Stress-Test Research Papers Before Submission with PaperJury
Use PaperJury when a Claude Code workflow should review a LaTeX research paper, classify reviewer-style objections, revise bounded issues, and verify the result before submission.
npx skills add agentskillexchange/skills --skill stress-test-research-papers-before-submission-with-paperjury
Use PaperJury when a research-paper author or assistant needs an adversarial pre-submission review workflow inside Claude Code. The operator installs the PaperJury skill or plugin, points the agent at a paper project, runs review, direct-edit, or auto mode, and lets the workflow produce evidence-linked issues, verdicts, bounded patches, and verification checks before any final changes are accepted.
What this skill actually does
Invoke this instead of using Claude Code normally when the paper needs reviewer-style pressure testing rather than generic rewriting: claims should be challenged, issues should be classified as valid or invalid, risky edits should require confirmation, and missing experiments or unsupported claims should be handed back to the author.
The scope boundary is pre-submission review and bounded LaTeX revision for research papers. This is not a general academic research assistant, literature-review tool, or writing product card; it is a Claude Code skill with a specific review, verdict, revise, and verify loop.
Inputs and prerequisites: Claude Code; PaperJury Claude Code plugin or skill files; a local LaTeX paper project; npm for doctor checks where used..
Setup notes: Install in Claude Code with /plugin marketplace add Spark-To-Paper-Skills/paperjury and /plugin install paperjury@Spark-To-Paper-Skills. Alternatively, clone https://github.com/Spark-To-Paper-Skills/paperjury into ~/.claude/skills/paperjury or a project .claude/skills/paperjury directory, then run npm run doctor inside the skill directory if you need an integrity check.
Source and verification boundary: use https://spark-to-paper-skills.github.io/paperjury/ as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a Claude Code workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.