Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run project commands with just. Check for justfile in project root, list available tasks, execute common operations like test, build, lint. Triggers on: run tests, build project, list tasks, check available commands, run script, project commands.
.claude/skills/aiskillstore-task-runner/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -4% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | -32% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | -51% | 0% |
Execute project-specific commands using just, a modern command runner that's simpler than make and works cross-platform.
| Tool | Command | Use For | |------|---------|---------| | just | just | List available recipes | | just | just test | Run specific recipe |
bash# List all available recipes just # Run a recipe just test just build just lint # Run recipe with arguments just deploy production # Run specific recipe from subdirectory just --justfile backend/justfile test
just# Example justfile # Run tests test: pytest tests/ # Build project build: npm run build # Lint code lint: ruff check . eslint src/ # Start development server dev: npm run dev # Clean build artifacts clean: rm -rf dist/ build/ *.egg-info/ # Deploy to environment deploy env: ./scripts/deploy.sh {{env}}
bash# Check if justfile exists just --summary # Show recipe details just --show test # List recipes with descriptions just --list
just to see available project commandsjust testjust buildAlways check for a justfile when entering a new project:
bashjust --list
This shows what commands are available without reading documentation.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 4,785 | 5,757 | +20% | 1 | 1 | 0% | 332 | 592 | +78% | 0 | 0 | — |
case-02 | fail→pass | 5,180 | 1,950 | -62% | 1 | 1 | 0% | 660 | 635 | -4% | 0 | 0 | — |
case-03 | fail→fail | 5,711 | 3,288 | -42% | 1 | 1 | 0% | 859 | 522 | -39% | 0 | 0 | — |
case-04 | pass→pass | 7,010 | 2,750 | -61% | 1 | 1 | 0% | 1,156 | 791 | -32% | 0 | 0 | — |
case-05 | pass→pass | 6,781 | 1,535 | -77% | 1 | 1 | 0% | 1,131 | 557 | -51% | 0 | 0 | — |
case-06 | pass→pass | 9,674 | 2,289 | -76% | 1 | 1 | 0% | 1,723 | 714 | -59% | 0 | 0 | — |
case-07 | pass→pass | 4,552 | 2,158 | -53% | 1 | 1 | 0% | 831 | 614 | -26% | 0 | 0 | — |
case-08 | pass→pass | 7,018 | 1,731 | -75% | 1 | 1 | 0% | 1,206 | 546 | -55% | 0 | 0 | — |
case-09 | pass→pass | 8,226 | 1,565 | -81% | 1 | 1 | 0% | 1,403 | 544 | -61% | 0 | 0 | — |
case-10 | pass→pass | 6,163 | 3,474 | -44% | 1 | 1 | 0% | 1,057 | 967 | -9% | 0 | 0 | — |
case-11 | fail→pass | 8,951 | 5,374 | -40% | 1 | 1 | 0% | 1,327 | 1,193 | -10% | 0 | 0 | — |
case-12 | pass→pass | 6,755 | 4,554 | -33% | 1 | 1 | 0% | 1,137 | 1,101 | -3% | 0 | 0 | — |
case-13 | pass→pass | 31,986 | 3,634 | -89% | 1 | 1 | 0% | 2,547 | 748 | -71% | 0 | 0 | — |
case-14 | pass→pass | 6,703 | 1,634 | -76% | 1 | 1 | 0% | 1,216 | 606 | -50% | 0 | 0 | — |
case-15 | pass→pass | 3,369 | 1,389 | -59% | 1 | 1 | 0% | 550 | 521 | -5% | 0 | 0 | — |
case-16 | fail→pass | 7,425 | 2,725 | -63% | 1 | 1 | 0% | 1,188 | 778 | -35% | 0 | 0 | — |
case-17 | pass→pass | 6,040 | 3,211 | -47% | 1 | 1 | 0% | 981 | 893 | -9% | 0 | 0 | — |
case-18 | pass→pass | 5,649 | 2,327 | -59% | 1 | 1 | 0% | 921 | 720 | -22% | 0 | 0 | — |
case-19 | pass→pass | 4,987 | 2,733 | -45% | 1 | 1 | 0% | 895 | 783 | -13% | 0 | 0 | — |
case-20 | pass→pass | 5,402 | 2,940 | -46% | 1 | 1 | 0% | 841 | 757 | -10% | 0 | 0 | — |
case-21 | pass→pass | 3,244 | 2,188 | -33% | 1 | 1 | 0% | 566 | 733 | +30% | 0 | 0 | — |
case-22 | pass→pass | 6,917 | 3,744 | -46% | 1 | 1 | 0% | 1,223 | 974 | -20% | 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 +14 percentage points is the difference between those two pass rates over the 21 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.