Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Sibyl 章节撰写 agent - 撰写论文特定章节
.claude/skills/majiayu000-sibyl-section-writer/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-17 | ✗→✓ | ▲ Improved | 214% | 0% |
| case-14 | ✗→✓ | ▲ Improved | 3222% | 0% |
| case-16 | ✗→✓ | ▲ Improved | 395% | 0% |
| case-20 | ✓→✓ | ＝ Same ✓ | 232% | 0% |
| case-21 | ✓→✓ | ＝ Same ✓ | -16% | 0% |
!SIBYL_WORKSPACE="$ARGUMENTS[0]" .venv/bin/python3 -c "from sibyl.orchestrate import render_skill_prompt; import os; ws = os.environ.get('SIBYL_WORKSPACE', ''); print(render_skill_prompt('section_writer', workspace_path=ws))"
AGENT_NAME: sibyl-section-writer AGENT_TIER: sibyl-standard SIBYL_ROOT: /Users/cwan0785/sibyl-system
Workspace path: $ARGUMENTS0] Section: $ARGUMENTS1] Section ID: $ARGUMENTS2] Write to: $ARGUMENTS0]/writing/sections/$ARGUMENTS2].md
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-09 | fail→fail | 15,657 | 10,979 | -30% | 1 | 1 | 0% | 237 | 479 | +102% | 0 | 0 | — |
case-10 | fail→fail | 14,566 | 10,830 | -26% | 1 | 1 | 0% | 119 | 452 | +280% | 0 | 0 | — |
case-11 | fail→fail | 10,518 | 6,594 | -37% | 1 | 1 | 0% | 303 | 541 | +79% | 0 | 0 | — |
case-17 | fail→pass | 10,019 | 6,959 | -31% | 1 | 1 | 0% | 153 | 480 | +214% | 0 | 0 | — |
case-18 | fail→fail | 11,163 | 11,457 | +3% | 1 | 1 | 0% | 316 | 549 | +74% | 0 | 0 | — |
case-19 | fail→fail | 10,036 | 16,031 | +60% | 1 | 1 | 0% | 261 | 520 | +99% | 0 | 0 | — |
case-01 | fail→fail | 10,284 | 11,879 | +16% | 1 | 1 | 0% | 267 | 813 | +204% | 0 | 0 | — |
case-02 | fail→fail | 15,664 | 15,596 | -0% | 1 | 1 | 0% | 299 | 766 | +156% | 0 | 0 | — |
case-03 | fail→fail | 12,662 | 43,237 | +241% | 1 | 1 | 0% | 483 | 464 | -4% | 0 | 0 | — |
case-04 | fail→fail | 11,212 | 12,298 | +10% | 1 | 1 | 0% | 349 | 665 | +91% | 0 | 0 | — |
case-05 | fail→fail | 35,012 | 10,174 | -71% | 1 | 1 | 0% | 3,800 | 458 | -88% | 0 | 0 | — |
case-06 | fail→fail | 9,948 | 6,149 | -38% | 1 | 1 | 0% | 197 | 490 | +149% | 0 | 0 | — |
case-07 | fail→fail | 15,778 | 23,813 | +51% | 1 | 1 | 0% | 282 | 415 | +47% | 0 | 0 | — |
case-08 | fail→fail | 5,323 | 6,029 | +13% | 1 | 1 | 0% | 285 | 482 | +69% | 0 | 0 | — |
case-12 | fail→fail | 44,398 | 10,536 | -76% | 1 | 1 | 0% | 5,730 | 464 | -92% | 0 | 0 | — |
case-13 | fail→fail | 20,056 | 16,582 | -17% | 1 | 1 | 0% | 2,319 | 500 | -78% | 0 | 0 | — |
case-14 | fail→pass | 15,492 | 40,634 | +162% | 1 | 1 | 0% | 181 | 6,012 | +3222% | 0 | 0 | — |
case-15 | fail→fail | 18,989 | 11,870 | -37% | 1 | 1 | 0% | 2,195 | 494 | -77% | 0 | 0 | — |
case-16 | fail→pass | 11,166 | 9,192 | -18% | 1 | 1 | 0% | 408 | 2,018 | +395% | 0 | 0 | — |
case-20 | pass→pass | 4,758 | 28,499 | +499% | 1 | 1 | 0% | 850 | 2,819 | +232% | 0 | 0 | — |
case-21 | pass→pass | 7,156 | 4,704 | -34% | 1 | 1 | 0% | 1,268 | 1,071 | -16% | 0 | 0 | — |
case-22 | pass→pass | 26,519 | 40,953 | +54% | 1 | 1 | 0% | 3,224 | 6,746 | +109% | 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 3 counted toward the lift figure. The other 19 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.