science-research-writing-skills

Plugin science-research-writing-skills from marketplace bernard-ng in bernard-ng/skills. Commands read from the repo's manifest — 1 month ago.

Skills (61)

• apply-design-patterns

  Use when a recurring design problem resembles a known pattern, a pattern name is mentioned (Strategy, State, Command, Facade, Mediator, Observer, Adapter, Composite, Template Method, double dispatch), the user asks "should I use X pattern" or "what pattern fits", or a design is about to take a patte

• assign-object-responsibilities

  Use when a class does too much, a service or controller gathers data from passive objects and decides everything, behavior is spread across data holders and helpers, a duty has no obvious owner, or the user asks "where should this logic live", "who is responsible for", or "this class is too big". Ap

• build-ubiquitous-language

  Use when domain terms are fuzzy, overloaded, or translated between people, tickets, and code, when the same word means different things, when developers and domain experts describe the same rule differently, when naming a new domain class or method, or the user asks for a glossary, "what should we c

• choose-control-style

  Use when event handling or workflow code has a large controller, many tiny handlers, long if-else or switch chains on type or state, a coordinator that knows every detail, repeated decision logic, or decision ownership is unclear, or the user asks "where should this decision be made", "should this b

• define-shared-roles

  Use when several classes or candidates look alike, a client branches on the kind of object it holds, one interface should serve many implementations, inheritance is proposed to share code, a candidate must fit a framework as well as the domain, or the user asks "should this be an interface or abstra

• describe-object-collaborations

  Use when writing a design document, architecture note, README section, pull request description, or ADR that explains how objects, services, or components cooperate, when asked to "draw a sequence diagram", "document this design", "explain how these classes interact", or "make a diagram of the flow"

• design-aggregates

  Use when drawing consistency boundaries in a domain model, when an invariant spans several objects, when transactions touch too many records or lock too much, when deciding what a repository returns, when object creation is complex or can leave an invalid state, when ORM entities expose their whole

• design-object-connections

  Use when objects reach through other objects (a.b().c().d()), a client sends a stream of tiny requests, raw strings and numbers are passed between collaborators, a constructor or global is used to find a helper, many outsiders call into one neighborhood, or the user asks about coupling, the Law of D

• design-reliable-collaborations

  Use when object requests cross trust or system boundaries, a workflow has retries, timeouts, or partial failure, errors are only logged and swallowed, exceptions are thrown without a clear handler, validation is duplicated everywhere, or the user asks "how should we handle failure", "what if this ca

• design-variation-points

  Use when requirements mention future variants, plugins, extensibility, configurability, or "make it flexible", repeated changes touch the same behavior, a proposed abstraction is justified by flexibility, config flags are multiplying, or a change request might need a hook. Apply automatically to dec

• discover-object-roles

  Use when a feature has unclear objects or classes, a domain model just turns nouns into classes, a proposed class has no clear purpose, a service or module boundary is being drawn, or the user asks "what classes do I need", "find the objects", "model this domain", or "sketch CRC cards". Apply automa

• distill-core-domain

  Use when deciding where design effort, the strongest developers, or refactoring time should go, when generic code such as dates, permissions, or workflow crowds out the business heart, when a large codebase has no clear statement of what matters, when choosing between building, buying, or outsourcin

• domain-driven-design

  Use when a task involves a business domain with real rules and vocabulary, such as "model the domain", "domain model", "DDD", "entities and value objects", "aggregate", "bounded context", "core domain", "business logic is scattered across services", "anemic model", or before building a feature whose

• evolve-large-scale-structure

  Use when a domain model has grown into many modules and nobody knows where new code belongs, when each area solves the same problem differently, when people need a shared way to describe the whole system, when configurable roles or rules tempt the model into endless flags, when plug-in components mu

• frame-design-problem

  Use when starting a new feature, system, or module whose design is not yet clear, when requirements are vague, conflicting, or only user-facing, when asked "design a system for", "where do I start", or "what is hard about this", or before proposing any classes. Apply automatically at the start of no

