Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Extract decisions, lessons, patterns, and surprises from completed phase artifacts
.claude/skills/davepoon-gsd-extract-learnings/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-05 | ✓→✗ | ▼ Worse | -80% | 0% |
| case-06 | ✓→✗ | ▼ Worse | -87% | 0% |
| case-07 | ✓→✗ | ▼ Worse | -86% | 0% |
| case-08 | ✓→✗ | ▼ Worse | -76% | 0% |
| case-09 | ✓→✗ | ▼ Worse | -82% | 0% |
<objective> Extract structured learnings from completed phase artifacts (PLAN.md, SUMMARY.md, VERIFICATION.md, UAT.md, STATE.md) into a LEARNINGS.md file that captures decisions, lessons learned, patterns discovered, and surprises encountered. </objective>
<execution_context> @${CLAUDE_PLUGIN_ROOT}/workflows/extract-learnings.md </execution_context>
Execute the extract-learnings workflow from @${CLAUDE_PLUGIN_ROOT}/workflows/extract-learnings.md end-to-end.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 5,071 | 4,714 | -7% | 1 | 1 | 0% | 845 | 419 | -50% | 0 | 0 | — |
case-02 | fail→fail | 10,163 | 3,428 | -66% | 1 | 1 | 0% | 1,735 | 330 | -81% | 0 | 0 | — |
case-03 | fail→fail | 9,583 | 6,401 | -33% | 1 | 1 | 0% | 1,601 | 352 | -78% | 0 | 0 | — |
case-04 | fail→fail | 4,158 | 5,339 | +28% | 1 | 1 | 0% | 239 | 421 | +76% | 0 | 0 | — |
case-05 | pass→fail | 13,922 | 4,534 | -67% | 1 | 1 | 0% | 2,291 | 449 | -80% | 0 | 0 | — |
case-06 | pass→fail | 16,621 | 4,860 | -71% | 1 | 1 | 0% | 2,544 | 330 | -87% | 0 | 0 | — |
case-07 | pass→fail | 17,000 | 3,256 | -81% | 1 | 1 | 0% | 2,764 | 383 | -86% | 0 | 0 | — |
case-08 | pass→fail | 8,886 | 4,155 | -53% | 1 | 1 | 0% | 1,429 | 340 | -76% | 0 | 0 | — |
case-09 | pass→fail | 10,331 | 4,226 | -59% | 1 | 1 | 0% | 1,647 | 300 | -82% | 0 | 0 | — |
case-10 | fail→fail | 5,245 | 5,738 | +9% | 1 | 1 | 0% | 280 | 376 | +34% | 0 | 0 | — |
case-11 | fail→fail | 5,309 | 5,328 | +0% | 1 | 1 | 0% | 196 | 349 | +78% | 0 | 0 | — |
case-12 | fail→fail | 5,591 | 3,934 | -30% | 1 | 1 | 0% | 831 | 283 | -66% | 0 | 0 | — |
case-13 | fail→fail | 3,411 | 4,096 | +20% | 1 | 1 | 0% | 410 | 323 | -21% | 0 | 0 | — |
case-14 | pass→fail | 7,262 | 3,643 | -50% | 1 | 1 | 0% | 1,256 | 295 | -77% | 0 | 0 | — |
case-15 | pass→fail | 8,800 | 5,918 | -33% | 1 | 1 | 0% | 1,391 | 426 | -69% | 0 | 0 | — |
case-16 | pass→fail | 6,953 | 4,680 | -33% | 1 | 1 | 0% | 1,087 | 380 | -65% | 0 | 0 | — |
case-17 | pass→fail | 8,187 | 4,988 | -39% | 1 | 1 | 0% | 1,175 | 379 | -68% | 0 | 0 | — |
case-18 | pass→pass | 5,631 | 4,011 | -29% | 1 | 1 | 0% | 934 | 735 | -21% | 0 | 0 | — |
case-19 | pass→fail | 7,875 | 3,571 | -55% | 1 | 1 | 0% | 1,225 | 322 | -74% | 0 | 0 | — |
case-20 | pass→fail | 20,649 | 6,801 | -67% | 1 | 1 | 0% | 3,309 | 393 | -88% | 0 | 0 | — |
case-21 | fail→fail | 16,795 | 7,743 | -54% | 1 | 1 | 0% | 275 | 643 | +134% | 0 | 0 | — |
case-22 | pass→fail | 10,521 | 5,847 | -44% | 1 | 1 | 0% | 1,255 | 456 | -64% | 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 1 counted toward the lift figure. The other 21 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. A headline lift is not published for this run.
Other measured skills in the registry, with their headline benchmark lift.