Six Authors Supply Three Quarters of the Directory
Directories are usually described as ecosystems. Ours is better described as a shelf stocked by a dozen publishers.
We track the source repository for every skill, which gives us an author count. Across 765 skills:
| Rank | Author | Skills |
|---|---|---|
| 1 | aaron-he-zhu | 119 |
| 2 | calesthio | 89 |
| 3 | gamedev-skills | 74 |
| 4 | phuryn | 68 |
| 5 | Jeffallan | 67 |
| 6 | agiprolabs | 67 |
| 7 | seranking | 32 |
| 8 | addyosmani | 24 |
| 9 | Jakeschincariol | 22 |
| 10 | JimLiu | 21 |
65 distinct authors. The top six supply 535 of 765 skills — 70%. The top ten supply 76%.
The median author has three skills.
This is not what an ecosystem looks like
An ecosystem has a long tail of independents. This has a head of large catalogues and a short tail. The six largest accounts are not hobbyists shipping a skill they needed: they are systematic publishers, in several cases with a hundred or more skills in one repository.
That produces a specific texture in the directory:
- Category averages become author averages. We made this point with nine categories and 591 skills; it holds more strongly at 765. The "video skills average 8.72" finding is substantially a statement about one author's catalogue.
- Quality is correlated within an author. Skills from the same catalogue share conventions, structure, and often the same evaluator judgment. Four of the skills scoring below 7.0 come from a single author's family.
- A single import can move the whole distribution. The 119-skill catalogue is 15.6% of the directory. If it improved by half a point, the site-wide average would move by roughly 0.08 — visible, from one author's work.
What this means for the numbers we publish
Category averages, score distributions, and "the state of the ecosystem" stats are all, to a first approximation, weighted averages over a dozen catalogues. They describe the supply, not the field.
That is worth stating explicitly because it changes how to read them. When we say "marketing skills average 9.30 on trigger quality," a reader reasonably imagines hundreds of independent authors converging on a convention. What actually happened is that one or two large marketing catalogues were scored, and their conventions dominate the average.
The flip side
For a new author, the concentration is an opening rather than a barrier. The directory is not saturated with independent work — there are 65 authors for 765 skills, and the median publishes three. A single genuinely good skill can be one of the few independent entries in its category, which is a more distinctive position than being the 120th skill in a large catalogue.
The supply is concentrated. The demand is flat. Those two facts together are the most interesting thing in the data.
