Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Use when you need to install the embedded project commands into command directories (.github/commands, .claude/commands, .cursor/command, .codex/commands), selecting the destination interactively and copying the embedded command definitions from project assets. This should trigger for requests such as Install embedded commands; Bootstrap .cursor/command; Bootstrap .claude/commands; Copy project commands; Install project command suite. Part of Plinth Toolkit
.claude/skills/jabrena-004-commands-installation/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-02 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-03 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -58% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 25% | 0% |
Install a predefined set of embedded project commands from repository assets into a user-selected target directory. This is an interactive skill.
What is covered in this Skill?
.github/commands, .claude/commands, .cursor/commands, or .codex/commands)assets/commandsThis skill installs only the embedded project commands bundle and must ask for destination before writing files.
assets/commands/For detailed guidance, examples, and constraints, see references/004-commands-installation.md.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | pass→pass | 11,307 | 2,509 | -78% | 1 | 1 | 0% | 1,639 | 621 | -62% | 0 | 0 | — |
case-01 | fail→pass | 6,739 | 2,105 | -69% | 1 | 1 | 0% | 1,031 | 536 | -48% | 0 | 0 | — |
case-02 | fail→pass | 3,416 | 2,910 | -15% | 1 | 1 | 0% | 480 | 674 | +40% | 0 | 0 | — |
case-03 | fail→pass | 6,674 | 2,286 | -66% | 1 | 1 | 0% | 898 | 632 | -30% | 0 | 0 | — |
case-04 | fail→pass | 11,268 | 3,205 | -72% | 1 | 1 | 0% | 1,853 | 786 | -58% | 0 | 0 | — |
case-20 | pass→pass | 3,085 | 2,839 | -8% | 1 | 1 | 0% | 413 | 673 | +63% | 0 | 0 | — |
case-05 | fail→pass | 5,201 | 6,114 | +18% | 1 | 1 | 0% | 799 | 995 | +25% | 0 | 0 | — |
case-06 | fail→pass | 10,676 | 1,830 | -83% | 1 | 1 | 0% | 1,917 | 598 | -69% | 0 | 0 | — |
case-07 | fail→pass | 8,644 | 4,356 | -50% | 1 | 1 | 0% | 1,446 | 949 | -34% | 0 | 0 | — |
case-08 | pass→pass | 9,629 | 3,040 | -68% | 1 | 1 | 0% | 1,498 | 752 | -50% | 0 | 0 | — |
case-09 | pass→pass | 10,803 | 4,341 | -60% | 1 | 1 | 0% | 1,800 | 938 | -48% | 0 | 0 | — |
case-10 | fail→pass | 8,778 | 2,774 | -68% | 1 | 1 | 0% | 1,314 | 649 | -51% | 0 | 0 | — |
case-11 | pass→pass | 8,169 | 2,466 | -70% | 1 | 1 | 0% | 1,317 | 657 | -50% | 0 | 0 | — |
case-12 | fail→pass | 5,620 | 3,354 | -40% | 1 | 1 | 0% | 867 | 701 | -19% | 0 | 0 | — |
case-13 | fail→pass | 12,177 | 2,198 | -82% | 1 | 1 | 0% | 1,887 | 610 | -68% | 0 | 0 | — |
case-15 | pass→pass | 2,264 | 1,398 | -38% | 1 | 1 | 0% | 449 | 502 | +12% | 0 | 0 | — |
case-16 | fail→pass | 6,241 | 3,118 | -50% | 1 | 1 | 0% | 884 | 745 | -16% | 0 | 0 | — |
case-17 | pass→pass | 10,687 | 4,949 | -54% | 1 | 1 | 0% | 2,093 | 1,062 | -49% | 0 | 0 | — |
case-18 | pass→pass | 9,443 | 2,172 | -77% | 1 | 1 | 0% | 1,379 | 619 | -55% | 0 | 0 | — |
case-19 | fail→pass | 11,879 | 2,466 | -79% | 1 | 1 | 0% | 2,078 | 749 | -64% | 0 | 0 | — |
case-21 | pass→pass | 2,432 | 2,939 | +21% | 1 | 1 | 0% | 417 | 658 | +58% | 0 | 0 | — |
case-22 | pass→pass | 2,853 | 3,307 | +16% | 1 | 1 | 0% | 440 | 738 | +68% | 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.