Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Split current work into small reviewable PRs. Use when the user asks to split a chat, set of changes, branch, or PR into smaller reviewable pieces.
.claude/skills/kunanonj-cursor-split-to-prs/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-12 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-16 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-23 | ✗→✓ | ▲ Improved | -36% | 0% |
| case-04 | ✓→✗ | ▼ Worse | -53% | 0% |
| case-05 | ✓→✓ | ＝ Same ✓ | 1% | 0% |
Turn one pile of work into a few small PRs.
reset --hard, clean -fdx, branch deletion, force-push, history rewrite) without explicit approval.main, so do not assume there is already a safe branch.git add . / git add -A.Compare the current work to the repo's default branch, including committed and uncommitted changes. Summarize the real slices you see, and use the chat history to recover intent.
Before proposing slices, find ownership signals for the touched paths (CODEOWNERS, nested ownership files, tools/ownership/PRODUCTOWNERS, or repo equivalents) and use them to identify natural reviewer boundaries.
Use judgment on detail. Usually PR titles are enough. Add a one-line scope note only when a title is unclear. Show a Mermaid diagram when there are multiple slices.
Optimize for reviewer-aligned PRs with minimal unrelated diff: split independent owners or concerns, keep tightly coupled changes together, and when stacking is necessary, order foundations before consumers.
Default to independent PRs off the default branch. Stack PRs only when the dependency is real.
Ask for approval before starting.
bash SHA=$(git stash create "pre-split") if [ -n "$SHA" ]; then git update-ref "refs/backup/pre-split-$(date +%s)" "$SHA" fi
Keep it short: PR titles and URLs, plus anything left on the starting branch or working tree. Do not delete the backup ref or original branch unless the user asks.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 9,100 | 4,719 | -48% | 1 | 1 | 0% | 1,585 | 711 | -55% | 0 | 0 | — |
case-02 | fail→fail | 2,371 | 2,590 | +9% | 1 | 1 | 0% | 270 | 805 | +198% | 0 | 0 | — |
case-03 | fail→fail | 7,746 | 2,632 | -66% | 1 | 1 | 0% | 1,202 | 744 | -38% | 0 | 0 | — |
case-04 | pass→fail | 9,541 | 5,939 | -38% | 1 | 1 | 0% | 1,613 | 756 | -53% | 0 | 0 | — |
case-05 | pass→pass | 13,587 | 8,642 | -36% | 1 | 1 | 0% | 2,004 | 2,015 | +1% | 0 | 0 | — |
case-06 | fail→fail | 7,150 | 7,328 | +2% | 1 | 1 | 0% | 803 | 1,540 | +92% | 0 | 0 | — |
case-07 | fail→fail | 9,639 | 5,982 | -38% | 1 | 1 | 0% | 1,687 | 1,038 | -38% | 0 | 0 | — |
case-08 | pass→pass | 4,733 | 2,536 | -46% | 1 | 1 | 0% | 892 | 984 | +10% | 0 | 0 | — |
case-09 | fail→fail | 7,506 | 2,420 | -68% | 1 | 1 | 0% | 1,388 | 748 | -46% | 0 | 0 | — |
case-10 | pass→pass | 8,926 | 6,677 | -25% | 1 | 1 | 0% | 1,600 | 1,669 | +4% | 0 | 0 | — |
case-11 | pass→pass | 9,749 | 5,227 | -46% | 1 | 1 | 0% | 1,717 | 1,403 | -18% | 0 | 0 | — |
case-12 | fail→pass | 9,229 | 5,158 | -44% | 1 | 1 | 0% | 1,705 | 1,350 | -21% | 0 | 0 | — |
case-13 | pass→pass | 6,215 | 2,207 | -64% | 1 | 1 | 0% | 1,116 | 861 | -23% | 0 | 0 | — |
case-14 | pass→pass | 7,090 | 2,575 | -64% | 1 | 1 | 0% | 1,246 | 906 | -27% | 0 | 0 | — |
case-15 | pass→pass | 6,723 | 2,594 | -61% | 1 | 1 | 0% | 1,213 | 904 | -25% | 0 | 0 | — |
case-16 | fail→pass | 9,569 | 3,560 | -63% | 1 | 1 | 0% | 1,546 | 1,139 | -26% | 0 | 0 | — |
case-17 | pass→pass | 9,179 | 3,913 | -57% | 1 | 1 | 0% | 1,404 | 1,112 | -21% | 0 | 0 | — |
case-18 | fail→fail | 11,291 | 2,902 | -74% | 1 | 1 | 0% | 1,979 | 943 | -52% | 0 | 0 | — |
case-19 | pass→pass | 9,903 | 4,917 | -50% | 1 | 1 | 0% | 1,653 | 1,227 | -26% | 0 | 0 | — |
case-20 | pass→pass | 12,246 | 4,512 | -63% | 1 | 1 | 0% | 2,085 | 1,235 | -41% | 0 | 0 | — |
case-21 | pass→pass | 10,837 | 6,177 | -43% | 1 | 1 | 0% | 1,821 | 1,507 | -17% | 0 | 0 | — |
case-22 | pass→pass | 11,955 | 8,606 | -28% | 1 | 1 | 0% | 2,272 | 2,389 | +5% | 0 | 0 | — |
case-23 | fail→pass | 8,671 | 2,466 | -72% | 1 | 1 | 0% | 1,408 | 903 | -36% | 0 | 0 | — |
case-24 | pass→pass | 12,823 | 4,006 | -69% | 1 | 1 | 0% | 2,013 | 1,066 | -47% | 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. 24 cases were attempted, and 22 counted toward the lift figure. The other 2 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 +8 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.