paper-review-lite
Pre-submission audit: argument, numerics, refs, writing, figures, replication.
Profile →npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill paper-review-lite --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
What it does
Runs a comprehensive pre-submission review of an academic paper using parallel sub-agents, each examining a different dimension (argument, numbers, references, writing, figures, replication, etc.). Each sub-agent must cite a direct quote from the manuscript for every critical findings, attack the argument or data rather than authors, and produce structured outputs to be saved in a scratch directory. The final output is a structured pre-submit report with severity-ranked findings and a journal-readiness checklist.
How it works
- The skill delegates to ~11 sub-agents operating inside a Claude Code session.
- Each sub-agent examines a distinct dimension (content/argument, numbers, references, DOI, writing, CONSORT/randomization, pre-registration, figures/tables, replication).
- Sub-agents run in parallel and write their findings to dedicated files under a scratch directory .review-tmp/ with names agent-1-content.md, agent-2-numbers.md, etc.
- Findings must include: Severity, Location, Issue, Fix, and a Quote (verbatim span from manuscript) for each CRITICAL/RECOMMENDED finding.
- Cross-checkers will later read the produced files to synthesize the final cross-check outputs.
When to use it
Use this skill for a fast, in-session pre-submission audit of an academic manuscript to identify high-impact, fix-before-submission issues. It is intended as a lightweight companion to a fuller reviewer2/presubmit pipeline, not a replacement for deeper adversarial checks.
What it can touch
- Assumes access to the manuscript text and its structure to extract quotes.
- Writes findings to .review-tmp/ subdirectory, with files named agent-1-content.md, agent-2-numbers.md, agent-3-references.md, agent-4-doi.md, agent-5-writing.md, agent-6-consort.md, agent-7-pre-registration.md, agent-8-figures.md, agent-9-replication.md.
Caveats
- This is described as a fast in-flow check, not the full multi-stage reviewer2/presubmit pipeline.
- Results depend on the manuscript being accessible in the environment so quotes can be drawn directly.
- The output files are written to a scratch directory (.review-tmp/) during Phase 2, and cross-checks rely on exact formatting of the provided findings.
# Paper Pre-Submission Review (Lite) ## Heritage and scope This is the in-session, Claude-Code-native counterpart to [`presubmit`](https://github.com/scdenney/presubmit) — our port of the [reviewer2](https://github.com/isitcredible/reviewer2) adversarial peer-review pipeline to Anthropic Claude. The design inherits two things from that lineage: 1. **A Critical-Reviewer posture.** Review sub-agents
What does the paper-review-lite skill do?
Pre-submission audit: argument, numerics, refs, writing, figures, replication.
How do I install it?
Run `npx skills add brycewang-stanford/Auto-Empirical-Research-Skills --skill paper-review-lite --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Auto-Empirical-Research-Skills, a repository with 3,244 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.