Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Creates new Claude Code slash commands following best practices. Guides through command structure, naming, arguments, and frontmatter. Use when user wants to create a command, build a slash command, or asks about command best practices.
.claude/skills/aiskillstore-creating-commands/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -14% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 7% | 0% |
| case-10 | ✗→✓ | ▲ Improved | 18% | 0% |
Guides creation of Claude Code slash commands using documented best practices.
For a new command:
For reviewing existing command:
Progress:
- [ ] Gather requirements (purpose, arguments, scope)
- [ ] Choose template (basic, with-args, workflow)
- [ ] Generate command file
- [ ] Validate against checklistAsk user with AskUserQuestion:
| Type | Template | When to Use | |------|----------|-------------| | Basic | templates/basic.md | No arguments, simple prompt | | With Args | templates/with-args.md | Single or multiple arguments | | Workflow | templates/workflow.md | Integrates with skills/agents |
Create in appropriate location:
.claude/commands/ - Project commands (git-tracked)~/.claude/commands/ - Personal commands (your machine only)Run through checklist before finishing:
Validation Checklist:
- [ ] Name: lowercase with hyphens only
- [ ] Name: descriptive, verb-noun format preferred
- [ ] Description: explains what command does
- [ ] Arguments: documented with argument-hint if used
- [ ] Prompt: clear, actionable instructions
- [ ] Tools: allowed-tools declared if neededFormat: verb-noun or action (lowercase, hyphens)
fix-issuereview-prrun-testsoptimizeConstraints:
markdown--- description: Brief explanation shown in help allowed-tools: - Bash(bash:*) - WebSearch argument-hint: "param_name" --- Your command prompt here. Use $ARGUMENTS for all args or $1, $2 for positional.
| Syntax | Usage | Example | |--------|-------|---------| | $ARGUMENTS | All arguments as one string | /cmd foo bar → foo bar | | $1, $2 | Positional arguments | /cmd foo bar → $1=foo, $2=bar | | @file | Include file contents | Review @src/main.ts |
| Field | Purpose | Required | |-------|---------|----------| | description | Shown in help, enables auto-invoke | Recommended | | allowed-tools | Tools the command can use | If using tools | | argument-hint | Documents expected args | If has args | | model | Specific model to use | Optional |
markdown--- description: Run all tests and report failures allowed-tools: - Bash(npm:*) --- Run the test suite and summarize results: 1. Execute `npm test` 2. If failures, show failing tests with context 3. Suggest fixes for common issues
See reference.md for detailed best practices and anti-patterns.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 24,642 | 9,757 | -60% | 1 | 1 | 0% | 3,454 | 1,652 | -52% | 0 | 0 | — |
case-02 | fail→fail | 14,518 | 14,187 | -2% | 1 | 1 | 0% | 2,385 | 3,187 | +34% | 0 | 0 | — |
case-03 | fail→pass | 16,626 | 11,369 | -32% | 1 | 1 | 0% | 3,061 | 2,885 | -6% | 0 | 0 | — |
case-04 | pass→pass | 8,671 | 14,301 | +65% | 1 | 1 | 0% | 1,834 | 2,614 | +43% | 0 | 0 | — |
case-05 | pass→fail | 14,872 | 16,816 | +13% | 1 | 1 | 0% | 2,655 | 4,134 | +56% | 0 | 0 | — |
case-06 | pass→pass | 11,752 | 11,079 | -6% | 1 | 1 | 0% | 1,177 | 1,932 | +64% | 0 | 0 | — |
case-16 | fail→pass | 14,417 | 7,712 | -47% | 1 | 1 | 0% | 1,517 | 1,294 | -15% | 0 | 0 | — |
case-07 | pass→pass | 8,611 | 4,096 | -52% | 1 | 1 | 0% | 1,003 | 1,487 | +48% | 0 | 0 | — |
case-08 | fail→pass | 13,778 | 7,737 | -44% | 1 | 1 | 0% | 1,498 | 1,282 | -14% | 0 | 0 | — |
case-09 | fail→pass | 15,060 | 4,444 | -70% | 1 | 1 | 0% | 1,510 | 1,616 | +7% | 0 | 0 | — |
case-10 | fail→pass | 11,596 | 2,867 | -75% | 1 | 1 | 0% | 1,092 | 1,289 | +18% | 0 | 0 | — |
case-11 | pass→pass | 10,286 | 2,903 | -72% | 1 | 1 | 0% | 878 | 1,353 | +54% | 0 | 0 | — |
case-12 | pass→pass | 14,563 | 10,066 | -31% | 1 | 1 | 0% | 1,543 | 1,712 | +11% | 0 | 0 | — |
case-13 | fail→pass | 15,418 | 7,202 | -53% | 1 | 1 | 0% | 1,730 | 2,156 | +25% | 0 | 0 | — |
case-14 | pass→pass | 10,475 | 7,572 | -28% | 1 | 1 | 0% | 907 | 1,253 | +38% | 0 | 0 | — |
case-15 | pass→pass | 12,958 | 4,940 | -62% | 1 | 1 | 0% | 1,380 | 1,694 | +23% | 0 | 0 | — |
case-17 | pass→pass | 7,819 | 4,591 | -41% | 1 | 1 | 0% | 1,271 | 1,663 | +31% | 0 | 0 | — |
case-18 | pass→pass | 13,988 | 9,642 | -31% | 1 | 1 | 0% | 2,152 | 1,561 | -27% | 0 | 0 | — |
case-19 | pass→pass | 14,293 | 8,242 | -42% | 1 | 1 | 0% | 1,467 | 1,373 | -6% | 0 | 0 | — |
case-20 | pass→pass | 17,796 | 12,683 | -29% | 1 | 1 | 0% | 2,155 | 2,157 | +0% | 0 | 0 | — |
case-21 | pass→pass | 9,579 | 4,757 | -50% | 1 | 1 | 0% | 1,575 | 1,699 | +8% | 0 | 0 | — |
case-22 | fail→pass | 14,779 | 2,012 | -86% | 1 | 1 | 0% | 1,465 | 1,112 | -24% | 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 +27 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.