NewBest Agent Skills
Daily rankings of agent skills across registries, with Jev judging social relevance, listing quality and category.
Added to Jevfast
NewHow it uses Jev
The daily collection pipeline uses Jev to judge whether social posts concern a skill, whether a skill listing is maintained and legitimate, and which category fits; ranking data retains original per-platform counts alongside the judgments.
What Jev decides
Illustrative Jev requestExample answers · not a recorded Jev response · Source ↗
Question 1 · is skill
YOUR APP
INSTRUCTION
Is listing a real AI agent skill?
STATE
One skill listing: name, registry, publisher/source repository and description.
JEV · NOUL
YesNo
App workflow
- Collect skill signals
Aggregate skill listings and install or growth counts from supported registries, along with relevant social posts.
- Judge relevance and quality
Ask Jev whether posts concern each skill, whether listings are maintained, and which category applies.
- Build daily rankings
Combine judged records with within-platform percentile scores rather than adding incomparable raw counts.
- Publish auditable results
Refresh ranking tables and CSVs while retaining per-platform counts and judgment probabilities.
Why it is interesting
It uses typed semantic judgments to filter noisy cross-registry and social signals, while publishing probabilities and raw platform counts so readers can inspect the basis for rankings.