List your company

Skill invocation in LLM agents on 30 artifact-rich tasks (test)

18Data Computation WinsQuality-First
4.487.9911.515.01Mar 2, 2026

Evaluation results

MethodLinks
Quality-FirstEcosystem Size (|S|)=2002026.0318151266117617702031100
Quality-FirstEcosystem Size (|S|)=200K2026.0315721554133894111815100
Quality-FirstEcosystem Size (|S|)=1K2026.0314461176132991052022100
Efficiency-FirstEcosystem Size (|S|)=2002026.03132971076991185143758.5
w/ Full Pool.Ecosystem Size (|S|)=1K2026.03111129510142810410441648.1
Efficiency-FirstEcosystem Size (|S|)=200K2026.031176127514461059162689
Simplicity-FirstEcosystem Size (|S|)=2002026.03104107988791266133853.6
Efficiency-FirstEcosystem Size (|S|)=1K2026.0310410115815188106106876.1
Simplicity-FirstEcosystem Size (|S|)=1K2026.03931210597314987116756.3
VanillaEcosystem Size (|S|)=1K2026.03921356136216561344160
Simplicity-FirstEcosystem Size (|S|)=200K2026.038313115883131257143756
w/ Full Pool.Ecosystem Size (|S|)=200K2026.0374134515122108313461417.2
VanillaEcosystem Size (|S|)=2002026.03651345155811251725170
VanillaEcosystem Size (|S|)=200K2026.03651338136216105904200
w/ Full Pool.Ecosystem Size (|S|)=2002026.035415118513383417241824.3