econ-audit
Econ Audit — Adversarial Econometrics Review
v1.0 — Adversarial reviewer that catches econometric specification errors, estimation mistakes, and silent analytical failures. Asks: "Will this produce correct economic conclusions?"
Review econometric code (Stata, R, or Python) for specification errors, estimation mistakes, and analytical choices that could produce wrong conclusions — even when the code runs without errors. This is the "hostile referee" for your analysis code.
Argument: $ARGUMENTS
- Path to a file (
.do,.R,.py) or a directory
Modes (append to argument):
spec(default) — Single-file specification review: clustering, controls, functional form, samplefull— Deep review with project context: reads data documentation, related files, pre-analysis plancompare— Compare two specification files or pre/post versions for specification drift
Flags:
pap:path/to/pap.pdf— Compare against a pre-analysis planvars:outcome1,outcome2— Focus audit on specific outcome variablesdesign:rct|did|iv|rd|panel— Specify research design (auto-detected if omitted)severity:high— Only report high-severity issues
More from thinkingwithagents/skills
academic-beamer-deck
>
13review-paper
Adversarial paper review simulating a skeptical referee — checks identification, statistical claims, robustness, and presentation against real referee patterns
9lit-review
Structured multi-session literature review workflow — scaffolds reviews, tracks papers, generates slide decks or documents, and runs referee passes for canonical paper coverage
9research-brainstorm
Brainstorm and stress-test research ideas as a senior scholar colleague would. Runs a multi-turn dialogue that clarifies the seed question (or generates one from cold start), probes it conversationally, launches a parallel deep literature scan across published and working-paper sources, critiques the question adversarially, generates 2–3 alternative framings, assesses feasibility (delegating to the find-data skill when available), and writes a research brief to the working directory. Trigger when the user asks to brainstorm a research idea, stress-test a research question, workshop a project, develop a new paper idea, assess novelty of a question, evaluate whether an idea is worth pursuing, refine a research direction, or check whether an idea clears a top-journal bar. Also trigger on phrases like "brainstorm with me", "is this novel", "has anyone done this", "what should I work on", "help me think through this idea", "workshop this question", "stress-test this", "poke holes in this", "new research idea", "research project idea", or when the user describes a half-formed question and asks for feedback. Use regardless of field but tuned for empirical economics and adjacent social sciences.
9code-review
Structured code review for research code (Stata, R, Python) against DIME, Gentzkow-Shapiro, AEA, and IPA standards — catches silent failures, reproducibility risks, and style issues
8find-data
Help researchers identify and catalog relevant datasets for empirical research. Use this skill whenever a user asks to find data, locate datasets, identify data sources, or asks "what data exists for...", "where can I get data on...", "is there a dataset that...", or similar. Also trigger when the user describes a research question and needs to know what data could support it, asks about data availability for a particular topic or geography, wants to compare data sources, or mentions needing panel data, cross-sectional data, or time-series data for a project. Trigger on mentions of "find data", "data search", "dataset", "data source", "microdata", "administrative data", "survey data", "public-use data", "restricted data", or when the user specifies parameters like time windows, geographic levels, or data frequency in the context of research. Also trigger when the user asks for help downloading, scraping, or structuring a dataset that has been identified.
8