anthropics/skills

Skills from anthropics/skills were tested in 1 task class; 1 more class has a qualifying skill pinned and not run. Results are given as a distribution over classes, never as one number.

Across the classes it was tested in

How many task classes each result fell in. Counts of classes, not scores.
What happenedClassesWhich
Helped on some1Skill authoring

There is no total here and there will not be one. A skill-authoring result is a share of a specification's requirements and a voice result is a blind win rate; they are measured on different task sets against different answer keys, so a single number over both would be a figure neither instrument reported. The numbers are on the skill pages, per class, where each was measured.

Per model

Claude Haiku 4.5

Claude Haiku 4.5: classes by result.
What happenedClassesWhich
Helped clearly1Skill authoring

GPT-5 mini

GPT-5 mini: classes by result.
What happenedClassesWhich
Helped clearly1Skill authoring

Gemini 3.1 Flash Lite

Gemini 3.1 Flash Lite: classes by result.
What happenedClassesWhich
No reliable difference1Skill authoring

Each model is named by the string the vendor actually served, not by the one the run requested.

Coverage, class by class

What was done in each class for this repository. The full grid is on the coverage page.
ClassStateWhat that means
Skill authoringTestedA skill from this repository is pinned in this class and has been run, with the skill loaded and without it, on the same task pairs.
Accessibility auditNo qualifying skillThe selection rule was run over this repository for this class and found nothing that qualifies. The candidates it weighed, and the sentence each was weighed on, are in the cohort record.
Spec writingPinned, not runA skill from this repository qualifies in this class under the published selection rule, and its bytes are committed here, but no model has been run against it. Either it holds a slot the run has not reached, or it qualified and lost the output-format tiebreak while the cohort still capped a class at one external skill.
On-page auditNo qualifying skillThe selection rule was run over this repository for this class and found nothing that qualifies. The candidates it weighed, and the sentence each was weighed on, are in the cohort record.
VoiceNo qualifying skillThe selection rule was run over this repository for this class and found nothing that qualifies. The candidates it weighed, and the sentence each was weighed on, are in the cohort record.
Code reviewNo qualifying skillThe selection rule was run over this repository for this class and found nothing that qualifies. The candidates it weighed, and the sentence each was weighed on, are in the cohort record.
Test writingDeclared emptyThis class has declared its external slot empty and said so. The rule has not been run over this repository for it, and no candidate has been weighed.

Identity pins

Every skill from this repository that holds a slot, and the commit its arm was given.
ClassPathCommitContent hashDate tested
Skill authoringskills/skill-creator3b3fad96af16e51ce07ea7fd2026-08-30

Licence status, per pin

What the licence check found beside each set of pinned bytes, run against that pin's own commit.
Pinned bytesLicenceLicence fileChecked at
skills/skill-creator slotApache-2.0skills/skill-creator/LICENSE.txt3b3fad96af16 2026-08-31
skills/doc-coauthoring rejectedNo licence filenone3b3fad96af16 2026-08-31
skills/brand-guidelines rejectedApache-2.0skills/brand-guidelines/LICENSE.txt3b3fad96af16 2026-08-31

Licence is recorded per pin, not per repository, because a repository-level answer is not always a true one: a repository that carries no licence at its root and licenses each skill directory separately has no single answer, and one of the pins held here has no licence file at all. Each row is a check run against that pin's own commit. A row reading not recorded means no check has been recorded for those bytes, which is a gap rather than a finding; a row reading no licence file means the check ran and found none, which is a finding. Nothing here is inferred from a repository name.

  • skills/skill-creator: Apache-2.0, in a LICENSE.txt inside the skill directory. anthropics/skills carries no licence at its repository root, so this per-skill file is the only licence covering these bytes. The file is byte-identical to the one beside skills/brand-guidelines, blob 4f881c52d1f72f4cfb720e339e2d35c3058d01a9.
  • skills/doc-coauthoring: license: none found; bytes retained in this private repository for measurement only, never rendered
  • skills/brand-guidelines: Apache-2.0, in a LICENSE.txt inside the skill directory. anthropics/skills carries no licence at its repository root, so this per-skill file is the only licence covering these bytes. The file is byte-identical to the one beside skills/skill-creator, blob 4f881c52d1f72f4cfb720e339e2d35c3058d01a9.

Everything on this page is derived at build time from the committed cohort record and the run records. R4 allows the identity pin and nothing else from the artifact repository: there is no description of any skill here, no README text, no star count and no install line.