Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Run support-safe iTerm diagnostics. Use when iTerm agent control, AppleScript, pane targeting, badges, screenshots, permissions, or layout automation is not working, or when the user asks for an iTerm health check.
.claude/skills/antonio-mello-ai-iterm-diagnostics/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-02 | ✗→✓ | ▲ Improved | -65% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -7% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -38% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -36% | 0% |
Use this skill to collect a read-only health report for iTerm automation. Do not dump shell history, environment variables, prompts, tokens, or command output from panes.
bashskills/iterm-diagnostics/scripts/iterm-diagnostics
Use context only when the issue depends on pane identity:
bashskills/iterm-diagnostics/scripts/iterm-diagnostics --include-context
Run a visible badge test only when the user is ready for a small UI change:
bashskills/iterm-diagnostics/scripts/iterm-diagnostics --test-badge "diag"
osascript is available.com.googlecode.iterm2.iterm_bundle fails: iTerm is missing, renamed, or AppleScript cannot resolve it.current_session fails: iTerm is closed, has no windows, or Automation permission is blocked.list_sessions fails: AppleScript access is blocked or the iTerm scripting dictionary changed.identify reports caller-fallback-current: the agent runtime has no TTY, so targeting is based on the current iTerm focus.--include-context only when session names and TTYs are needed.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-05 | pass→pass | 6,937 | 5,991 | -14% | 1 | 1 | 0% | 1,304 | 1,597 | +22% | 0 | 0 | — |
case-18 | pass→pass | 5,987 | 3,382 | -44% | 1 | 1 | 0% | 996 | 926 | -7% | 0 | 0 | — |
case-01 | fail→fail | 8,256 | 4,304 | -48% | 1 | 1 | 0% | 1,557 | 640 | -59% | 0 | 0 | — |
case-02 | fail→pass | 10,997 | 1,685 | -85% | 1 | 1 | 0% | 2,082 | 725 | -65% | 0 | 0 | — |
case-03 | fail→fail | 5,006 | 2,790 | -44% | 1 | 1 | 0% | 889 | 758 | -15% | 0 | 0 | — |
case-04 | fail→pass | 9,124 | 6,505 | -29% | 1 | 1 | 0% | 1,766 | 1,644 | -7% | 0 | 0 | — |
case-06 | pass→pass | 6,896 | 5,671 | -18% | 1 | 1 | 0% | 1,250 | 1,598 | +28% | 0 | 0 | — |
case-07 | fail→pass | 9,287 | 2,631 | -72% | 1 | 1 | 0% | 1,784 | 946 | -47% | 0 | 0 | — |
case-08 | fail→pass | 6,566 | 2,141 | -67% | 1 | 1 | 0% | 1,299 | 808 | -38% | 0 | 0 | — |
case-09 | pass→pass | 10,614 | 4,735 | -55% | 1 | 1 | 0% | 1,912 | 1,266 | -34% | 0 | 0 | — |
case-10 | fail→pass | 9,805 | 3,605 | -63% | 1 | 1 | 0% | 1,653 | 1,057 | -36% | 0 | 0 | — |
case-11 | fail→pass | 10,979 | 9,785 | -11% | 1 | 1 | 0% | 1,881 | 2,105 | +12% | 0 | 0 | — |
case-12 | fail→pass | 7,939 | 2,238 | -72% | 1 | 1 | 0% | 1,263 | 777 | -38% | 0 | 0 | — |
case-13 | pass→pass | 4,816 | 2,583 | -46% | 1 | 1 | 0% | 845 | 827 | -2% | 0 | 0 | — |
case-14 | fail→pass | 7,006 | 2,539 | -64% | 1 | 1 | 0% | 1,192 | 888 | -26% | 0 | 0 | — |
case-15 | fail→pass | 11,707 | 1,925 | -84% | 1 | 1 | 0% | 2,119 | 764 | -64% | 0 | 0 | — |
case-16 | pass→pass | 10,720 | 2,455 | -77% | 1 | 1 | 0% | 1,724 | 875 | -49% | 0 | 0 | — |
case-17 | pass→pass | 3,616 | 1,783 | -51% | 1 | 1 | 0% | 562 | 652 | +16% | 0 | 0 | — |
case-19 | fail→pass | 4,831 | 1,330 | -72% | 1 | 1 | 0% | 797 | 661 | -17% | 0 | 0 | — |
case-20 | pass→pass | 9,284 | 2,560 | -72% | 1 | 1 | 0% | 1,471 | 867 | -41% | 0 | 0 | — |
case-21 | pass→pass | 6,781 | 3,595 | -47% | 1 | 1 | 0% | 1,176 | 781 | -34% | 0 | 0 | — |
case-22 | fail→pass | 5,661 | 1,685 | -70% | 1 | 1 | 0% | 947 | 649 | -31% | 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 21 counted toward the lift figure. The other 1 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 +50 percentage points is the difference between those two pass rates over the 21 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.