Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Clear description of what this skill does and when to use it. Include trigger keywords and contexts inline, e.g. "Use when user wants to X, Y, or Z."
.claude/skills/resciencelab-skill-name/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 4 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 12% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -29% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -13% | 0% |
Brief description of the skill and its purpose.
List any setup requirements:
Example setup:
bashexport SKILL_API_KEY="your_api_key"
How to use the skill quickly:
bashcd <skill_directory> python3 scripts/command.py --option value
bashpython3 scripts/script.py "input"
Output:
Expected output herebashpython3 scripts/script.py "input" --flag --option value
All commands run from the skill directory.
bashpython3 scripts/script1.py --help python3 scripts/script1.py "param1" --option value
bashpython3 scripts/script2.py "param1" "param2"
script1.py - Description of what this script doesscript2.py - Description of what this script doesSymptom: Description of the problem
Solution:
Symptom: Description of the problem
Solution:
See examples/ directory for full workflow examples.
The YAML frontmatter at the top of this file is required:
| Field | Type | Required | Description | |-------|------|----------|-------------| | name | string | ✓ | Unique identifier (kebab-case) | | description | string | ✓ | What the skill does and when to use it. Include trigger keywords and "Use when..." contexts inline. |
skills/your-skill-name/scripts/examples/skills.json with your skill entry| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-03 | fail→pass | 10,669 | 4,979 | -53% | 1 | 1 | 0% | 1,760 | 1,735 | -1% | 0 | 0 | — |
case-01 | fail→pass | 29,736 | 13,104 | -56% | 1 | 1 | 0% | 3,052 | 3,428 | +12% | 0 | 0 | — |
case-02 | fail→pass | 10,062 | 5,428 | -46% | 1 | 1 | 0% | 2,005 | 1,428 | -29% | 0 | 0 | — |
case-04 | pass→pass | 7,924 | 2,128 | -73% | 1 | 1 | 0% | 1,151 | 937 | -19% | 0 | 0 | — |
case-05 | pass→pass | 6,378 | 2,212 | -65% | 1 | 1 | 0% | 1,108 | 1,017 | -8% | 0 | 0 | — |
case-06 | pass→pass | 10,911 | 4,151 | -62% | 1 | 1 | 0% | 1,878 | 1,425 | -24% | 0 | 0 | — |
case-07 | pass→pass | 9,998 | 4,663 | -53% | 1 | 1 | 0% | 1,448 | 1,539 | +6% | 0 | 0 | — |
case-08 | fail→pass | 8,817 | 1,668 | -81% | 1 | 1 | 0% | 1,150 | 910 | -21% | 0 | 0 | — |
case-09 | fail→pass | 9,128 | 3,423 | -63% | 1 | 1 | 0% | 1,401 | 1,221 | -13% | 0 | 0 | — |
case-10 | fail→pass | 9,238 | 2,043 | -78% | 1 | 1 | 0% | 1,392 | 1,018 | -27% | 0 | 0 | — |
case-11 | pass→pass | 13,561 | 9,240 | -32% | 1 | 1 | 0% | 2,221 | 2,301 | +4% | 0 | 0 | — |
case-12 | fail→pass | 5,583 | 2,443 | -56% | 1 | 1 | 0% | 949 | 1,069 | +13% | 0 | 0 | — |
case-13 | fail→pass | 12,289 | 3,153 | -74% | 1 | 1 | 0% | 2,025 | 1,183 | -42% | 0 | 0 | — |
case-14 | fail→pass | 13,395 | 1,988 | -85% | 1 | 1 | 0% | 2,039 | 1,000 | -51% | 0 | 0 | — |
case-15 | fail→fail | 11,759 | 2,894 | -75% | 1 | 1 | 0% | 1,915 | 1,066 | -44% | 0 | 0 | — |
case-16 | fail→pass | 9,205 | 2,181 | -76% | 1 | 1 | 0% | 1,725 | 1,006 | -42% | 0 | 0 | — |
case-17 | pass→pass | 5,606 | 1,986 | -65% | 1 | 1 | 0% | 768 | 934 | +22% | 0 | 0 | — |
case-18 | pass→pass | 11,230 | 1,348 | -88% | 1 | 1 | 0% | 1,981 | 844 | -57% | 0 | 0 | — |
case-19 | fail→pass | 13,255 | 3,896 | -71% | 1 | 1 | 0% | 2,208 | 1,302 | -41% | 0 | 0 | — |
case-20 | pass→pass | 8,308 | 8,602 | +4% | 1 | 1 | 0% | 1,765 | 2,494 | +41% | 0 | 0 | — |
case-21 | pass→pass | 12,046 | 7,700 | -36% | 1 | 1 | 0% | 2,485 | 1,936 | -22% | 0 | 0 | — |
case-22 | fail→pass | 9,139 | 8,149 | -11% | 1 | 1 | 0% | 1,762 | 2,322 | +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 +55 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.