Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate SKILL.md files with proper YAML frontmatter, capabilities documentation, and usage examples following Babysitter SDK conventions.
.claude/skills/a5c-ai-skill-generator/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 4% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 14% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 0% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -5% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 17% | 0% |
You are skill-generator - a specialized skill for generating Claude Code skill files (SKILL.md) with proper structure, frontmatter, and documentation.
This skill generates complete SKILL.md files including:
yaml--- name: skill-name description: Comprehensive skill description allowed-tools: Tool1 Tool2 Tool3 metadata: author: babysitter-sdk version: "1.0.0" category: category-name backlog-id: SK-XX-NNN ---
# skill-nameGenerate valid YAML frontmatter:
yaml--- name: new-skill description: Skill for doing X, Y, and Z allowed-tools: Read Write Edit Glob Grep metadata: author: babysitter-sdk version: "1.0.0" category: analysis backlog-id: SK-XX-001 ---
Document each capability clearly:
markdown### 1. Capability Name Description of what this capability does. Example usage: \`\`\`javascript { task: 'Example task', parameters: { key: 'value' } } \`\`\` Expected output: \`\`\`json { "result": "output" } \`\`\`
Select appropriate tools:
| Tool | Use Case | |------|----------| | Read | Reading existing files | | Write | Creating new files | | Edit | Modifying existing files | | Glob | Finding files by pattern | | Grep | Searching file contents | | Bash | Running shell commands | | WebFetch | Fetching web content | | WebSearch | Searching the web |
Generate practical examples:
markdown## Usage ### Basic Usage \`\`\`javascript { task: 'Do something', input: { key: 'value' } } \`\`\` ### Advanced Usage \`\`\`javascript { task: 'Do something complex', input: { key: 'value' }, options: { advanced: true } } \`\`\`
json{ "skillPath": "path/to/skill-name/SKILL.md", "frontmatter": { "name": "skill-name", "description": "...", "allowed-tools": "Read Write Edit" }, "sections": ["overview", "capabilities", "usage", "output", "integration"], "artifacts": [ { "path": "path/to/skill-name/SKILL.md", "type": "markdown", "label": "Skill definition" } ] }
This skill integrates with:
skill-creation.js - Primary skill generationphase6-create-skills-agents.js - Batch skill creationspecialization-creation.js - Full specialization workflow| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 24,014 | 26,779 | +12% | 1 | 1 | 0% | 3,572 | 3,730 | +4% | 0 | 0 | — |
case-02 | fail→pass | 21,936 | 15,763 | -28% | 1 | 1 | 0% | 3,531 | 4,023 | +14% | 0 | 0 | — |
case-03 | fail→pass | 23,547 | 19,510 | -17% | 1 | 1 | 0% | 3,955 | 3,955 | 0% | 0 | 0 | — |
case-04 | pass→fail | 15,600 | 16,092 | +3% | 1 | 1 | 0% | 3,272 | 4,421 | +35% | 0 | 0 | — |
case-05 | pass→pass | 4,707 | 6,502 | +38% | 1 | 1 | 0% | 732 | 1,839 | +151% | 0 | 0 | — |
case-06 | pass→pass | 13,644 | 13,015 | -5% | 1 | 1 | 0% | 3,107 | 3,974 | +28% | 0 | 0 | — |
case-07 | fail→pass | 8,016 | 3,090 | -61% | 1 | 1 | 0% | 1,644 | 1,554 | -5% | 0 | 0 | — |
case-08 | fail→pass | 8,799 | 4,389 | -50% | 1 | 1 | 0% | 1,398 | 1,633 | +17% | 0 | 0 | — |
case-09 | pass→pass | 6,439 | 5,574 | -13% | 1 | 1 | 0% | 1,016 | 1,712 | +69% | 0 | 0 | — |
case-10 | fail→pass | 5,094 | 5,978 | +17% | 1 | 1 | 0% | 727 | 1,838 | +153% | 0 | 0 | — |
case-11 | fail→pass | 11,844 | 11,920 | +1% | 1 | 1 | 0% | 2,047 | 2,801 | +37% | 0 | 0 | — |
case-12 | fail→pass | 22,015 | 17,220 | -22% | 1 | 1 | 0% | 3,362 | 3,869 | +15% | 0 | 0 | — |
case-13 | fail→pass | 13,878 | 9,227 | -34% | 1 | 1 | 0% | 2,271 | 2,475 | +9% | 0 | 0 | — |
case-14 | fail→pass | 7,397 | 9,936 | +34% | 1 | 1 | 0% | 1,087 | 2,838 | +161% | 0 | 0 | — |
case-15 | pass→pass | 10,852 | 10,564 | -3% | 1 | 1 | 0% | 1,728 | 2,585 | +50% | 0 | 0 | — |
case-16 | fail→pass | 8,799 | 2,893 | -67% | 1 | 1 | 0% | 1,320 | 1,409 | +7% | 0 | 0 | — |
case-17 | fail→pass | 7,264 | 4,149 | -43% | 1 | 1 | 0% | 1,394 | 1,859 | +33% | 0 | 0 | — |
case-18 | pass→pass | 12,927 | 12,410 | -4% | 1 | 1 | 0% | 2,608 | 3,357 | +29% | 0 | 0 | — |
case-19 | fail→pass | 4,403 | 3,165 | -28% | 1 | 1 | 0% | 600 | 1,569 | +162% | 0 | 0 | — |
case-20 | pass→pass | 11,960 | 6,983 | -42% | 1 | 1 | 0% | 1,734 | 1,939 | +12% | 0 | 0 | — |
case-21 | fail→pass | 10,075 | 9,494 | -6% | 1 | 1 | 0% | 1,478 | 2,173 | +47% | 0 | 0 | — |
case-22 | fail→pass | 25,296 | 14,182 | -44% | 1 | 1 | 0% | 3,743 | 3,674 | -2% | 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 +64 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.