Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Walk through the Elixir/Phoenix plugin commands, workflow, and features in 6 interactive sections. Use when a new user wants to learn what the plugin offers or needs a refresher on available commands.
.claude/skills/oliver-kriska-phx-intro/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 121% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -6% | 0% |
| case-09 | ✓→✗ | ▼ Worse | -61% | 0% |
| case-07 | ✓→✓ | = Same ✓ | -13% | 0% |
Interactive walkthrough of the Elixir/Phoenix plugin in 6 sections (~5 min).
$ARGUMENTS may contain --section N to jump to a specific section (1-6)references/tutorial-content.md for all section content$ARGUMENTS for --section N flag (1-6)--section N specified, jump directly to that sectionFor each section:
— as visible response text emitted BEFORE the AskUserQuestion call; content composed only in thinking is invisible to the user
AskUserQuestion with options:| N | Title | |---|-------| | 1 | Welcome | | 2 | Core Workflow Commands | | 3 | Knowledge & Safety Net | | 4 | Hooks & Behavioral Rules | | 5 | Init, Review & Gaps | | 6 | Cheat Sheet & Next Steps |
After Section 6 (or when user stops):
You're all set! Try `/skill:phx-plan` with your next feature to see the workflow in action.
Run `/skill:phx-intro --section N` anytime to revisit a specific section.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-07 | pass→pass | 9,453 | 5,182 | -45% | 1 | 1 | 0% | 1,592 | 1,382 | -13% | 0 | 0 | — |
case-01 | fail→fail | 6,809 | 5,425 | -20% | 1 | 1 | 0% | 1,099 | 756 | -31% | 0 | 0 | — |
case-02 | fail→fail | 9,869 | 4,717 | -52% | 1 | 1 | 0% | 1,764 | 791 | -55% | 0 | 0 | — |
case-03 | fail→fail | 9,552 | 5,324 | -44% | 1 | 1 | 0% | 1,637 | 982 | -40% | 0 | 0 | — |
case-04 | fail→fail | 18,274 | 4,240 | -77% | 1 | 1 | 0% | 3,852 | 788 | -80% | 0 | 0 | — |
case-05 | fail→fail | 6,211 | 3,748 | -40% | 1 | 1 | 0% | 884 | 677 | -23% | 0 | 0 | — |
case-06 | fail→fail | 12,225 | 4,637 | -62% | 1 | 1 | 0% | 2,048 | 668 | -67% | 0 | 0 | — |
case-08 | fail→fail | 9,769 | 16,010 | +64% | 1 | 1 | 0% | 1,633 | 961 | -41% | 0 | 0 | — |
case-09 | pass→fail | 11,546 | 4,120 | -64% | 1 | 1 | 0% | 1,843 | 717 | -61% | 0 | 0 | — |
case-10 | fail→pass | 5,890 | 2,252 | -62% | 1 | 1 | 0% | 907 | 914 | +1% | 0 | 0 | — |
case-11 | fail→fail | 15,164 | 4,561 | -70% | 1 | 1 | 0% | 2,808 | 747 | -73% | 0 | 0 | — |
case-12 | fail→fail | 10,999 | 4,616 | -58% | 1 | 1 | 0% | 1,892 | 775 | -59% | 0 | 0 | — |
case-13 | fail→pass | 5,329 | 6,060 | +14% | 1 | 1 | 0% | 725 | 1,601 | +121% | 0 | 0 | — |
case-14 | fail→fail | 7,281 | 2,305 | -68% | 1 | 1 | 0% | 1,114 | 892 | -20% | 0 | 0 | — |
case-15 | fail→fail | 3,912 | 8,789 | +125% | 1 | 1 | 0% | 603 | 976 | +62% | 0 | 0 | — |
case-16 | fail→fail | 8,865 | 6,026 | -32% | 1 | 1 | 0% | 1,337 | 857 | -36% | 0 | 0 | — |
case-17 | fail→fail | 8,523 | 4,130 | -52% | 1 | 1 | 0% | 1,340 | 695 | -48% | 0 | 0 | — |
case-18 | fail→pass | 5,944 | 2,543 | -57% | 1 | 1 | 0% | 995 | 934 | -6% | 0 | 0 | — |
case-19 | fail→fail | 8,254 | 5,739 | -30% | 1 | 1 | 0% | 1,266 | 815 | -36% | 0 | 0 | — |
case-20 | pass→pass | 4,077 | 7,282 | +79% | 1 | 1 | 0% | 750 | 1,726 | +130% | 0 | 0 | — |
case-21 | pass→pass | 12,784 | 11,601 | -9% | 1 | 1 | 0% | 2,155 | 2,555 | +19% | 0 | 0 | — |
case-22 | pass→pass | 11,451 | 16,611 | +45% | 1 | 1 | 0% | 2,422 | 3,886 | +60% | 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 8 counted toward the lift figure. The other 14 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 +9 percentage points is the difference between those two pass rates over the 8 comparable cases. 5 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.