LinklyAI

best-skills: Compare Agent Skill Rankings Across Platforms

Best Skills publishes daily rankings of agent skills using install counts, growth, maintenance, publisher verification, and social activity across several ecosystems. Its CSV data helps developers compare skills without treating platform-specific counts as directly equivalent.

not specifiedOther licenseAgent SkillsAI AgentsAnalytics
Stars
628
Forks
18
Stars, last 17 days
+28

Our take

Health score

36/100

How it is scored
Activity
43
Community
31
Issues
47
Pull requests
21

+28 stars in the last 17 days

  • Fewer contributors than 86% of the projects we track

Overview

Best Skills aggregates public skill listings and activity signals from skills.sh, ClawHub, Tencent SkillHub, GitHub, and social platforms into nine rankings. It addresses a comparison problem: each registry measures activity differently, so the project keeps platform counts separate and ranks skills using normalized comparisons rather than adding unlike counts together.

The project is useful both as a browsable directory at linkly.ai/skills and as an open data source for developers, researchers, and agents. Daily CSV snapshots include source counts alongside rankings, making it possible to inspect or re-rank the data.

Key Features

  • Publishes nine rankings, including overall, installs, weekly trends, social buzz, verified publishers, and rising skills.
  • Aggregates install and download data from skills.sh, ClawHub, and Tencent SkillHub, plus GitHub repository information.
  • Tracks social signals from X, Hacker News, Bluesky, and GitHub search.
  • Preserves original per-platform counts in CSV files instead of combining them into a single raw total.
  • Uses a percentile-based composite score for the overall ranking, with coverage information indicating how many registries contribute.
  • Provides dated CSV snapshots and a latest directory for current data.
  • Documents its data sources, scoring approach, judgments, and limitations in the repository.

Use Cases

  • Developers choosing an agent skill can use the overall ranking as a starting point, then check source counts and coverage before installing.
  • Skill authors can monitor install growth, social attention, and how their listings compare across registries.
  • Researchers and analysts can download CSV snapshots to study skill ecosystem trends or create their own rankings.
  • Agent builders can direct an agent to the project's llms.txt for instructions on finding and interpreting the ranking data.

Getting Started

To add the repository's skill to a compatible agent setup, run:

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

For current rankings, visit linkly.ai/skills or browse the CSVs in the repository. See the README for details on using the datasets and interpreting their columns.

Alternatives

  • awesome-agent-skills: It is a curated directory for browsing agent skills by provider and task, rather than a daily ranking based on usage, growth, and maintenance data.
ProjectLanguageLicenseStarsStatus
best-skillsnot specifiedOther628Not checked yet
awesome-agent-skillsnot specifiedMIT35.2kActive

Considerations

  • Counts come from different platforms and reflect each platform's own methodology. They should not be treated as directly comparable totals.
  • Rankings and collected data change over time. Use dated snapshots when reproducibility matters.
  • Coverage varies by skill and ranking; a composite score may draw on only some of the listed registries.
  • Social mentions and skill quality involve judgment and attribution limits. The README describes the project's filtering and scoring methodology.
  • Repository metadata reports the license as NOASSERTION, while the README states that data and documentation are released under CC BY 4.0. Check the repository's license files before reuse.

Found this useful?

Share it with someone who would like best-skills.

Comparisons