Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Summarize academic papers, extract key findings, and identify research gaps
.claude/skills/brycewang-stanford-lit-review/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 412% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -20% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 8% | 0% |
| case-12 | ✓→✗ | ▼ Worse | 16% | 0% |
| case-05 | ✓→✓ | = Same ✓ | 11% | 0% |
Conduct a structured literature search and synthesis on the given topic.
Input: $ARGUMENTS — a topic, paper title, research question, or phenomenon to investigate.
$ARGUMENTS. If a specific paper is named, use it as the anchor.master_supporting_docs/supporting_papers/ for uploaded papersWebSearch to find recent publications (if available)WebFetch to access working paper repositories (if available).bib file for papers already in the projectquality_reports/lit_review_[sanitized_topic].mdmarkdown# Literature Review: [Topic] **Date:** [YYYY-MM-DD] **Query:** [Original query from user] ## Summary [2-3 paragraph overview of the state of the literature] ## Key Papers ### [Author (Year)] — [Short Title] - **Main contribution:** [1-2 sentences] - **Method:** [Identification strategy / data] - **Key finding:** [Result with effect size if available] - **Relevance:** [Why it matters for our research] [Repeat for 5-15 papers, ordered by relevance] ## Thematic Organization ### Theoretical Contributions [Grouped discussion] ### Empirical Findings [Grouped discussion with comparison across studies] ### Methodological Innovations [Methods relevant to the topic] ## Gaps and Opportunities 1. [Gap 1 — what's missing and why it matters] 2. [Gap 2] 3. [Gap 3] ## Suggested Next Steps - [Concrete actions: papers to read, data to obtain, methods to consider] ## BibTeX Entries
@article{...}
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | fail→fail | 3,339 | 1,362 | -59% | 1 | 1 | 0% | 619 | 812 | +31% | 0 | 0 | — |
case-01 | fail→fail | 5,002 | 3,009 | -40% | 1 | 1 | 0% | 337 | 970 | +188% | 0 | 0 | — |
case-02 | fail→fail | 5,549 | 4,138 | -25% | 1 | 1 | 0% | 464 | 939 | +102% | 0 | 0 | — |
case-03 | fail→fail | 30,145 | 33,292 | +10% | 1 | 1 | 0% | 6,228 | 946 | -85% | 0 | 0 | — |
case-04 | fail→pass | 7,365 | 26,633 | +262% | 1 | 1 | 0% | 1,051 | 5,380 | +412% | 0 | 0 | — |
case-05 | pass→pass | 13,111 | 13,687 | +4% | 1 | 1 | 0% | 2,530 | 2,808 | +11% | 0 | 0 | — |
case-06 | fail→fail | 16,843 | 9,743 | -42% | 1 | 1 | 0% | 2,619 | 1,618 | -38% | 0 | 0 | — |
case-07 | pass→pass | 6,922 | 3,872 | -44% | 1 | 1 | 0% | 1,231 | 1,242 | +1% | 0 | 0 | — |
case-08 | pass→pass | 10,251 | 14,129 | +38% | 1 | 1 | 0% | 1,886 | 3,178 | +69% | 0 | 0 | — |
case-09 | pass→pass | 14,784 | 24,922 | +69% | 1 | 1 | 0% | 2,468 | 5,144 | +108% | 0 | 0 | — |
case-20 | pass→pass | 18,867 | 28,201 | +49% | 1 | 1 | 0% | 3,639 | 6,580 | +81% | 0 | 0 | — |
case-10 | fail→pass | 7,772 | 2,798 | -64% | 1 | 1 | 0% | 1,458 | 1,161 | -20% | 0 | 0 | — |
case-11 | fail→fail | 4,322 | 1,741 | -60% | 1 | 1 | 0% | 754 | 930 | +23% | 0 | 0 | — |
case-12 | pass→fail | 14,019 | 13,544 | -3% | 1 | 1 | 0% | 2,407 | 2,801 | +16% | 0 | 0 | — |
case-13 | fail→fail | 11,179 | 2,486 | -78% | 1 | 1 | 0% | 2,185 | 1,094 | -50% | 0 | 0 | — |
case-15 | fail→fail | 10,000 | 1,347 | -87% | 1 | 1 | 0% | 1,663 | 846 | -49% | 0 | 0 | — |
case-16 | fail→fail | 6,875 | 1,490 | -78% | 1 | 1 | 0% | 1,384 | 901 | -35% | 0 | 0 | — |
case-17 | fail→fail | 11,358 | 2,555 | -78% | 1 | 1 | 0% | 1,793 | 1,029 | -43% | 0 | 0 | — |
case-18 | fail→pass | 9,468 | 7,632 | -19% | 1 | 1 | 0% | 1,765 | 1,905 | +8% | 0 | 0 | — |
case-19 | fail→fail | 5,019 | 13,685 | +173% | 1 | 1 | 0% | 916 | 2,781 | +204% | 0 | 0 | — |
case-21 | fail→fail | 3,934 | 3,820 | -3% | 1 | 1 | 0% | 809 | 900 | +11% | 0 | 0 | — |
case-22 | pass→pass | 9,059 | 25,717 | +184% | 1 | 1 | 0% | 1,709 | 5,568 | +226% | 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 17 counted toward the lift figure. The other 5 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 +9 percentage points is the difference between those two pass rates over the 17 comparable cases. 2 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.