mhe-identification — Claude Code skill: install and usage
mhe-identification evaluates causal identification strategies and drafts endogeneity sections for empirical papers using principles from applied econometrics.
no license ★ 240 updated 2026-09-18
npx skills add franklee16/academic-research-skills --skill mhe-identification -a claude-codeWhat mhe-identification does
This skill brings the empirical econometrics mindset from Mostly Harmless Econometrics directly to research audits. It inspects whether an empirical claim genuinely isolates exogenous variation rather than reporting partial correlations. The evaluation checks potential outcomes, bad controls, instrument strength via first-stage F-statistics, parallel trends in difference-in-differences setups, and recent staggered estimation issues.
Two operational tracks are available. Running the critic mode scrutinizes identification arguments and audits standard error clustering choices. Running the drafter mode produces an identification section styled for academic working papers, respecting conventions from working-paper-format.md. When invoked without an explicit mode argument, it inspects the draft as a critic.
What it helps with
- Audit causal identification setups for bad control variables
- Detect TWFE bias in difference-in-differences designs with staggered timing
- Draft endogeneity sections using formal econometric notation
- Verify standard error clustering levels against assignment units
Example prompts for the agent
- /mhe-identification critic Review the identification strategy in this staggered DiD draft
- /mhe-identification draft Write the endogeneity section for an IV regression model
- Evaluate whether the parallel trends assumption holds in this empirical setup
How to install mhe-identification
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill mhe-identification -a claude-code2Manually: put the files in place
~/.claude/skills/mhe-identification/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/mhe-identification/
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill mhe-identification -a codex2Manually: put the files in place
.agents/skills/mhe-identification/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill mhe-identification -a cursor2Manually: put the files in place
.cursor/skills/mhe-identification/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install franklee16/academic-research-skills review-revise/mhe-identification/SKILL.md2Manually: put the files in place
.github/skills/mhe-identification/SKILL.md1Skills CLI (Vercel)
npx skills add franklee16/academic-research-skills --skill mhe-identification -a gemini-cli2Platform CLI
gemini skills install https://github.com/franklee16/academic-research-skills --path review-revise/mhe-identificationManually: put the files in place
.devin/skills/mhe-identification/SKILL.mdThe legacy .windsurf/skills/ path still works
Claude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about mhe-identification
How do I install mhe-identification in Claude Code?
Run npx skills add franklee16/academic-research-skills --skill mhe-identification -a claude-code in your terminal. You can also place the configuration file manually at ~/.claude/skills/mhe-identification/SKILL.md.
What does mhe-identification do?
It audits empirical research papers through an applied causal inference framework and drafts endogeneity sections. The tool flags omitted variable bias, checks instrumental variable relevance, and reviews difference-in-differences assumptions.
Which operating modes does mhe-identification support?
It supports a critic mode for reviewing empirical designs and a drafter mode for writing paper sections. If the query does not declare a mode, the skill defaults to critic mode.