• make-implicit-concepts-explicit

  Use when a business rule, constraint, process, or policy is buried in conditionals, flags, comments, or helper methods, when eligibility or validation logic is repeated or scattered across layers, when status fields and boolean combinations encode states, when a "process" has no object, when swappab

• map-bounded-contexts

  Use when more than one model, team, service, or legacy system shares a domain, when one concept such as Customer or Product means different things in different places, when a shared database or shared library couples teams, when integrating with an external or legacy system, when splitting a monolit

• model-domain-building-blocks

  Use when deciding whether a domain concept should be an entity, a value object, a domain service, or a module, when designing identity and equality, when a class is mutable but should not be, when "Service" or "Manager" classes hold all the logic, when associations are bidirectional or many to many,

• name-design-objects

  Use when naming or renaming a class, interface, module, service, or role, when names such as Manager, Handler, Helper, Util, Processor, Data, Info, or Record appear, when two things share a name, or when a name hides what the object really does. Apply automatically whenever a new design candidate is

• place-objects-in-layers

  Use when deciding where objects belong in a layered, hexagonal, pipeline, or framework-driven architecture, when a framework dictates control flow or hooks, when UI, application, domain, and technical code are mixed, when dependencies point the wrong way, or the user asks about "layers", "ports and

• refactor-toward-deeper-insight

  Use when every new requirement adds a special case to the same domain code, when domain experts keep correcting the model, when a design feels shallow or clumsy even though tests pass, when there is a temptation to rewrite or to add yet another flag, when a recurring problem looks like a known busin

• responsibility-driven-design

  Use when a task involves designing, reviewing, or refactoring how objects, classes, modules, or services divide work and cooperate, including requests like "design the classes for", "where should this logic live", "this class does too much", "model this domain", "CRC cards", "split this service", or

• review-object-design

  Use when reviewing a design, pull request, or class structure for quality, when asked "is this good design", "review my classes", "why is this hard to change", or "refactor this", when a class has dozens of methods or a controller gathers data and decides, when one change ripples through many classe

• shape-supple-design

  Use when a domain model is correct but hard to use, combine, test, or change, when callers must know implementation details or call order, when methods have hidden side effects, when a change in one class ripples through many, when interfaces are named for how they work rather than what they mean, w

• solve-revealing-design-problems

  Use when a design problem keeps resisting solution, each attempt exposes a new twist, stakeholders disagree on what "solved" means, the same part keeps producing bugs, or the user says "we're stuck", "this keeps getting harder", "no clean solution", or "wicked problem". Apply automatically to separa

• trace-object-collaborations

  Use when a use case crosses several objects, a client sends many low-level messages, handoffs or who-calls-whom are unclear, a design is about to be coded without a walkthrough, or the user asks to "walk through the scenario", "role-play the design", "simulate", or "draw the sequence". Apply automat

• academic-vocabulary

  Use when scientific prose needs more precise or natural academic wording, repeats generic verbs such as show, do, use, or get, sounds informal, softens or sharpens criticism, or uses a word whose field meaning is uncertain. Apply automatically when polishing text, and when choosing reporting verbs,

• active-passive-voice

  Use when drafting or editing sentences in which the actor matters, such as Methods text, cited versus own work, "we" and "I" decisions, agentless passives, or "this paper presents" constructions, or when a reader could not tell who did something. Apply automatically to any Methods section, and to th

• articles-determiners

  Use when a, an, the, or no article, or this, that, each, or another determiner, appears before a technical noun in scientific text, or when a reader could ask "which one?". Apply automatically on any language edit of Methods text and on text by writers whose first language lacks articles.

• causal-language

  Use when a draft says or implies that one thing caused, produced, led to, explains, or is linked to another, including words such as caused, due to, because of, effect of, resulted in, influences, associated with, or relationship, or when design and evidence do not clearly support the cause claimed.

