Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Check if specific lines were executed using gcov data
.claude/skills/gadievron-line-execution-checker/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 50% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -56% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -32% | 0% |
Fast tool to check if specific source lines were executed during test runs.
bashg++ -O3 -std=c++17 line_checker.cpp -o line-checker
bash# Single line ./line-checker file.c:42 # Multiple lines ./line-checker file.c:42 main.c:100 util.c:55
file.c:42 EXECUTED (5 times)
main.c:100 NOT EXECUTED
util.c:55 EXECUTED (12 times)Coverage data must exist from prior test run with --coverage flag.
"Was line X of file.c executed?" or "Check if these lines were covered"
.gcda files exist: find . -name "*.gcda" -print -quitg++ -O3 -std=c++17 line_checker.cpp -o line-checker./line-checker file.c:XUser: "Was line 127 in parser.c executed?"
bash./line-checker parser.c:127 # Output: parser.c:127 EXECUTED (3 times)
Response: "Yes, line 127 was executed 3 times during testing."
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 3,323 | 2,915 | -12% | 1 | 1 | 0% | 605 | 910 | +50% | 0 | 0 | — |
case-02 | fail→fail | 3,611 | 4,171 | +16% | 1 | 1 | 0% | 595 | 616 | +4% | 0 | 0 | — |
case-03 | fail→fail | 3,510 | 6,164 | +76% | 1 | 1 | 0% | 606 | 625 | +3% | 0 | 0 | — |
case-08 | fail→pass | 10,050 | 2,220 | -78% | 1 | 1 | 0% | 1,639 | 720 | -56% | 0 | 0 | — |
case-04 | pass→pass | 11,184 | 4,577 | -59% | 1 | 1 | 0% | 2,132 | 1,312 | -38% | 0 | 0 | — |
case-05 | pass→pass | 12,376 | 6,950 | -44% | 1 | 1 | 0% | 2,362 | 1,709 | -28% | 0 | 0 | — |
case-06 | pass→pass | 12,535 | 7,861 | -37% | 1 | 1 | 0% | 2,387 | 2,009 | -16% | 0 | 0 | — |
case-07 | pass→pass | 10,610 | 1,994 | -81% | 1 | 1 | 0% | 1,685 | 720 | -57% | 0 | 0 | — |
case-09 | fail→pass | 3,228 | 1,445 | -55% | 1 | 1 | 0% | 489 | 556 | +14% | 0 | 0 | — |
case-10 | fail→pass | 5,553 | 2,038 | -63% | 1 | 1 | 0% | 939 | 740 | -21% | 0 | 0 | — |
case-11 | pass→pass | 6,639 | 1,713 | -74% | 1 | 1 | 0% | 1,101 | 620 | -44% | 0 | 0 | — |
case-12 | pass→pass | 4,219 | 1,400 | -67% | 1 | 1 | 0% | 656 | 554 | -16% | 0 | 0 | — |
case-17 | pass→pass | 2,545 | 1,815 | -29% | 1 | 1 | 0% | 429 | 662 | +54% | 0 | 0 | — |
case-13 | pass→pass | 4,973 | 1,387 | -72% | 1 | 1 | 0% | 775 | 567 | -27% | 0 | 0 | — |
case-14 | pass→pass | 8,514 | 1,794 | -79% | 1 | 1 | 0% | 1,293 | 630 | -51% | 0 | 0 | — |
case-15 | fail→pass | 5,323 | 1,722 | -68% | 1 | 1 | 0% | 1,082 | 741 | -32% | 0 | 0 | — |
case-16 | pass→pass | 6,773 | 2,081 | -69% | 1 | 1 | 0% | 1,193 | 789 | -34% | 0 | 0 | — |
case-18 | pass→pass | 5,260 | 3,076 | -42% | 1 | 1 | 0% | 853 | 877 | +3% | 0 | 0 | — |
case-19 | fail→pass | 6,843 | 1,734 | -75% | 1 | 1 | 0% | 1,238 | 664 | -46% | 0 | 0 | — |
case-20 | fail→pass | 6,506 | 1,785 | -73% | 1 | 1 | 0% | 1,040 | 683 | -34% | 0 | 0 | — |
case-21 | fail→pass | 8,299 | 1,745 | -79% | 1 | 1 | 0% | 1,548 | 699 | -55% | 0 | 0 | — |
case-22 | fail→pass | 1,974 | 1,847 | -6% | 1 | 1 | 0% | 300 | 658 | +119% | 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 20 counted toward the lift figure. The other 2 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 +41 percentage points is the difference between those two pass rates over the 20 comparable cases.
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.