draft-detective

Use this skill to find and catalogue every abbreviation and acronym occurrence in a document — recording, for each occurrence, any accompanying inline definition, how many times the abbreviation has appeared so far, where it appears, the definition listed in any Abbreviations section, and whether the occurrence is exempt from compliance checks. Invoke when asked to extract, list, or inventory the abbreviations/acronyms in a document (e.g. as input to an abbreviation compliance check).

Use this skill to check that a document's abbreviations and acronyms follow publication style — each is defined inline at its first use, listed in a dedicated Abbreviations section, used consistently, and not given conflicting meanings. Invoke when asked to review or validate a document's abbreviations/acronyms for compliance (e.g. per a manual of style).

Use this skill to validate the author biography section of a research publication — that each author bio has the right sentence count, states position and affiliation, describes research focus, and names the highest degree. Invoke when asked to check a document's "About the Authors" / author biography section.

Use this skill to validate a research publication's preface / introductory section against publication requirements — that it establishes context, explains objectives, identifies the audience, situates the work in the literature, states the contribution, and defines scope. Invoke when asked to check a document's preface, introduction, or "About This" section.

Use this skill to check a document for passive-voice sentences and for sentences whose inanimate subject hides who is responsible for an action ("the evaluation will assess" when several organizations share the work). Reports each occurrence with a proposed active rewrite, or asks the author to name the actor when the text does not say who it is. Invoke when asked to check active voice, passive voice, or unclear actors in a document.

Use this skill to check that a document uses neutral, objective language. Flags trigger words (certainty language without evidence), advocacy language (unsupported recommendations or opinion framing), and subjective tone (value judgments or emotional language). Invoke when asked to review a document's tone, neutrality, or objectivity.

Use this skill as the entry point when the user gives a generic, unspecified request to use Draft Detective on their document — e.g. "run draft detective", "analyze my document", "review this document", "check my paper", "what can you check?", "what can you do?". It lists everything Draft Detective offers and asks the user what to run. Do NOT use it when the user has already named a specific check or operation (run that skill directly).

Use this skill to check whether an in-text citation is actually supported by the source it cites — deciding, for each cited claim, whether the cited source supports it, only partially supports it, does not support it (is silent on it or contradicts it), or cannot be checked. Invoke when asked to validate that a document's claims are backed by their cited references (claim / citation substantiation), given access to the cited sources.

Use this skill to check that a document contains all required top-level sections — About This, Acknowledgements, Methods, Results, Conclusion, References, and a conditional Appendix. Invoke when asked to verify a document's structure or that its required sections/content are present.

Use this skill to check that every figure and table in a document is properly titled/captioned, consistently numbered, referenced in the body text, and that all body-text references resolve to an actual figure or table. Invoke when asked to review or validate the figures and tables in a document for completeness and consistency.

Use this skill to analyze a document for logically invalid inferences — conclusions drawn but not supported by the premises, or reasoning containing logical fallacies. Runs three independent passes and consolidates them into a single, double-checked, severity-ranked list (quoting the key sentence and explaining each flaw). Invoke when asked to check the logical validity of a document's reasoning or arguments.

Use this skill to report document review findings with clear evidence, severity, and actionable guidance. Present findings naturally in chat and written reviews; use the standard issue schema when a workflow, tool, or user requires structured output.

Use this skill to perform a literature review of a document — surface high-quality academic sources the document should cite or discuss but doesn't, both supporting and conflicting, using web search. Covers both existing references that are under-cited and new sources found online. Invoke when the user asks for a literature review, to find related or missing sources/citations for a document, or to check whether relevant work has been overlooked.

Use this skill to produce a "live report" addendum for a document — find high-quality literature published AFTER the document's publication date that updates, challenges, supports, or adds context to its claims, using web search, and recommend what the authors should update. Invoke when the user asks whether a document's findings have been updated or contradicted by newer research, or wants an addendum of newer evidence.

Use this skill to compare a paper's methodology against standard practice in its scientific field (methodological alignment) — using web search to characterize the field baseline, then assessing similarities, differences, missing components, rigor/risks, and improvements. Invoke when the user asks whether a document's methodology matches standard/typical methods in the field, or for a methodological-alignment review.

Use this skill to extract a detailed, reproducible description of the methodology a research document used to obtain its results, and to classify how reproducible that methodology is. Invoke when the user asks to extract, summarize, or describe a paper's methods/methodology, or to assess whether a study's methodology is reproducible.

Use this skill to check that each recommendation in a document is directly supported by the document's own findings, classifying it as supported, partially supported, or unsupported. Invoke when asked to verify that recommendations are backed by the evidence presented in the same document.

Use this skill to locate and download the full original content of a bibliographic reference from the web — given a citation, find a direct URL to the full text (not an abstract or metadata page), download and save it, verify it matches the reference, and report the outcome. Invoke when the user asks to download, fetch, or retrieve the full text/PDF of one or more references or citations. Handles one reference at a time; for several, apply the procedure to each.

Use this skill to find and extract all bibliographic references from a document's reference/bibliography section, returning the complete text of each entry. Invoke when the user asks to extract, list, or pull out all the references, citations, or the bibliography from a document.

Use this skill to validate a bibliographic reference (journal article, preprint, book, webpage, press release, government report, etc.) by searching for it online and comparing the citation to authoritative sources. Invoke when the user asks to check, verify, fact-check, or validate a citation or list of references — confirming author, title, publisher, year, and identifier (DOI / arXiv ID / ISBN / ISSN) against the actual published work.

Use this skill to extract a research document's main results and classify how reproducible each one is (fully reproducible, reproducible with web search, reproducible with external uploads, or not reproducible) from the information in the paper. Invoke when the user asks whether their results could be reproduced from the document alone, or to assess the reproducibility of a paper's findings.

Helps authors work through a peer-review cycle: making sense of what the reviewers asked for, planning the revision, and writing the paperwork that closes the loop with the reviewers and the quality-assurance manager. Use it whenever someone has a draft plus one or more reviewer memos (also called review reports, referee reports, or reviewer comments) and needs to act on them: to understand or organize the feedback, plan revisions from it, respond point by point, or check whether a revised draft addresses what was raised. Produces, singly or together: a revision-planning summary, one response memo per reviewer, and a consolidated reviewer coverage report for a QA manager. Trigger even when the user does not say "skill" or name an output: "respond to these reviewer comments", "draft a response memo to the reviewer", "did my revision address the review", "what does this reviewer want me to change", "check my responses cover everything the reviewers raised", "help me deal with this referee report".

Use this skill to produce a rigorous peer review and a devil's-advocate rebuttal of a research document, in the persona of a demanding senior researcher ("Reviewer 2"). Invoke when asked to peer-review, critique, or stress-test a draft paper or report and you want a structured review (summary, strengths, weaknesses, prioritized next steps) plus a separate rebuttal.

How Draft Detective writes. Use this skill for the wording of any prose a person will read: review findings and issue descriptions, reports, memos, summaries, recommendations, and ordinary chat answers. It is a companion skill rather than a task: load it alongside whichever skill is handling the work, before drafting, and let it govern how the output reads. It covers the shared baseline (objective, plain, anchored in the text), the critique voice for when you are the one naming problems, the author voice for when you are drafting words the user will send as their own, and the generic-AI tells to avoid (sycophancy, hype, empty hedging, filler, exclamation marks, em dashes). Reach for it whenever output quality depends on sounding like an experienced research colleague instead of a chatbot, or when someone asks you to make a draft answer sound less like AI.