Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create a new Cursor plugin scaffold with a valid manifest, component directories, and marketplace wiring. Use when starting a new plugin or adding a plugin to a multi-plugin repository.
.claude/skills/kunanonj-cursor-plugin-create-create-plugin-scaffold/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-14 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-01 | ✗→✓ | ▲ Improved | 1% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -19% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 26% | 0% |
| case-04 | ✗→✓ | ▲ Improved | 23% | 0% |
You need to create a new Cursor plugin from scratch and make it ready for local use or marketplace submission.
rules, skills, agents, commands, hooks, mcpServers)single-plugin or multi-plugin marketplace)By default, create the plugin inside the user's local plugin directory:
~/.cursor/plugins/local/<plugin-name>/This path makes the plugin immediately available to Cursor without any install step. If the user explicitly asks to create the plugin elsewhere (e.g. inside an existing repo or a specific directory), respect that choice instead.
~/.cursor/plugins/local/<plugin-name>/.cursor-plugin/plugin.jsonREADME.mdLICENSECHANGELOG.mdplugin.json:nameversion, description, author, license, keywords.mdc with description, alwaysApply, optional globsskills/<skill-name>/SKILL.md with name, descriptionagents/*.md with name, descriptioncommands/*.(md|txt) with name, description.cursor-plugin/marketplace.json, add plugin entry:namesourcedescription, keywords, category, tags)~/.cursor/plugins/local/<plugin-name>/ unless the user provides a different path.plugin.json~/.cursor/plugins/local/ and ready for use| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | fail→pass | 15,098 | 9,430 | -38% | 1 | 1 | 0% | 3,113 | 2,849 | -8% | 0 | 0 | — |
case-01 | fail→pass | 9,079 | 11,712 | +29% | 1 | 1 | 0% | 2,111 | 2,141 | +1% | 0 | 0 | — |
case-02 | fail→pass | 16,047 | 9,964 | -38% | 1 | 1 | 0% | 3,828 | 3,095 | -19% | 0 | 0 | — |
case-03 | fail→pass | 8,683 | 8,798 | +1% | 1 | 1 | 0% | 2,126 | 2,679 | +26% | 0 | 0 | — |
case-04 | fail→pass | 8,596 | 8,460 | -2% | 1 | 1 | 0% | 2,051 | 2,519 | +23% | 0 | 0 | — |
case-05 | fail→pass | 8,003 | 9,574 | +20% | 1 | 1 | 0% | 1,930 | 3,095 | +60% | 0 | 0 | — |
case-06 | pass→pass | 9,388 | 7,022 | -25% | 1 | 1 | 0% | 1,863 | 2,242 | +20% | 0 | 0 | — |
case-07 | fail→pass | 10,684 | 7,916 | -26% | 1 | 1 | 0% | 2,318 | 2,588 | +12% | 0 | 0 | — |
case-08 | fail→pass | 18,208 | 10,637 | -42% | 1 | 1 | 0% | 3,997 | 3,174 | -21% | 0 | 0 | — |
case-09 | fail→pass | 10,537 | 8,688 | -18% | 1 | 1 | 0% | 2,546 | 2,811 | +10% | 0 | 0 | — |
case-10 | fail→pass | 7,137 | 7,926 | +11% | 1 | 1 | 0% | 1,551 | 2,531 | +63% | 0 | 0 | — |
case-11 | pass→pass | 6,992 | 9,680 | +38% | 1 | 1 | 0% | 1,675 | 3,122 | +86% | 0 | 0 | — |
case-12 | fail→pass | 10,068 | 7,479 | -26% | 1 | 1 | 0% | 1,985 | 2,195 | +11% | 0 | 0 | — |
case-13 | fail→pass | 7,349 | 5,873 | -20% | 1 | 1 | 0% | 1,599 | 1,953 | +22% | 0 | 0 | — |
case-15 | pass→pass | 15,458 | 11,212 | -27% | 1 | 1 | 0% | 3,846 | 3,395 | -12% | 0 | 0 | — |
case-16 | pass→pass | 6,636 | 9,104 | +37% | 1 | 1 | 0% | 1,669 | 2,823 | +69% | 0 | 0 | — |
case-17 | fail→pass | 11,870 | 11,276 | -5% | 1 | 1 | 0% | 2,776 | 3,281 | +18% | 0 | 0 | — |
case-18 | fail→fail | 1,783 | 6,862 | +285% | 1 | 1 | 0% | 316 | 1,629 | +416% | 0 | 0 | — |
case-19 | pass→pass | 6,689 | 6,003 | -10% | 1 | 1 | 0% | 1,596 | 2,036 | +28% | 0 | 0 | — |
case-20 | pass→pass | 9,334 | 7,336 | -21% | 1 | 1 | 0% | 2,106 | 2,403 | +14% | 0 | 0 | — |
case-21 | pass→pass | 8,236 | 8,815 | +7% | 1 | 1 | 0% | 1,685 | 2,433 | +44% | 0 | 0 | — |
case-22 | fail→fail | 14,190 | 3,479 | -75% | 1 | 1 | 0% | 3,375 | 1,128 | -67% | 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 +59 percentage points is the difference between those two pass rates over the 22 comparable cases. 1 case got worse with the skill loaded, and it is 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.