Filter and clean
5 scenarios, 5 run by an agent. Each run shows its result, time and new tokens.
| Scenario | Codexwithout deft | Codexwith deft, skill installed | Codexwith deft, skill as agents.md | Claude Opus 5.5without deft | Claude Opus 5.5with deft, skill installed | Claude Codewithout deft | Claude Codewith deft |
|---|---|---|---|---|---|---|---|
| Filter rowsFrom gdp.csv keep only the rows for China and save them as china.csv with the header. | pass12s · 5.9k | pass19s · 9.2k | pass16s · 6.2k | pass9s · 8.9k | pass37s · 12.2k | not run | not run |
| Drop columnsKeep only Country Name and Value from gdp.csv and save it as slim.csv. | pass13s · 5.8k | pass16s · 8.7k | not run | pass15s · 8.8k | pass12s · 11.1k | not run | not run |
| Remove duplicate rowsKeep one row per Country Name in gdp.csv and save it as countries-once.csv. | pass14s · 3.7k | pass17s · 9.3k | not run | pass29s · 10.9k | pass25s · 12.4k | not run | not run |
| Largest values firstSort gdp.csv by Value from largest to smallest and save it as ranked.csv. | pass12s · 3.5k | pass19s · 9.9k | not run | pass10s · 9.0k | pass20s · 11.2k | not run | not run |
| Take part of a tableSave the header and first 10 rows of countries.csv as top10.csv. | pass7s · 2.8k | pass16s · 5.0k | pass10s · 5.6k | pass6s · 8.3k | pass17s · 11.0k | not run | not run |