• citation-attribution

  Use when claims about prior work, consensus, novelty, borrowed methods, or quoted wording appear in scientific writing, and when checking where citations sit in a sentence, who is credited, or whether a source supports the wording. Apply automatically while drafting any Introduction, Discussion, or

• claim-strength

  Use when a scientific claim may be stated too strongly, too weakly, or inconsistently across Abstract, Results, Discussion, and Title, including words such as prove, clearly, demonstrates, obviously, may, tends to, or suggests, or when a request asks for stronger, safer, or more cautious wording. Ap

• contribution-mapping

  Use when writing or checking how a study's results relate to existing research and what the study adds, including consistent with, extends, contradicts, novelty or "first" claims, achievement statements, implications, and applications in a Discussion, Conclusion, or Abstract. Apply automatically whe

• evaluative-language

  Use when results, figures, tables, or numbers are reported without saying what they mean, when "bare numbers" appear, when tone toward findings is flat, inflated, or inconsistent, or when describing achievements and surprising results. Apply automatically while drafting Results, Discussion, and Abst

• figure-table-commentary

  Use when writing or checking figure and table callouts, captions, or the prose that comments on a graph, table, image, equation, or other display, or when text says "as shown in Figure" without saying what is shown. Apply automatically whenever a draft cites a figure or table, or the user supplies a

• formulate-research-gap

  Use when a paper needs the reason the study exists, such as a research gap, unresolved problem, limitation of earlier work, contradiction, or testable prediction, or when asked to "justify the research", "state the problem", or "what is new here". Apply automatically after a literature summary and b

• frequency-language

  Use when a method or result says how often something happened or how consistently, with words such as always, usually, often, sometimes, rarely, never, regularly, or frequently, or when a step is described without saying how many times. Apply automatically to Results sentences about repeated trials

• grammar-and-meaning

  Use when editing scientific sentences for grammar, clarity, or "does this say what I mean", especially for non-native English writers, translated text, or sentences that look correct but may state the wrong claim. Apply automatically on any proofreading, language-polish, or "fix my English" request

• limitations-future-work

  Use when a draft mentions, hides, or omits problems, deviations, imperfect or unexpected results, small samples, instrument limits, or future work, in Methods, Results, Discussion, or Conclusion, or when text sounds apologetic, defensive, or evasive. Apply automatically while writing a Discussion or

• literature-review-positioning

  Use when summarising, organising, or reviewing prior research for an Introduction, literature review chapter, related work section, or Discussion comparison, including requests like "write the background", "summarise these papers", or "who did what". Apply automatically whenever text lists several c

• methods-reproducibility

  Use when auditing, expanding, or checking a Methods, Experimental, Procedure, or analysis description for missing, vague, or inconsistent detail, or when asked "could someone repeat this", "is my methodology clear", or "what am I missing". Apply automatically to any procedure text that uses words li

• modality

  Use when a sentence contains may, might, could, can, cannot, should, must, need to, have to, ought to, or similar modal verbs, or expresses possibility, ability, probability, certainty, advice, or necessity in scientific writing. Apply automatically in Discussion, Conclusion, Abstract, and Results i

• modifier-placement

  Use when a time, place, manner, degree, frequency, or prepositional phrase could attach to more than one word in a scientific sentence, or when a sentence has two or more adverbs, "with", "using", "after", "only", "respectively", or "previously". Apply automatically to dense Methods and Results sent

• noun-phrases-prepositions

  Use when a title, heading, caption, or dense sentence contains stacked nouns, long compound noun phrases, or prepositions such as of, in, for, with, by, on, from, and to that could change the meaning. Apply automatically when drafting or editing titles, headings, figure captions, abstracts, and Meth

• paper-consistency-audit

  Use when a complete or near-complete manuscript, thesis chapter set, or multi-section draft needs a cross-section review, or when asked to "check my paper", "does everything match", or "review before submission". Apply automatically before final polish and whenever sections were written or revised a

• paragraph-coherence

  Use when a section reads like a list of unrelated sentences, paragraphs run too long or are cut into one-sentence fragments, a paragraph lacks a clear point, or the order of ideas is hard to follow. Apply automatically when planning a section outline and when reviewing any draft at paragraph level.

