Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Edit an existing weft workflow template. Use only when user types /wf-edit-template.
.claude/skills/ccplugins-wf-edit-template/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 18% | 0% |
| case-19 | ✗→✓ | ▲ Improved | -68% | 0% |
| case-22 | ✗→✓ | ▲ Improved | 101% | 0% |
| case-16 | ✓→✗ | ▼ Worse | -23% | 0% |
| case-21 | ✓→✗ | ▼ Worse | -39% | 0% |
Load an existing template, modify it interactively, and save.
$ARGUMENTS
If no arguments, list templates and ask which to edit:
bashpython3 "${CLAUDE_PLUGIN_ROOT}/core/cli.py" start
Load the template preview:
bashpython3 "${CLAUDE_PLUGIN_ROOT}/core/cli.py" preview <name>
Display the current template with all step details, guards, and policies.
Ask: "What would you like to change?"
Supported modifications:
Show the modified template and ask: "Save these changes?"
Ask whether to overwrite the original or save as a new name.
If saving to the project-local directory:
bashecho '<template_json>' | python3 "${CLAUDE_PLUGIN_ROOT}/core/cli.py" save-template
If the original is a plugin-bundled template, always save as a project-local override (don't modify plugin files).
Tell the user the template was saved and how to start it.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 16,284 | 7,472 | -54% | 1 | 1 | 0% | 2,419 | 657 | -73% | 0 | 0 | — |
case-02 | fail→fail | 2,256 | 7,318 | +224% | 1 | 1 | 0% | 254 | 669 | +163% | 0 | 0 | — |
case-03 | fail→fail | 14,651 | 7,790 | -47% | 1 | 1 | 0% | 2,113 | 741 | -65% | 0 | 0 | — |
case-04 | fail→fail | 11,453 | 5,347 | -53% | 1 | 1 | 0% | 1,586 | 1,095 | -31% | 0 | 0 | — |
case-05 | fail→fail | 12,694 | 13,563 | +7% | 1 | 1 | 0% | 2,282 | 777 | -66% | 0 | 0 | — |
case-06 | fail→fail | 11,744 | 3,815 | -68% | 1 | 1 | 0% | 1,856 | 813 | -56% | 0 | 0 | — |
case-07 | fail→fail | 7,453 | 6,995 | -6% | 1 | 1 | 0% | 1,144 | 655 | -43% | 0 | 0 | — |
case-08 | fail→fail | 15,528 | 7,918 | -49% | 1 | 1 | 0% | 2,375 | 726 | -69% | 0 | 0 | — |
case-09 | fail→fail | 15,785 | 8,947 | -43% | 1 | 1 | 0% | 2,275 | 937 | -59% | 0 | 0 | — |
case-10 | fail→fail | 11,729 | 6,628 | -43% | 1 | 1 | 0% | 2,065 | 649 | -69% | 0 | 0 | — |
case-11 | fail→fail | 12,128 | 6,605 | -46% | 1 | 1 | 0% | 1,893 | 673 | -64% | 0 | 0 | — |
case-12 | fail→fail | 12,771 | 7,067 | -45% | 1 | 1 | 0% | 1,953 | 711 | -64% | 0 | 0 | — |
case-13 | fail→fail | 9,287 | 4,552 | -51% | 1 | 1 | 0% | 1,220 | 883 | -28% | 0 | 0 | — |
case-14 | fail→fail | 9,631 | 7,057 | -27% | 1 | 1 | 0% | 1,130 | 858 | -24% | 0 | 0 | — |
case-15 | fail→pass | 13,978 | 12,425 | -11% | 1 | 1 | 0% | 1,951 | 2,298 | +18% | 0 | 0 | — |
case-16 | pass→fail | 9,980 | 4,983 | -50% | 1 | 1 | 0% | 1,512 | 1,165 | -23% | 0 | 0 | — |
case-17 | fail→fail | 13,032 | 4,607 | -65% | 1 | 1 | 0% | 1,790 | 947 | -47% | 0 | 0 | — |
case-18 | fail→fail | 15,476 | 7,060 | -54% | 1 | 1 | 0% | 2,317 | 672 | -71% | 0 | 0 | — |
case-19 | fail→pass | 20,793 | 4,071 | -80% | 1 | 1 | 0% | 2,915 | 925 | -68% | 0 | 0 | — |
case-20 | pass→pass | 10,596 | 5,383 | -49% | 1 | 1 | 0% | 1,622 | 1,072 | -34% | 0 | 0 | — |
case-21 | pass→fail | 12,201 | 11,501 | -6% | 1 | 1 | 0% | 1,840 | 1,127 | -39% | 0 | 0 | — |
case-22 | fail→pass | 8,632 | 14,066 | +63% | 1 | 1 | 0% | 1,196 | 2,399 | +101% | 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 12 counted toward the lift figure. The other 10 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 0 percentage points is the difference between those two pass rates over the 12 comparable cases. 6 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.