maxwell2732/codex-r-empirical-research

maxwell2732/codex-r-empirical-research

Agent Claude Code

Substantive domain review for empirical economics analyses. Template agent — customize the 5 review lenses for your sub-field. Checks identification correctness, derivation steps, citation fidelity, code-theory alignment, and logical consistency.

maxwell2732/codex-r-empirical-research

Agent Claude Code

Spec-review agent for empirical economics. Checks clustering level, FE absorption, weights, IV first-stage strength, multiple-hypothesis correction, sample selection, DiD assumptions, and SE method choice. Use on any analysis script before merging to main.

maxwell2732/codex-r-empirical-research

Agent Claude Code

Holistic pedagogical review for academic slides. Checks narrative arc, prerequisite assumptions, worked examples, notation clarity, and deck-level pacing. Use after content is drafted.

maxwell2732/codex-r-empirical-research

Agent Claude Code

Implementation planning specialist for R projects. Use for feature implementation, architectural decisions, or complex refactoring.

maxwell2732/codex-r-empirical-research

Agent Claude Code

Expert proofreading agent for academic lecture slides. Reviews for grammar, typos, overflow, and consistency. Use proactively after creating or modifying lecture content.

maxwell2732/codex-r-empirical-research

Agent Claude Code

Verifies that numerical claims about R analysis results actually appear in the corresponding R log file or output table. Refuses to validate claims with no log provenance. Use before any commit or report that contains numerical results.

maxwell2732/codex-r-empirical-research

Agent Claude Code

R code reviewer for empirical economics scripts. Checks reproducibility, logging hygiene, naming, magic numbers, table/figure quality, and adherence to project conventions. Use after writing or modifying any .R file.

maxwell2732/codex-r-empirical-research

Agent Claude Code

End-to-end verification agent for the R pipeline. Confirms scripts run cleanly, logs exist with no errors, output files are present and fresh, and reports render. Use proactively before committing or creating PRs.