # always a copy of the most recent day
└── index/
    └── first-seen.csv      # cumulative first-seen dates (powers rising-stars)

One folder per day, CSV only. Composite scores are always accompanied by the raw per-platform numbers they were derived from — start from data/latest/ if you just want today's lists.

See docs/methodology.md for data sources, normalization rules, and known limitations.

Install/download counts are collected from public endpoints of skills.sh, ClawHub (per its third-party directory policy; skill pages link back to their canonical ClawHub listing), Tencent SkillHub, and the GitHub API. Social signals come from X, Hacker News (Algolia), Bluesky, and GitHub search. This project is not endorsed by or affiliated with any of these platforms. Counts reflect each platform's own methodology and are never summed across platforms.

Data and documentation are released under CC BY 4.0 — free to use, republish, and build on, commercial use included.

Include this line wherever the data appears:

Data from Best Skills by @BlueeonY — CC BY 4.0

Keep the link to this repository intact, and keep the per-platform attributions listed above. That is the whole requirement — no separate permission is needed.

Built something with it? @BlueeonY would love to see it — not required by the license, just appreciated.

Built by @BlueeonY at Linkly AI — an AI-powered local knowledge base with agent skills support.

Found this useful? A ⭐ helps more people find it.