Review LaTeX Files

Run the comprehensive LaTeX review protocol focused on hardcoded numeric results.

Steps

1. •

   Identify files to review:

   • •If an argument is a specific .tex filename: review that file only
   • •If the argument is all: review all .tex files in latex/
2. •

   For each file, follow the review protocol below.

   • •Save report to quality_reports/[filename]_tex_review.md
3. •

   After all reviews complete, present a summary:

   • •Total issues found per file
   • •Breakdown by severity (Critical / High / Medium / Low)
   • •Top 3 most critical issues
4. •

   IMPORTANT: Do NOT edit any LaTeX source files unless auto-fixing (see Phase 2).

Review Protocol

Detect hardcoded numeric results in LaTeX manuscript prose and, when possible, automatically replace them with dynamically generated macros from the code pipeline.

Background

The project's code pipeline (R/Julia) generates .txt files containing \newcommand definitions in output/numbers/. The latex/Makefile sets TEXINPUTS to include ../output/numbers/, so the manuscript uses plain filenames:

\input{revenue_estimate.txt}
% Then in prose:
% a U.S. carbon tariff raises \$\revenueEstimate\ billion

Phase 1: Detection (Read-Only Scan)

Step 1: Build the Dynamic Value Registry

1. •Glob for output/numbers/*.txt and all \input{...} lines in .tex files
2. •Read each .txt file and extract \newcommand definitions
3. •Build a registry: {macro_name: value, ...}

Step 2: Scan Prose Sections Read the .tex files. For each line, skip if inside:

• •Math environments: equation, align, gather, multline, inline $...$
• •Table environments: tabular, table, longtable
• •Preamble (before \begin{document})
• •Comment lines, \ref, \eqref, \label, \cite commands
• •LaTeX layout commands: \vspace, \hspace, \setcounter, \setlength

Step 3: Classify Numbers

Always OK (never flag):

• •Section/equation/figure/table references
• •Years (1900--2100)
• •Structural descriptors: "12-sector", "N-country", "3-digit"
• •Numbers inside math or table environments
• •Page numbers, ordinals in prose

Suspicious (flag for review):

• •Dollar amounts: \$N.NN followed by "billion"/"million"/"trillion"
• •Percentages: N\%, N percent, N percentage points
• •Parenthetical numbers resembling standard errors: (0.042)
• •Numbers preceded by result verbs: "raises", "increases", "lowers", "reduces"
• •Decimal numbers in prose context

Step 4: Cross-Reference Against Registry

• •If an existing macro holds approximately this value: Critical -- macro exists but prose uses hardcoded number
• •If no matching macro: proceed to Phase 2

Phase 2: Source Tracing and Auto-Fix

Step 1: Trace the Source

1. •Grep code/ for the numeric value, nearby variable names, or contextual keywords
2. •Examine Makefiles to find which scripts produce output files
3. •Identify the specific script and line that computes the value

Step 2: Decide

• •Auto-fix if exactly one script clearly produces the value
• •Escalate to user if multiple candidates, no match, or external source

Step 3: Auto-Fix (When Source Is Clear)

1. •Add stats generation code to the source script
2. •Update latex/Makefile SOURCES
3. •Replace hardcoded number with macro in manuscript
4. •Rebuild and verify the generated number matches

Template State

If output/numbers/ does not exist AND code/ has no scripts, report:

No code pipeline found. Hardcoded number detection is informational only.

Report Format

# Hardcoded Number Review: [filename].tex
**Date:** [YYYY-MM-DD]
**Reviewer:** review-tex skill

## Summary
- **Auto-fixed:** N
- **Escalated:** N
- **Clean:** No hardcoded results detected

## Auto-Fixed
[List: line, old value, new macro, generated value]

## Escalated Issues

### Issue N: [Brief description]
- **File:** `latex/[file]:[line_number]`
- **Severity:** [Critical / High / Medium]
- **Hardcoded value:** [the number]
- **Context:** [surrounding sentence]
- **Candidate sources:** [script paths, if any]
- **Action needed:** [What the user should decide]

Severity Levels