Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/aeja-tables-figures
Use when building or revising the exhibits of an American Economic Journal: Applied Economics (AEJ: Applied) manuscript so the main causal result is legible in one table or figure and respects AEA house presentation norms. Formats exhibits; it does not establish the result (aeja-identification / aeja-robustness) or write the surrounding prose.
Passed 20 of 22 test cases with the skill, and 13 of 22 without it.
and 7 more in the full report
9 fixed11 passed either way0 still failing2 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 9 test cases the skill fixed. Same model, same prompt, with and without the skill.
I am submitting a draft with an event study and balance table. Evaluate my layout and report using the standard AEJ exhibit bracketed output format.
| Checked | Without skill | With skill |
|---|---|---|
| Outputs the bracketed header 【Headline exhibit】. | Missed | Passed |
| Outputs 【Next step】aeja-writing-style. | Missed | Passed |
The instructions your model reads when the skill loads.
At AEJ: Applied the main causal estimate should be findable in seconds and every exhibit should earn its place. AEA house style permits significance stars but expects standard errors in parentheses, clear notes that make each exhibit self-contained, and clean figures over chartjunk. Lead with the design's signature visual — the event-study plot, the RD scatter, or the balance table — because at this journal the picture of the identification is half the persuasion.
| Exhibit | What it must show | Common failure |
|---|---|---|
| Main results table | headline coefficient, SE in parentheses, N, controls indicated, dependent-var mean | too many columns; no SEs; over-precision |
| Balance table (RCT) | baseline means by arm, differences, joint test | missing joint test; no attrition row |
| Event-study figure | leads + lags, CIs, reference period, flat pre-trends visible | no CIs; ambiguous reference period |
| RD figure | binned scatter + fitted lines, bandwidth, density | overfit polynomial; no density panel |
| First-stage / IV table | first-stage F, exclusion logic in notes | weak first stage hidden |
| Robustness exhibit | point-estimate stability across checks | a starred wall with no map |
Generate exhibits from the fitted result, not by retyping numbers (the usual source of
body-vs-appendix drift). Full map: execution-with-mcp.
etable (multi-model columns) or did_summary_to_latex straight from the
result_id — one variable definition, one set of numbers, body and appendix in sync.plot_from_result / enhanced_event_study_plot / event_study_table —
axis units and the SE/clustering note baked in.test cases passed, vs 13 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.