prose plugins

by lulucatdev

Enforce serious, scholarly, textbook-grade written register across Codex agent output. Reinjects the constraint after startup, resume, clear, and context compaction. Provides a style profile, core and extended imitation examples, and an active review skill for prose audit and rewriting.

by Lauri Gates

Prose transformation and style control - synthesis, distillation, tone, voice, clarity, and consistency for written text

by Melodic Software

Detects and removes AI-writing tells (slop) in tracked markdown prose: em dashes, emoji formatting, AI vocabulary, negative parallelisms, chatbot phrases, filler, stacked hedging, citation artifacts, and model-era phrases, from a catalog based on Wikipedia's Signs of AI writing plus an evidence-graded model-vocabulary inventory. Read-only audit by default (deterministic detector plus judgment rubric); an explicit fix action rewrites findings behind a semantic-diff guard.

by Melodic Software

Finds prose in tracked markdown that restates content an external source owns (vendor docs, blogs, articles) without attribution, confirms the source, and converts the copy into a pointer, a citation, or a dated stamped record. Documentation attribution, not software supply chain. Scripts do only mechanical work; judgment is LLM work. Read-only audit by default; explicit fix and sweep actions apply changes behind a semantic-diff guard and live pointer checks.

by Melodic Software

Prose a scanning reader can use. /writing:be-concise puts the bottom line first, cuts words the meaning does not need, and keeps structure scannable and tone factual. Bare, it sets a standing posture; given a target, it reshapes it and reports word counts. For tickets, PR descriptions, changelogs, READMEs, and status updates. Never drops a decision, number, ask, error, or warning. Paraphrases NN/g, GOV.UK, US plain-language guidelines, Google and Microsoft style guides, and BLUF.

by Uday Sharma

Rewrite prose so it reads like a person wrote it. Uses research-informed structural editing, genre profiles, and a bundled deterministic pre-output gate. The gate separates hard typography failures from advisory editorial findings and never claims an authorship or detector verdict.

by Wildcat Labs

Receipted delivery through merge and closure, with ranked frontier, audit, fuzzing, prose, specification, debugging, hardening, telemetry and measurement skills.

by Jason Matthew

Positive writing system for durable written artifacts — READMEs, ADRs, PR descriptions, docs, runbooks. Reader-expectation prose, Diátaxis document jobs, plain-language structure.

by Andrew Vasilyev

Holds everything Claude Code writes for you to one writing standard: the answer first, a reason for every rule, one term for one thing, and plain English for a reader who learned it as a second language. It keeps up to 700 characters in context on every turn, and loads a skill of about 5,000 tokens each time you write.

by Andrew Vasilyev

Stops Claude Code from publishing a commit message, pull request, issue, comment or release note that uses a stock idiom from a fixed list, a line of bold text in place of a heading, or text hidden in a file the gate can't read. A program checks each by exact match and quotes what it found, so it costs no model call and keeps nothing in context.

by Ivan Gudak

Pluggable prose style checker and writing aid for documentation, planning documents, and release notes. Ships a vendor-neutral baseline — terminology, word list, voice and tone, grammar, formatting, UI interactions, inclusive language, and prose accessibility — distilled from the Microsoft, Google, and Apple style guides, Chicago, the Conscious Style Guide, and WCAG. Point it at your own organization's style guide and the overlay augments or overrides that baseline; every resolution miss falls back silently. Provides /prose-review-pr, /prose-review-docs (with --fix), /prose-style-refresh, a prose-style-checker agent, a prose-fixer agent, and a prose-style-rules skill.

by Simon Heimlicher

Prose craft through a matched triple (/architect-prose, /author-prose, /audit-prose) over caller-supplied kind layers: interface, documentation, copy

by Simon Heimlicher

Prose craft through a matched triple (/architect-prose, /author-prose, /audit-prose) over caller-supplied kind layers: interface, documentation, copy

by Simon Heimlicher

Prose craft through a matched triple (/architect-prose, /author-prose, /audit-prose) over caller-supplied kind layers: interface, documentation, copy

by Alexander Kurganov

Engineering prose standard for everything an agent writes: user replies, prose, code comments, review comments, commit messages, change descriptions, error messages and agent instructions. Five layers — sentence structure, plain vocabulary, comment hygiene, artifact formats, slop pruning — each stated as a rule a reader can check rather than a preference. The core rules are one file, rules/prose-discipline.md, injected at session start by the hook in hooks/ on hosts that run plugin hooks, and loaded as an always-apply rule by hosts that read plugin rule files; the skill carries the protocol and the depth behind every rule, including the substitution table, the Conventional Comments labels and a calibration corpus. Code blocks, identifiers, commands, URLs and quoted text are never flagged, and established domain terms are kept. No credentials, no MCP servers, no settings files, and no checker script: enforcement is model judgement over rule texts a person can read and argue with.

by Jordan Richlen

Route every response element to exactly one voice: prose to human-voice (verdict-first, confidence-tagged), machine-read artifacts to machine-voice (compressed traces, logs, status lines); plus ai-writing-mistakes, a wording pass that strips the tells marking prose as machine-written, and second-opinion, an offer-only subagent validation pipeline that never runs unbidden.

by Thom Conaty

Strips performed-quality tics from Claude's prose. Ships an optional Dewormer output style plus a dewormer skill holding the full ~90-term caution list across nine families: anthropomorphised code-speak, self-important spotlighting, performative candor, unsolicited validation, adverb inflation, the em-dash reframe, hedging connective tissue, metaphor soup, and sign-off tics.

by Alastair Zeved

Two-agent editorial system: an editor agent applies five named schemata, an evaluator agent independently scores the output against measurable features before seeing the editor's trace, and they loop until a reader without the author's domain expertise can follow the reasoning. Built from cross-domain translation of forensic linguistics, species counterpoint, oral-formulaic composition, and collaborative voice reconstruction.

by Luis Fominaya

Writes English prose to a house style with the enlint rule set. Puts the style card in the agent's instructions, lints every answer and every document it writes for free, and rewrites flagged documents in the background on the harness's own cheap model.

by Silvan Büchi

Word and punctuation choice. Removes the vocabulary and punctuation patterns that mark writing as AI generated.

by blader

Rewrite AI-sounding text so it reads naturally without changing what it says.

by blader

Rewrite AI-sounding text so it reads naturally without changing what it says.

by Scott Hysom

Claude Code and Cowork. Learns a project's house prose style from edits already made, records it as prose-style.md with stable rule ids, and conforms the rest of the documents to it. Ships prose.py, so it must be installed as a plugin rather than proposed as a single skill.

by doyled-it

Narrative-nonfiction craft plus an AI-tells blacklist, for writing prose that reads human and stays sourced.