replicate

Compare paper-reported results against Stata translations using gold-standard targets with tolerances.

220|358|Updated Apr 29, 2026
One-click install
npx skills add https://github.com/maxwell2732/codex-stata-for-economists --skill replicate-maxwell2732
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: replicate
Source: https://github.com/maxwell2732/codex-stata-for-economists/tree/main/.claude/skills/replicate
Command: npx skills add https://github.com/maxwell2732/codex-stata-for-economists --skill replicate-maxwell2732

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Help researchers reproduce a published paper’s results in a Stata-based workflow by building an inventory of gold-standard targets, translating the original analysis faithfully, and performing tolerance-respecting comparisons to detect divergences.

Core Features & Use Cases

  • Replication protocol execution: Applies .claude/rules/replication-protocol.md end-to-end, including target inventorying and comparison.
  • Gold-standard targets with tolerances: Records table/figure targets (with MUST/SHOULD/MAY tiers) and explicit tolerance expectations for pass/fail decisions.
  • Stata translation and funnel debugging: Translates the original code line-by-line into project Stata dofiles and investigates discrepancies via sample restrictions, variable construction, SE/df adjustments, and command-default differences.

Example use case: starting from a paper where you want to verify a key ATT/ATE estimate and first-stage diagnostics, you replicate the targeted tables and produce a structured report indicating which targets passed within tolerance and which require investigation.

Quick Start

Run replicate on a paper by providing its short name or target file, for example: /replicate AbadieDiamondHainmueller2010.

Frequently Asked Questions about replicate

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I replicate a published paper's regression results in Stata?

To replicate regression results in Stata, you build an inventory of gold-standard targets, translate the original analysis line-by-line into Stata dofiles, and compare outputs using predefined tolerances to detect divergences.

What is tolerance checking in research replication workflows?

Tolerance checking in research replication compares your Stata output against gold-standard targets using explicit thresholds. It enforces PASS/FAIL reporting through log validation to determine if estimates match within acceptable bounds.

Why does my Stata replication produce different standard errors than the original paper?

Stata replications diverge from original papers due to sample restrictions, variable construction differences, standard error adjustments, degrees of freedom changes, and command-default differences, which require funnel debugging to investigate.

Can I use this replication protocol to audit any published econometrics paper?

Yes, the replication protocol applies when auditing a published paper, onboarding a new analysis by replicating first, or validating a method on a known benchmark, provided you can translate the analysis into a Stata pipeline.

What's the best way to verify a paper's ATT estimate and first-stage diagnostics?

The best way to verify an ATT estimate and first-stage diagnostics is to replicate the targeted tables using gold-standard targets with MUST/SHOULD/MAY tiers, producing a structured report indicating which estimates passed within tolerance.

How do I start a replication of a specific paper using the command line?

Start a replication by running the replicate command with a paper's short name or target file, such as /replicate AbadieDiamondHainmueller2010, to execute the end-to-end Stata pipeline protocol.