Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate .shellcheckrc configuration with appropriate rules, exclusions, and severity settings for shell script linting.
.claude/skills/a5c-ai-shellcheck-config-generator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -9% | 0% |
| case-21 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-22 | ✗→✓ | ▲ Improved | -22% | 0% |
| case-01 | ✓→✓ | ＝ Same ✓ | 24% | 0% |
Generate ShellCheck configuration for shell script linting.
Invoke this skill when you need to:
ini# .shellcheckrc - ShellCheck configuration # https://www.shellcheck.net/wiki/ # Default shell dialect (sh, bash, dash, ksh) shell=bash # Source path for sourced files source-path=SCRIPTDIR source-path=lib/ # External sources (follow source commands) external-sources=true # Severity level: error, warning, info, style severity=warning # Disable specific warnings # SC1090: Can't follow non-constant source disable=SC1090 # SC1091: Not following (sourced file not found) disable=SC1091 # SC2034: Variable appears unused (often used in sourced scripts) # disable=SC2034 # SC2086: Double quote to prevent globbing (sometimes intentional) # disable=SC2086 # SC2155: Declare and assign separately # disable=SC2155 # Enable optional checks # Check for unquoted variables that could be empty enable=check-unassigned-uppercase # Check for extra masking of return values enable=check-extra-masked-returns
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 6,992 | 5,242 | -25% | 1 | 1 | 0% | 1,300 | 1,617 | +24% | 0 | 0 | — |
case-02 | fail→fail | 8,223 | 7,946 | -3% | 1 | 1 | 0% | 1,597 | 1,987 | +24% | 0 | 0 | — |
case-03 | fail→pass | 10,202 | 8,126 | -20% | 1 | 1 | 0% | 1,874 | 2,106 | +12% | 0 | 0 | — |
case-04 | fail→fail | 9,075 | 8,361 | -8% | 1 | 1 | 0% | 1,761 | 1,945 | +10% | 0 | 0 | — |
case-05 | fail→fail | 9,754 | 6,546 | -33% | 1 | 1 | 0% | 1,528 | 1,776 | +16% | 0 | 0 | — |
case-06 | fail→fail | 8,931 | 6,065 | -32% | 1 | 1 | 0% | 1,576 | 1,623 | +3% | 0 | 0 | — |
case-07 | fail→fail | 13,767 | 12,917 | -6% | 1 | 1 | 0% | 2,462 | 2,841 | +15% | 0 | 0 | — |
case-08 | fail→fail | 16,400 | 15,214 | -7% | 1 | 1 | 0% | 2,712 | 3,303 | +22% | 0 | 0 | — |
case-09 | fail→pass | 6,918 | 2,713 | -61% | 1 | 1 | 0% | 1,211 | 1,103 | -9% | 0 | 0 | — |
case-10 | fail→fail | 8,395 | 3,634 | -57% | 1 | 1 | 0% | 1,589 | 1,253 | -21% | 0 | 0 | — |
case-11 | fail→fail | 3,650 | 2,902 | -20% | 1 | 1 | 0% | 573 | 1,038 | +81% | 0 | 0 | — |
case-12 | fail→fail | 4,165 | 2,822 | -32% | 1 | 1 | 0% | 652 | 1,010 | +55% | 0 | 0 | — |
case-13 | fail→fail | 3,602 | 2,076 | -42% | 1 | 1 | 0% | 609 | 925 | +52% | 0 | 0 | — |
case-14 | pass→pass | 5,003 | 2,046 | -59% | 1 | 1 | 0% | 720 | 917 | +27% | 0 | 0 | — |
case-15 | fail→fail | 6,836 | 1,489 | -78% | 1 | 1 | 0% | 1,215 | 810 | -33% | 0 | 0 | — |
case-16 | fail→fail | 10,096 | 9,072 | -10% | 1 | 1 | 0% | 1,595 | 2,270 | +42% | 0 | 0 | — |
case-17 | fail→fail | 11,450 | 12,113 | +6% | 1 | 1 | 0% | 2,168 | 2,580 | +19% | 0 | 0 | — |
case-18 | fail→fail | 12,609 | 9,659 | -23% | 1 | 1 | 0% | 1,907 | 2,290 | +20% | 0 | 0 | — |
case-19 | fail→fail | 11,827 | 10,760 | -9% | 1 | 1 | 0% | 2,029 | 2,270 | +12% | 0 | 0 | — |
case-20 | fail→fail | 12,868 | 11,581 | -10% | 1 | 1 | 0% | 2,332 | 2,403 | +3% | 0 | 0 | — |
case-21 | fail→pass | 8,686 | 4,814 | -45% | 1 | 1 | 0% | 1,694 | 1,472 | -13% | 0 | 0 | — |
case-22 | fail→pass | 10,860 | 5,229 | -52% | 1 | 1 | 0% | 2,130 | 1,664 | -22% | 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. The headline lift of +18 percentage points is the difference between those two pass rates over the 22 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.