Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Diagnose Ruflo health, then report system, MCP server, and active-agent status without changing the installation
.claude/skills/ruvnet-ruflo-status/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | -53% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -62% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -45% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -59% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -44% | 0% |
Use this skill when the user asks for Ruflo health, diagnostics, MCP server state, or active-agent status.
Run these commands in order:
bashnpx @claude-flow/cli@latest doctor npx @claude-flow/cli@latest status
Summarize the diagnostics and status together. Do not repair, reset, start, stop, install, or otherwise change anything during the default workflow.
Only when the user explicitly asks to fix or auto-repair the installation, replace the first command with:
bashnpx @claude-flow/cli@latest doctor --fix npx @claude-flow/cli@latest status
Report what the repair changed and the resulting status. Never infer authorization for doctor --fix from a request to check or diagnose Ruflo.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 12,921 | 4,461 | -65% | 1 | 1 | 0% | 1,593 | 363 | -77% | 0 | 0 | — |
case-02 | fail→fail | 6,480 | 4,902 | -24% | 1 | 1 | 0% | 1,002 | 488 | -51% | 0 | 0 | — |
case-03 | fail→fail | 10,015 | 4,342 | -57% | 1 | 1 | 0% | 1,848 | 415 | -78% | 0 | 0 | — |
case-04 | fail→fail | 11,036 | 3,658 | -67% | 1 | 1 | 0% | 1,775 | 740 | -58% | 0 | 0 | — |
case-05 | fail→fail | 10,246 | 4,917 | -52% | 1 | 1 | 0% | 1,827 | 488 | -73% | 0 | 0 | — |
case-06 | fail→pass | 7,865 | 2,228 | -72% | 1 | 1 | 0% | 1,251 | 582 | -53% | 0 | 0 | — |
case-07 | pass→pass | 7,800 | 2,913 | -63% | 1 | 1 | 0% | 1,415 | 786 | -44% | 0 | 0 | — |
case-08 | fail→fail | 8,576 | 4,645 | -46% | 1 | 1 | 0% | 1,493 | 405 | -73% | 0 | 0 | — |
case-09 | fail→pass | 9,287 | 6,359 | -32% | 1 | 1 | 0% | 1,638 | 620 | -62% | 0 | 0 | — |
case-10 | fail→fail | 8,153 | 7,405 | -9% | 1 | 1 | 0% | 1,449 | 945 | -35% | 0 | 0 | — |
case-11 | fail→pass | 6,873 | 2,499 | -64% | 1 | 1 | 0% | 1,072 | 586 | -45% | 0 | 0 | — |
case-12 | fail→pass | 10,388 | 2,887 | -72% | 1 | 1 | 0% | 1,840 | 755 | -59% | 0 | 0 | — |
case-13 | fail→fail | 4,889 | 4,577 | -6% | 1 | 1 | 0% | 776 | 491 | -37% | 0 | 0 | — |
case-14 | fail→fail | 9,136 | 4,403 | -52% | 1 | 1 | 0% | 1,482 | 454 | -69% | 0 | 0 | — |
case-15 | fail→pass | 6,477 | 2,838 | -56% | 1 | 1 | 0% | 1,239 | 694 | -44% | 0 | 0 | — |
case-16 | pass→pass | 9,667 | 3,687 | -62% | 1 | 1 | 0% | 1,732 | 787 | -55% | 0 | 0 | — |
case-17 | pass→pass | 7,827 | 2,483 | -68% | 1 | 1 | 0% | 1,418 | 604 | -57% | 0 | 0 | — |
case-18 | pass→pass | 8,520 | 3,086 | -64% | 1 | 1 | 0% | 1,392 | 769 | -45% | 0 | 0 | — |
case-19 | pass→pass | 4,427 | 3,008 | -32% | 1 | 1 | 0% | 941 | 812 | -14% | 0 | 0 | — |
case-20 | pass→pass | 12,016 | 8,702 | -28% | 1 | 1 | 0% | 2,686 | 2,063 | -23% | 0 | 0 | — |
case-21 | pass→pass | 10,907 | 6,347 | -42% | 1 | 1 | 0% | 2,140 | 1,464 | -32% | 0 | 0 | — |
case-22 | pass→pass | 6,260 | 3,952 | -37% | 1 | 1 | 0% | 1,145 | 952 | -17% | 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 15 counted toward the lift figure. The other 7 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 +23 percentage points is the difference between those two pass rates over the 15 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.