arXiv cs.AI
  • Score 81
  • Official

SkillBuffBench: Measuring What Agent Skills Contribute across Harnesses, Models, Skill Access Modes and Memory

arXiv:2611.22017v1 Announce Type: new Abstract: Agent Skills, file-based packages of procedural knowledge that a coding agent can discover and load on demand, are now shipped by every major agent harness, yet the evidence for how much they help comes almost entirely from settings in which the right Skill is hand-picked for the task and the agent starts from a blank slate. Neither holds in deployme

So what

What this event means by reading role—not a longer recap.

  • BuilderIf you ship agent Skills, this protocol's three-level access factor and memory-on setting mirror real deployment, so watch its results before tuning your Skill
  • ResearcherThe design crosses Skill access, memory, four harnesses and multiple models with paired bootstrap uncertainty, but no results are reported yet in this version.

Score dimensions

Higher total means read first. Each bar is one factor we use to rank the system pool. How we score

RelevanceHow tightly this is about AI.
78
ImpactHow much this could change the field or the market.
55
NoveltyHow new this is versus a recap.
70
CredibilityHow much we trust the source.
92
ActionabilityWhether a reader can do something with it.
45