Content
New
$10.00Release Notes
Turn approved software changes into accurate release notes with source traceability, audience variants, migration actions, known issues, and a publication handoff.
release noteschangelog
T
talvio
Updated Jul 25, 2026
Turn a scientific manuscript into a traceable referee report with prioritized findings, answerable revision requests, and a recommendation tied to the evidence.
A useful peer review does more than walk through a long checklist. It shows which claims the manuscript makes, what evidence supports them, where the chain breaks, and which unresolved issues actually control the recommendation.
This skill produces that decision trail. It classifies the study design, chooses reporting guidance only when it fits, checks methods and analysis against the central claims, and separates missing reporting from a verified flaw. Every major comment names a locator, the observed evidence, why it matters, the smallest adequate response, and the effect on the recommendation. The finished report includes a claim-evidence map and a revision-response matrix, so editors can make a decision and authors can answer without decoding vague reviewer prose.
Use it for research articles, preprints, methods papers, systematic reviews, or revision rounds. It does not verify raw data, replace specialist statistical or ethical review, invent venue policy, or promise acceptance.
Input
Review this randomized trial manuscript for a journal that uses accept, minor revision, major revision, and reject. Check whether the primary claim follows from the registered outcomes, attrition, analysis population, effect estimates, and adverse-event reporting. Return an editor-ready report and author response matrix.
Output
A neutral study summary, claim-evidence map, recommendation controlled by the outcome-switching and analysis-population findings, numbered major comments with exact locators and minimum sufficient responses, minor corrections, questions that could change the verdict, and a revision-response matrix.
Input
Review this cross-sectional study linking social-media use with adolescent sleep. The abstract says heavy use causes poor sleep, the model reports odds ratios, and the supplement is missing. Keep revision requests proportionate to what this design can establish.
Provide the manuscript or the sections available, the study type or discipline if known, the target venue and recommendation scale if available, and any supplement, protocol, registration, data, code, or author response. Ask for a full referee report or a scoped review. The skill maps central claims to evidence, prioritizes findings by decision impact, requests the minimum sufficient repair, and derives a recommendation without guessing about missing material.
The manuscript, preprint, revision, or specific sections to review, with stable page, line, table, or figure locators where possible.
Study discipline, manuscript type, intended venue, recommendation labels, scope, and deadline when known. Missing context can remain explicit.
Available supplement, protocol, registration, reporting checklist, data, code, prior decision letter, and author response. The skill records anything not supplied as a limitation.
SKILL.md; design-aware scientific review rubric; decision-ready referee report template; Agent Skills interface metadata
No reviews yet.
Output
A design-calibrated review that separates verified reporting from missing material, traces causal and risk-language claims to the actual model, identifies confounding and missing-data limits, asks for bounded corrections or sensitivity checks, and avoids demanding an unrelated randomized trial.
Input
I only have the abstract, two result tables, and a request to recommend acceptance. Review what is here and tell me what an editor can decide now.
Output
A scoped readiness assessment that marks methods, protocol, analysis details, and supporting files as missing; reports only contradictions visible in the supplied tables and abstract; declines a whole-paper acceptance recommendation; and lists the minimum material needed for a full review.
Creator
Kkelvio4