auto-paper-improvement-loop

To automate LaTeX paper review and recompilation, this skill runs a two-round loop of LLM critique, targeted LaTeX source edits, and latexmk recompilation to iteratively polish academic manuscripts without manual intervention.

An autonomous paper improvement loop is a process where an LLM independently reviews academic writing, applies targeted fixes to LaTeX source code, and recompiles the document, using fresh review threads to prevent confirmation bias.

Yes, you can freeze your paper structure during automated LaTeX refinement by applying an optional edit whitelist, which restricts specific operations like citation additions to maintain document integrity during resubmission.

To run automated academic paper refinement, you need python3, latexmk for recompilation, pdfinfo for PDF inspection, jq for data processing, and integration with GPT-5.6-Sol to power the LLM-based review.

Reviewer independence prevents bias in automated academic review by using fresh, context-naive review threads for each evaluation cycle, ensuring the LLM evaluates the manuscript genuinely rather than confirming its own previous edits.

Automated multi-round refinement is used for research manuscripts to ensure theoretical rigor and presentation excellence before submission, catching structural flaws that a single review pass might miss.