# Review Cover Letter

> Review a submission cover letter for a scientific manuscript, checking that it states the novelty concretely, gives a real reason for choosing this journal, includes the declarations the journal requires, does not overclaim relative to the paper, and fits on one page. Use when asked to review, critique or check a cover letter before submission.

- Skill: `yuryalencar/review-cover-letter` (Agent Skill)
- Install (CLI): `npx skillmds@latest add yuryalencar/review-cover-letter`
- Raw SKILL.md: https://api.skillmd.com/api/skills/yuryalencar/review-cover-letter/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: yuryalencar (https://skillmd.com/u/yuryalencar)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/yuryalencar/review-cover-letter

---


# /review-cover-letter

The first document the editor reads. Most cover letters fail in one of two ways: they state no concrete
novelty, or they give no real reason for choosing this journal. Both are fatal to the letter's purpose and
both are quick to fix.

## Before anything else

1. **Load the style rules.** Invoke `paper-writing-style`. If absent, work from the condensed rules in
   `create-cover-letter`.
2. **Lock the language.** Language of the user's first message. Rewrites are English.
3. **Read the paper too** if available. The overclaiming check is impossible without it, and it is the
   check with the most at stake.

## Checks

**1. Is the novelty concrete?** The letter must state what the work shows and why it advances the field, in
specific terms. Flag:
   - Novelty stated as a topic rather than a finding: "we investigated the cycling behaviour of graphene
     electrodes" says nothing an abstract does not.
   - Novelty asserted rather than shown: "this represents a significant advance" with no content.
   - No novelty statement at all, which is commoner than it should be.

Quote what is there and propose the concrete version using the paper's actual contribution.

**2. Is there a real reason for this journal?** This is the check that separates a working letter from a
template. Flag:
   - No mention of the journal beyond its name in the first line.
   - A reason that would apply to any journal: "your esteemed journal", "the high impact factor", "the
     wide readership".
   - Flattery in place of a reason.

A real reason relates the contribution to the journal's scope, its readership, or a specific recent paper
it published. If the letter has none, say so and ask whether the work builds on or answers something that
journal has published; that is the strongest possible content for this paragraph.

**3. Overclaiming.** Compare the letter's claims against the paper. An editor who finds the letter
overselling relative to the manuscript starts reading sceptically, and this is where the temptation is
greatest. Check whether a conditional result became unconditional, whether values match, and whether the
letter claims generality the paper does not establish.

**4. Required elements.** Report any missing:
   - Editor addressed, by name and title where known
   - Manuscript title and article type
   - Originality statement: not published elsewhere, not under consideration elsewhere
   - All authors have approved the submission
   - Declarations the journal requires: conflicts of interest, funding, ethics, data availability
   - Suggested or excluded reviewers, if the journal invites them
   - Corresponding author contact details

A missing mandatory declaration causes an administrative rejection that has nothing to do with the
science, so flag these prominently even though they are boilerplate.

**5. Length.** One page. An editor reading fifty submissions will not read two. If it runs over, say which
paragraph to cut, and it is almost always a paper summary that duplicates the abstract.

**6. Does it summarize instead of arguing?** A letter that walks through the paper section by section has
misunderstood its job. The abstract is attached. The letter argues for the paper.

**7. Style and register.** Formal without stiffness. Flag `we are excited to`, `we are pleased to present`,
`groundbreaking`, `paradigm shift`, and the AI tells: `delve`, `harness`, `leverage`, `underscore`,
`pave the way`, `plays a crucial role`. Flag em dashes.

## Deliver

1. Failing checks, most serious first: novelty and journal fit, then overclaiming, then missing
   declarations, then length and style.
2. Each with the text quoted and a concrete replacement.
3. **One rewritten letter**, with `[TODO: ...]` where the researcher must supply a declaration, an editor's
   name, or the journal-fit reason.

If the letter passes, say so briefly. Cover letters are short.

## Hard rules

- **Never invent a journal's scope, a recent paper in it, or an editor's name.** No network access, so you
  do not know these. Ask, or leave a marked placeholder.
- **Never fabricate a declaration.** Ethics approval, conflicts of interest, funding and data availability
  are legally and professionally consequential. Leave a placeholder and say what is needed.
- **Never invent a result or strengthen a claim** to improve the letter.
- **Never emit an em dash.**
- If the paper is unavailable, say which checks you could not run.