A senior peer reviewer reads your draft and writes the report that decides whether it goes back for major revisions. Five required sections: central claim (in one sentence), contribution (what's new), what's working, what needs work (3–5 specific issues, each anchored to a quote), and the top three priorities ordered by impact.

A logical fallacy is a specific, named error in reasoning — Strawman, False dichotomy, Cherry-picking, Post hoc, etc. The check uses a 28-fallacy canonical taxonomy and flags only what fits one of those names. Every finding includes a verbatim quote, the broken inference step, a severity (minor / serious / critical), and a one-line repair. If your draft has none, the check says so plainly rather than inventing.

A weak argument is a claim your draft *needs* (load-bearing) that isn't well-supported on its own terms — insufficient evidence, unstated assumption, equivocation, sample-selection problems, or a conclusion broader than the evidence supports. The check ranks by load-bearingness × weakness, so the worst case (a critical claim that's also poorly supported) shows first.

Counter-arguments are the objections a thoughtful, fair-minded reader would raise to your central claim — not strawmen. The check produces 3–4 steelmanned objections with the reasoning behind each, plus what your draft would need to do to address them: concede, refute, narrow scope, or cite additional evidence.

Concrete suggestions are the kind of edit you'd actually mark in the margin of a draft — never "be more concise". Four buckets: structural (what to reorder, merge, cut), argumentative (what evidence or comparison is missing), methods/scope (what's unspecified for replication), prose-level (sentences to rewrite right now). Every item names a specific quote and a specific fix.

The clarity check finds the passages where a careful reader is most likely to stumble: an ambiguous pronoun reference, an undefined jargon term, two ideas compressed into one sentence, an inference that depends on an unstated premise, or a concept introduced before the prerequisite is established. Each finding includes a quote, the specific source of confusion, and the fix.

Five concrete next moves when you're stuck.

Treat the draft as a pitch — is it worth writing?

What is *new* in this paper? Three sharp candidate framings.

Limitations the author should acknowledge — calibrated to what the draft actually claims.

Could a competent researcher in your field reproduce what you did from this text alone?

Three alternative opening paragraphs in different registers.

Tighten the central argument into one strong sentence.

Three alternative endings — different shapes.

30% shorter, every claim preserved.

Formal register, qualified claims, citation hooks.

Plain language, fewer technical terms.

Five distinct titles in different styles.

Three single-sentence summaries — different angles.

Standard 4-section abstract suitable for journal submission.

5-7 post thread distilling the core argument.