• quantity-comparisons

  Use when reporting sizes, amounts, proportions, trends, differences, or comparisons, or when a draft uses words such as more, less, few, little, significant, substantial, slight, nearly, approximately, higher, or lower. Apply automatically to Results and Abstract text containing numbers, and to any

• science-research-writing

  Use for any request to plan, draft, restructure, review, or polish English scientific research writing, such as journal articles, theses, dissertations, manuscript sections, abstracts, titles, or edits for non-native English speakers. Start here, then route to the section and language skills automat

• scientific-conventions

  Use when scientific text contains units, symbols, abbreviations, acronyms, Latin phrases such as e.g. or et al., or Latin and Greek singular and plural forms such as data, criteria, phenomena, hypotheses, and bacteria, or when preparing text to meet journal style. Apply automatically as part of the

• scientific-style-editing

  Use when a scientific draft, section, or manuscript is ready for review, polishing, proofreading, or submission, or when asked to "edit", "improve the flow", "check my English", or "make it publishable". Apply automatically as the final pass after drafting any research text, and run it before return

• sequence-language

  Use when describing the order, timing, duration, or overlap of steps, events, or observations in a method or result, or when a draft relies on "then", "next", "after", "while", or "subsequently". Apply automatically to any procedure with several stages and any result that depends on when something h

• signalling-transitions

  Use when adjacent sentences or paragraphs in scientific writing feel disconnected, when connectors such as however, therefore, since, while, so, or conversely may misstate the relationship, or when a reader would not know what the next sentence is doing. Apply automatically to any draft with abrupt

• spatial-description

  Use when describing the physical arrangement, geometry, position, orientation, or connections of apparatus, instruments, samples, sites, or components, as in a Methods section, device description, or figure caption, or whenever text says where something is placed, attached, aligned, or located. Appl

• target-article-analysis

  Use when starting a paper, section, or edit for a specific journal or field whose conventions are unknown, when asked to "write like the journals in my field", match a journal's style, or check whether a draft follows field norms. Apply automatically before drafting any section when no sample articl

• tense-selection

  Use whenever scientific prose reports prior studies, a procedure, an aim, a finding, an interpretation, or a conclusion and the choice of past simple, present simple, or present perfect affects meaning, or when the user asks about tense. Apply automatically when drafting or editing any research sect

• write-abstract

  Use when writing, shortening, restructuring, or reviewing an abstract, summary, structured abstract, or synopsis of a research paper, thesis, or report, including requests to "summarise my paper", "cut this to 150 words", or "add keywords". Apply automatically once a manuscript or its main sections

• write-discussion-conclusion

  Use when writing, restructuring, or reviewing the Discussion, Conclusion, Conclusions, or closing section of a research paper or thesis, or any text that interprets findings, compares them with prior work, states a contribution, applications, limitations, or future work. Apply automatically when the

• write-introduction

  Use when writing, restructuring, or reviewing the Introduction, opening paragraphs, background, literature context, research gap, or aim statement of a research paper, thesis chapter, or report, including requests like "motivate my study", "write the first paragraph", or "state the aim". Apply autom

• write-methodology

  Use when writing, expanding, or reviewing the Methods, Materials and Methods, Experimental, Procedure, Simulation, or Model section of a research paper or thesis, or any text that says what was done, used, sampled, built, or analysed. Apply automatically to procedure descriptions, including requests

• write-results

  Use when writing, restructuring, or reviewing the Results, Analysis, or Data Analysis section of a research paper or thesis, or any text that reports findings, comparisons, trends, figures, or tables. Apply automatically when the user supplies numbers, a table, or a figure and asks to "describe", "r

• write-title

  Use when creating, shortening, checking, or choosing between titles, subtitles, or running heads for a research paper, thesis, or report, or when asked for keywords or a searchable name for a study. Apply automatically at the end of drafting a paper, and whenever a title contains stacked nouns, prep

Similar plugins