Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generates high-quality .agent/skills/ directories based on user requirements. Use when the user asks to "create a skill", "build a skill", or formalize a set of instructions into a system skill.
.claude/skills/valec3-generating-skills/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-07 | ✗→✓ | ▲ Improved | 55% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 162% | 0% |
| case-01 | ✗→✓ | ▲ Improved | -12% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -26% | 0% |
You are an expert developer specializing in creating "Skills" for the Antigravity agent environment. Your goal is to generate high-quality, predictable, and efficient .agent/skills/ directories based on user requirements.
Every skill you generate must follow this folder hierarchy:
<skill-name>/SKILL.md (Required: Main logic and instructions)scripts/ (Optional: Helper scripts)examples/ (Optional: Reference implementations)resources/ (Optional: Templates or assets)The SKILL.md must start with YAML frontmatter following these strict rules:
testing-code, managing-databases). Max 64 chars. Lowercase, numbers, and hyphens only. No "claude" or "anthropic" in the name.When writing the body of SKILL.md, adhere to these best practices:
SKILL.md under 500 lines. If more detail is needed, link to secondary files (e.g., [See ADVANCED.md](ADVANCED.md)) only one level deep./ for paths, never \.For complex tasks, include:
--help if they are unsure.When asked to create a skill, output the result in this format:
Path: .agent/skills/[skill-name]/
markdown--- name: [gerund-name] description: [3rd-person description] --- # [Skill Title] ## When to use this skill - [Trigger 1] - [Trigger 2] ## Workflow [Insert checklist or step-by-step guide here] ## Instructions [Specific logic, code snippets, or rules] ## Resources - [Link to scripts/ or resources/] [Supporting Files] (If applicable, provide the content for scripts/ or examples/)
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-07 | fail→pass | 10,028 | 12,860 | +28% | 1 | 1 | 0% | 1,939 | 2,996 | +55% | 0 | 0 | — |
case-13 | fail→pass | 15,651 | 11,179 | -29% | 1 | 1 | 0% | 2,715 | 2,689 | -1% | 0 | 0 | — |
case-06 | fail→pass | 7,211 | 11,974 | +66% | 1 | 1 | 0% | 1,236 | 3,237 | +162% | 0 | 0 | — |
case-01 | fail→pass | 15,150 | 9,425 | -38% | 1 | 1 | 0% | 3,141 | 2,756 | -12% | 0 | 0 | — |
case-02 | fail→pass | 17,725 | 10,453 | -41% | 1 | 1 | 0% | 3,858 | 2,847 | -26% | 0 | 0 | — |
case-03 | fail→pass | 24,332 | 12,628 | -48% | 1 | 1 | 0% | 5,266 | 3,385 | -36% | 0 | 0 | — |
case-04 | fail→fail | 17,416 | 8,907 | -49% | 1 | 1 | 0% | 3,865 | 2,478 | -36% | 0 | 0 | — |
case-05 | fail→pass | 13,870 | 9,639 | -31% | 1 | 1 | 0% | 2,633 | 2,617 | -1% | 0 | 0 | — |
case-08 | fail→pass | 16,717 | 7,773 | -54% | 1 | 1 | 0% | 3,067 | 2,344 | -24% | 0 | 0 | — |
case-09 | fail→pass | 11,779 | 8,186 | -31% | 1 | 1 | 0% | 2,200 | 2,414 | +10% | 0 | 0 | — |
case-10 | fail→fail | 17,876 | 10,302 | -42% | 1 | 1 | 0% | 3,310 | 2,874 | -13% | 0 | 0 | — |
case-11 | fail→pass | 14,632 | 9,346 | -36% | 1 | 1 | 0% | 3,002 | 2,512 | -16% | 0 | 0 | — |
case-12 | fail→pass | 9,681 | 7,812 | -19% | 1 | 1 | 0% | 2,224 | 2,274 | +2% | 0 | 0 | — |
case-14 | fail→fail | 9,188 | 12,580 | +37% | 1 | 1 | 0% | 1,912 | 2,719 | +42% | 0 | 0 | — |
case-15 | fail→pass | 8,014 | 10,158 | +27% | 1 | 1 | 0% | 1,535 | 2,831 | +84% | 0 | 0 | — |
case-16 | fail→pass | 16,277 | 14,474 | -11% | 1 | 1 | 0% | 3,766 | 3,916 | +4% | 0 | 0 | — |
case-17 | fail→pass | 14,044 | 10,314 | -27% | 1 | 1 | 0% | 2,758 | 2,864 | +4% | 0 | 0 | — |
case-18 | fail→pass | 8,469 | 9,039 | +7% | 1 | 1 | 0% | 1,462 | 2,378 | +63% | 0 | 0 | — |
case-19 | fail→pass | 9,759 | 10,403 | +7% | 1 | 1 | 0% | 1,791 | 2,118 | +18% | 0 | 0 | — |
case-20 | pass→pass | 4,170 | 10,934 | +162% | 1 | 1 | 0% | 794 | 2,193 | +176% | 0 | 0 | — |
case-21 | pass→pass | 9,221 | 10,826 | +17% | 1 | 1 | 0% | 1,977 | 3,349 | +69% | 0 | 0 | — |
case-22 | pass→pass | 10,974 | 9,871 | -10% | 1 | 1 | 0% | 2,005 | 2,644 | +32% | 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 +73 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.