Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use this when user wants you to walk through (code or text) files in a EDITOR to either explain how some code works, or to show the user what changes you made, etc. You would typically use this repeatedly to show the user your changes or code files one by one, sometimes with specific line-numbers. This way the user is easily able to follow along in their favorite EDITOR as you point at various files possibly at specific line numbers within those files.
.claude/skills/pchalasani-code-walk-thru/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 17% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 45% | 0% |
Depending on which EDITOR the user says they are using, you will use a different "show-file" cli command that shows (in the EDITOR) a specific file, optionally at specific line-number, as in examples below. If no editor specified, you must ask the user which editor they are using.
IMPORTANT: you must walk thru the files ONE BY ONE, and you MUST wait for the user to say something before moving on to the next file, or to same file different line.
code --goto <file_path>:<line_number>pycharm --line <line_number> <file_path>intellij --line <line_number> <file_path>zed path/to/file.md:43vim +42 blah.py
nvim +42 blah.pyIf any of these fail tell the user to install the corresponding CLI tool for their editor.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | pass→pass | 1,810 | 1,865 | +3% | 1 | 1 | 0% | 309 | 585 | +89% | 0 | 0 | — |
case-20 | fail→fail | 6,901 | 9,330 | +35% | 1 | 1 | 0% | 799 | 1,353 | +69% | 0 | 0 | — |
case-01 | fail→pass | 5,075 | 5,456 | +8% | 1 | 1 | 0% | 924 | 1,148 | +24% | 0 | 0 | — |
case-02 | fail→pass | 6,154 | 2,160 | -65% | 1 | 1 | 0% | 1,074 | 687 | -36% | 0 | 0 | — |
case-03 | fail→pass | 4,604 | 4,561 | -1% | 1 | 1 | 0% | 818 | 951 | +16% | 0 | 0 | — |
case-04 | fail→pass | 7,128 | 2,046 | -71% | 1 | 1 | 0% | 570 | 665 | +17% | 0 | 0 | — |
case-05 | fail→pass | 2,695 | 2,037 | -24% | 1 | 1 | 0% | 432 | 628 | +45% | 0 | 0 | — |
case-06 | pass→pass | 3,983 | 2,346 | -41% | 1 | 1 | 0% | 619 | 713 | +15% | 0 | 0 | — |
case-21 | pass→fail | 13,594 | 3,787 | -72% | 1 | 1 | 0% | 3,019 | 877 | -71% | 0 | 0 | — |
case-07 | pass→pass | 2,880 | 2,542 | -12% | 1 | 1 | 0% | 469 | 646 | +38% | 0 | 0 | — |
case-08 | fail→pass | 3,436 | 1,938 | -44% | 1 | 1 | 0% | 577 | 575 | -0% | 0 | 0 | — |
case-09 | fail→pass | 4,493 | 1,780 | -60% | 1 | 1 | 0% | 736 | 535 | -27% | 0 | 0 | — |
case-10 | fail→pass | 4,233 | 1,717 | -59% | 1 | 1 | 0% | 743 | 555 | -25% | 0 | 0 | — |
case-22 | fail→fail | 5,743 | 2,589 | -55% | 1 | 1 | 0% | 960 | 663 | -31% | 0 | 0 | — |
case-11 | fail→pass | 2,298 | 1,932 | -16% | 1 | 1 | 0% | 374 | 634 | +70% | 0 | 0 | — |
case-12 | fail→pass | 6,206 | 1,886 | -70% | 1 | 1 | 0% | 413 | 624 | +51% | 0 | 0 | — |
case-13 | pass→pass | 2,911 | 2,036 | -30% | 1 | 1 | 0% | 503 | 598 | +19% | 0 | 0 | — |
case-14 | pass→pass | 2,240 | 1,871 | -16% | 1 | 1 | 0% | 385 | 568 | +48% | 0 | 0 | — |
case-16 | pass→pass | 6,724 | 2,462 | -63% | 1 | 1 | 0% | 1,282 | 757 | -41% | 0 | 0 | — |
case-17 | pass→pass | 9,542 | 4,205 | -56% | 1 | 1 | 0% | 1,649 | 979 | -41% | 0 | 0 | — |
case-18 | fail→pass | 4,026 | 3,248 | -19% | 1 | 1 | 0% | 585 | 782 | +34% | 0 | 0 | — |
case-19 | fail→pass | 3,734 | 2,146 | -43% | 1 | 1 | 0% | 538 | 583 | +8% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 21 counted toward the lift figure. The other 1 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +50 percentage points is the difference between those two pass rates over the 21 comparable cases. 1 case got worse with the skill loaded, and it is included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.