Multi-step jobs
1 scenario, 1 run by an agent. Each run shows its result, time and new tokens.
| Scenario | Codexwithout deft | Codexwith deft, skill installed | Claude Opus 5.5without deft | Claude Opus 5.5with deft, skill installed | Claude Codewithout deft | Claude Codewith deft |
|---|---|---|---|---|---|---|
| Clean a dataset into a reportgdp.csv mixes countries with regions such as World. Using the country list in countries.csv, make gdp-2015.xlsx with only real countries for 2015, with the columns Country Name, Code and Value, largest first. | pass58s · 15.7k | pass24s · 12.2k | pass29s · 15.9k | pass28s · 14.2k | not run | not run |