agentsclimarketplace

Ngeo referee strategy

Skill brycewang-stanford/Awesome-Journal-Skills/Nature-Geoscience-Skills/skills/ngeo-referee-strategy

Use when choosing suggested and excluded referees for a Nature Geoscience submission and pre-empting the objections Earth-science referees are likely to raise. Plans referee handling; does not write the manuscript or the rebuttal.From its SKILL.md

Install
npx -y skills add brycewang-stanford/Awesome-Journal-Skills --skill ngeo-referee-strategy

Assembled from the repository path, not quoted from the project. Check it against their README if it does not work.

SKILL.md

6.2 KB, ~1.3k tokens by cl100k_base, as published. Nobody here has run it

Nature Geoscience Referee Strategy (ngeo-referee-strategy)

When to trigger

  • The Editorial Manager form asks for suggested / excluded reviewers
  • You want to anticipate the objections a Nature Geoscience referee will raise and address them in advance
  • A prior manuscript was declined and you are rethinking who reviews it
  • You are coordinating the cover-letter referee section (see ngeo-cover-letter)

How Nature Geoscience review works (durable shape)

An in-house editor first triages on broad interest and Earth-system importance; only a minority of submissions are sent out. Those that pass go to referees (commonly two to three) who assess both technical soundness and significance/breadth. A referee can recommend rejection on interest grounds even when the work is correct — so your strategy must defend the science and the venue fit. Review is single-blind by default, with an opt-in double-blind route (choose it at submission and anonymise the manuscript accordingly). Resubmission and appeal routes exist (handled in ngeo-revision).

Choosing suggested referees

PrinciplePractice
Competent on the methodCan actually judge your instruments, proxy, or model
Spans the breadth claimInclude someone from the neighboring Earth-science field you claim to reach
Covers the evidence typeIf it is field + model, cover both the observational and modelling sides
No conflict of interestNot recent collaborators, advisors, or same-institution
Constructive, not partisanKnown for fair, rigorous reports — not reflexive gate-keeping
Grasps the significanceWill recognise why the result is broad, not just correct

Suggest a small set that together can vouch for correctness, the evidence, and breadth. Do not stack suggestions only with subfield insiders — editors may read that as evading the broad-interest test. Consider global diversity of expertise; Earth-system questions often need both a regional expert and a global-perspective referee.

Choosing excluded referees

  • Exclude only with a brief, professional, factual rationale (a direct competitor with a conflict, a documented dispute). Editors discount emotional or vague objections.
  • Do not over-exclude; a long exclusion list signals defensiveness. Nature journals typically cap the number you may exclude.

Worked example: a referee slate

Suppose the Article reports a proxy-based ocean-circulation reconstruction whose headline consequence bears on climate sensitivity:

  1. Method judge — a paleoceanographer who runs the same proxy and can interrogate the calibration and age model.
  2. Modelling witness — a climate modeller who can assess the data–model comparison and the circulation inference.
  3. Breadth witness — a climate-dynamics scientist who answers "why Nature Geoscience and not a paleo community journal," i.e. why it bears on a first-order question.

The three-role logic transfers to any subfield: one referee for the measurement or model, one for the neighboring field you claim to reach, one independent senior figure. If you cannot name a credible breadth witness, that is diagnostic — revisit ngeo-scope-fit.

Pre-empting the likely objections (do this in the manuscript, not just the letter)

Anticipate the standard Nature Geoscience referee moves and disarm them in advance:

  • "This is regional / not broad enough." → Strengthen the broad-interest framing in the opening and cover letter (ngeo-results-framing, ngeo-cover-letter).
  • "The model is not grounded in observations." → Show the data–model validation in the main text (ngeo-methods).
  • "Uncertainty is not quantified." → State propagated uncertainty on the headline number (ngeo-methods).
  • "Incremental over Ref. [X]." → Make the qualitative Earth-system advance explicit; cite and differentiate.
  • "Can't follow it without the Supplementary." → Ensure the main text stands alone (ngeo-supplementary).
  • "Over-claimed attribution/impact." → Calibrate claims to evidence (ngeo-writing-style).

Checklist

  • Suggested referees can judge correctness, the evidence type, AND breadth
  • At least one suggested referee represents the claimed neighboring field
  • Observational and modelling sides both covered if the study is mixed
  • No conflicts of interest among suggestions
  • Excluded referees (if any) have a brief, factual, professional rationale
  • The exclusion list is short (within any journal cap)
  • The manuscript already pre-empts the top objections (breadth, grounding, uncertainty)
  • Cover-letter referee section is consistent with the Editorial Manager form
  • Double-blind opt-in decision made and manuscript anonymised if chosen

Anti-patterns

  • Suggesting only close-network insiders (reads as breadth evasion)
  • Excluding referees with emotional or unsubstantiated reasons
  • A long exclusion list that signals defensiveness
  • Suggesting a recent collaborator (conflict) as an "independent" referee
  • Ignoring the breadth or grounding objection and hoping the science alone carries it

Output format

【Suggested referees】N — cover correctness + evidence + breadth?  yes / adjust
【Neighboring-field referee included】yes / add
【Obs + model sides covered】yes / n/a / add
【Conflicts checked】clean / fix
【Excluded referees】N (short) — factual rationale?  yes / fix
【Top objections pre-empted】breadth / grounding / uncertainty / increment / stand-alone / over-claim
【Next】ngeo-submission (enter referees) → later ngeo-revision

Referee counts, exclusion caps, and double-blind mechanics evolve — verify on the official Nature Geoscience author pages (Checked: 2026-07-16).

What ships with it

Read from the repository

Just SKILL.md. No reference files, no scripts.

Keep looking

Skills are one crate of 325,949. Ordering is by how many stacks a row turns up in, so the top of any crate is what has actually been picked rather than what has the most stars.