Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Expert Go code reviewer specializing in idiomatic Go, concurrency patterns, error handling, and performance. Use for all Go code changes. MUST BE USED for Go projects.
.claude/skills/kunanonj-agent-go-reviewer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-09 | ✗→✓ | ▲ Improved | 9% | 0% |
| case-12 | ✗→✓ | ▲ Improved | 24% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 20% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 13% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 13% | 0% |
You are a senior Go code reviewer ensuring high standards of idiomatic Go and best practices.
When invoked:
git diff -- '*.go' to see recent Go file changesgo vet ./... and staticcheck ./... if available.go filesdatabase/sql queriesos/execfilepath.Clean + prefix checkInsecureSkipVerify: true_ to discard errorsreturn err without fmt.Errorf("context: %w", err)errors.Is(err, target) not err == targetcontext.Context)defer mu.Unlock()if/else instead of early returnstrings.Buildermake([]T, 0, cap)ctx context.Context should be first parameterbashgo vet ./... staticcheck ./... golangci-lint run go build -race ./... go test -race ./... govulncheck ./...
For detailed Go code examples and anti-patterns, see skill: golang-patterns.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 11,196 | 10,188 | -9% | 1 | 1 | 0% | 2,207 | 2,493 | +13% | 0 | 0 | — |
case-01 | fail→fail | 16,206 | 13,979 | -14% | 1 | 1 | 0% | 1,648 | 2,106 | +28% | 0 | 0 | — |
case-02 | fail→fail | 5,339 | 2,193 | -59% | 1 | 1 | 0% | 920 | 1,072 | +17% | 0 | 0 | — |
case-03 | fail→fail | 4,046 | 6,537 | +62% | 1 | 1 | 0% | 657 | 1,905 | +190% | 0 | 0 | — |
case-04 | pass→pass | 8,561 | 5,257 | -39% | 1 | 1 | 0% | 1,421 | 1,928 | +36% | 0 | 0 | — |
case-06 | fail→fail | 13,131 | 19,460 | +48% | 1 | 1 | 0% | 2,520 | 2,685 | +7% | 0 | 0 | — |
case-07 | pass→pass | 8,010 | 5,373 | -33% | 1 | 1 | 0% | 1,545 | 1,811 | +17% | 0 | 0 | — |
case-08 | pass→pass | 10,431 | 7,094 | -32% | 1 | 1 | 0% | 2,024 | 2,114 | +4% | 0 | 0 | — |
case-09 | fail→pass | 12,092 | 7,572 | -37% | 1 | 1 | 0% | 1,925 | 2,089 | +9% | 0 | 0 | — |
case-10 | pass→pass | 9,084 | 6,612 | -27% | 1 | 1 | 0% | 1,597 | 2,098 | +31% | 0 | 0 | — |
case-11 | pass→pass | 5,850 | 3,535 | -40% | 1 | 1 | 0% | 1,083 | 1,477 | +36% | 0 | 0 | — |
case-12 | fail→pass | 7,499 | 5,043 | -33% | 1 | 1 | 0% | 1,468 | 1,817 | +24% | 0 | 0 | — |
case-13 | fail→pass | 10,082 | 7,882 | -22% | 1 | 1 | 0% | 1,759 | 2,113 | +20% | 0 | 0 | — |
case-14 | fail→fail | 7,803 | 4,309 | -45% | 1 | 1 | 0% | 1,336 | 1,652 | +24% | 0 | 0 | — |
case-15 | pass→pass | 5,577 | 3,547 | -36% | 1 | 1 | 0% | 862 | 1,463 | +70% | 0 | 0 | — |
case-16 | pass→pass | 10,551 | 6,067 | -42% | 1 | 1 | 0% | 1,978 | 1,939 | -2% | 0 | 0 | — |
case-17 | pass→pass | 7,917 | 6,179 | -22% | 1 | 1 | 0% | 1,468 | 1,937 | +32% | 0 | 0 | — |
case-18 | fail→pass | 8,704 | 4,554 | -48% | 1 | 1 | 0% | 1,373 | 1,545 | +13% | 0 | 0 | — |
case-19 | pass→pass | 12,226 | 6,956 | -43% | 1 | 1 | 0% | 1,916 | 1,985 | +4% | 0 | 0 | — |
case-20 | pass→pass | 23,755 | 14,216 | -40% | 1 | 1 | 0% | 2,112 | 3,259 | +54% | 0 | 0 | — |
case-21 | pass→pass | 11,687 | 11,977 | +2% | 1 | 1 | 0% | 2,089 | 2,950 | +41% | 0 | 0 | — |
case-22 | pass→pass | 14,209 | 8,406 | -41% | 1 | 1 | 0% | 2,847 | 2,919 | +3% | 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.
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.