From jel-skills
Builds exhibits for JEL surveys: who-found-what summary tables, conceptual/framework figures, and meta-evidence plots. Ensures credible comparison of studies with self-contained captions and sourced cells.
How this skill is triggered — by the user, by Claude, or both
Slash command
/jel-skills:jel-tables-figuresThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
- The synthesis is done and the reader needs to *see* the field at a glance
JEL exhibits summarize across the literature; they do not present the author's own estimation. The workhorses:
| Exhibit | Purpose | Design notes |
|---|---|---|
| Who-found-what summary table | one row per study (or per design class): question/estimand, method, sample, finding (direction + magnitude), credibility note | rows ordered by the framework's cells, not chronology; columns let the reader compare comparable objects |
| Conceptual / framework figure | render the organizing spine — taxonomy tree, mechanism diagram, the simple unifying model | this is often the survey's signature exhibit; it should be restate-able from memory |
| Meta-evidence exhibit | a forest-style plot or funnel of effect sizes, a timeline of methods, a coverage map | use only when the estimates are commensurable; otherwise it manufactures false consensus |
jel-literature-synthesis).If you assemble effect sizes into a quantitative synthesis (forest plot, meta-regression), you are doing a meta-analysis, with all its assumptions — comparable estimands, publication-bias diagnostics, weighting. Only do this where the literature genuinely supports it; otherwise a qualitative who-found-what table is more honest than a spurious pooled number. JEL readers include the methodologists who would catch an invalid pooling.
When a figure from a surveyed paper is central, prefer a re-drawn synthesis figure (your own panel that places several studies on common axes) over copying one paper's exhibit. A re-drawn figure serves the survey's argument and avoids the consensus-by-accident of reprinting whichever paper had the prettiest chart; if you do reproduce an original figure, attribute it and secure any permission the AEA style guide requires.
【Exhibit set】<list: summary tables / conceptual figure / meta-evidence>
【Summary table】rows by framework cell; comparable estimands only? Y/N
【Credibility column】present for every finding? Y/N
【Conceptual figure】renders the spine; restate-able from memory? Y/N
【Meta-evidence】pools only commensurable estimates (or omitted)? Y/N
【Sourcing】every cell traces to the evidence matrix? Y/N
【Next step】→ jel-writing-style (weave exhibits into the synthesis prose)
npx claudepluginhub brycewang-stanford/awesome-journal-skills --plugin jel-skillsGenerates three exhibit types for Annual Review of Economics (ARE) reviews: who-found-what summary tables, conceptual/framework figures, and meta-evidence exhibits that synthesize across studies.
Builds review exhibits for Academy of Management Annals: framework figures, synthesis tables, search/coverage exhibits, and gap matrices. Does not produce regression tables.
Guides designing self-explanatory figures and tables for JEP articles aimed at non-specialists. Emphasizes one-claim exhibits, rounded numbers, direct labels, and minimal chartjunk.