Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Transform a brief or prompt into a structured, production-ready prompt via prompt-optimizer. File or text mode.
.claude/skills/automagik-dev-refine/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 193% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 75% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 140% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 97% | 0% |
| case-07 | ✓→✗ | ▼ Worse | -57% | 0% |
Runtime syntax: invoke the plugin copy through the active runtime's owner-qualified skill selector; use a bare selector only when intentionally selecting a user-tier copy (a separately installed personal copy; Genie no longer seeds this tier). Cross-skill prose below uses bare names as portable semantic routes; the orchestrator resolves the selector for the active runtime.
Transform any brief, draft, or one-liner into a production-ready structured prompt.
refine with text or a file path@ → file mode; otherwise → text mode.prompts/optimizer.md (relative to this skill's directory — skills/refine/prompts/optimizer.md). Its full contents are the refiner's system prompt.prompts/optimizer.md; user message = the input. Single turn./tmp/prompts/<slug>.md.| | File mode | Text mode | |---|-----------|-----------| | Invocation | refine @path/to/file.md | refine <text> | | Input | file contents (strip @ prefix) | the raw argument | | Output | overwrite the same file | /tmp/prompts/<slug>.md (mkdir -p /tmp/prompts/ first) | | Report | the updated file path | the created file path |
Slug: <unix-timestamp>-<word1>-<word2>-<word3> — first 3 words, lowercased, hyphenated. Example: 1708190400-fix-auth-bug.
The refiner is a single-turn subagent: input in, optimized prompt out.
prompts/optimizer.md — passed whole, never summarized./tmp/prompts/.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→fail | 8,881 | 6,964 | -22% | 1 | 1 | 0% | 1,335 | 938 | -30% | 0 | 0 | — |
case-15 | fail→pass | 5,966 | 11,933 | +100% | 1 | 1 | 0% | 898 | 2,635 | +193% | 0 | 0 | — |
case-01 | fail→fail | 10,218 | 5,550 | -46% | 1 | 1 | 0% | 1,631 | 984 | -40% | 0 | 0 | — |
case-02 | fail→fail | 5,200 | 7,625 | +47% | 1 | 1 | 0% | 324 | 1,116 | +244% | 0 | 0 | — |
case-03 | fail→pass | 10,938 | 14,845 | +36% | 1 | 1 | 0% | 1,623 | 2,841 | +75% | 0 | 0 | — |
case-04 | fail→fail | 10,494 | 7,251 | -31% | 1 | 1 | 0% | 1,690 | 1,820 | +8% | 0 | 0 | — |
case-05 | fail→fail | 5,830 | 10,123 | +74% | 1 | 1 | 0% | 192 | 1,002 | +422% | 0 | 0 | — |
case-06 | fail→pass | 7,685 | 15,905 | +107% | 1 | 1 | 0% | 1,246 | 2,992 | +140% | 0 | 0 | — |
case-07 | pass→fail | 12,054 | 4,161 | -65% | 1 | 1 | 0% | 1,954 | 848 | -57% | 0 | 0 | — |
case-08 | fail→fail | 8,397 | 7,851 | -7% | 1 | 1 | 0% | 1,472 | 972 | -34% | 0 | 0 | — |
case-09 | pass→pass | 6,251 | 7,134 | +14% | 1 | 1 | 0% | 1,046 | 1,770 | +69% | 0 | 0 | — |
case-11 | pass→pass | 7,212 | 9,033 | +25% | 1 | 1 | 0% | 1,125 | 2,130 | +89% | 0 | 0 | — |
case-12 | fail→fail | 5,683 | 6,914 | +22% | 1 | 1 | 0% | 815 | 1,057 | +30% | 0 | 0 | — |
case-13 | fail→pass | 6,582 | 9,377 | +42% | 1 | 1 | 0% | 1,115 | 2,196 | +97% | 0 | 0 | — |
case-14 | pass→pass | 6,905 | 18,215 | +164% | 1 | 1 | 0% | 631 | 1,895 | +200% | 0 | 0 | — |
case-16 | pass→fail | 7,681 | 9,026 | +18% | 1 | 1 | 0% | 705 | 1,194 | +69% | 0 | 0 | — |
case-17 | fail→fail | 12,531 | 5,704 | -54% | 1 | 1 | 0% | 2,292 | 952 | -58% | 0 | 0 | — |
case-18 | pass→pass | 10,237 | 13,366 | +31% | 1 | 1 | 0% | 1,591 | 2,441 | +53% | 0 | 0 | — |
case-19 | pass→fail | 12,924 | 5,667 | -56% | 1 | 1 | 0% | 2,071 | 938 | -55% | 0 | 0 | — |
case-20 | pass→pass | 7,551 | 5,623 | -26% | 1 | 1 | 0% | 1,181 | 1,474 | +25% | 0 | 0 | — |
case-21 | pass→pass | 2,541 | 4,225 | +66% | 1 | 1 | 0% | 320 | 1,340 | +319% | 0 | 0 | — |
case-22 | pass→pass | 6,850 | 6,297 | -8% | 1 | 1 | 0% | 1,010 | 1,586 | +57% | 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 12 counted toward the lift figure. The other 10 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 +5 percentage points is the difference between those two pass rates over the 12 comparable cases. 3 cases got worse with the skill loaded, and they are 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.