Copy this prompt into the AI assistant you're using.
Follow https://skillsmp.com/skill-install/prompt.md to install Agent Skills from https://github.com/EveryInc/compound-writing.Loading the repository README…
Showing 33 of 33 collected skills.
Run a multi-round deliberation between reviewers. Unlike /cw-panel (which synthesizes), /cw-debate has reviewers respond to each other's arguments across rounds until tensions resolve or reach acknowledged stalemate.
Review a draft for big-picture issues in argument, structure, stakes, evidence, and payoff. Use when a piece needs developmental editing, a top-level diagnosis, or substantial restructuring before sentence polish.
Compose existing Compound Writing skills and runtime capabilities for open-ended writing outcomes that no single skill covers, such as cross-draft analysis, source-to-outline comparison, or multi-pass revision.
Generate 3 hook options tailored to your content. Invoke with /cw-hook or /cw-hook [topic/context].
Reads your writing through the meanest, least-charitable lens. Challenges every claim, questions every assumption, pokes holes in the logic. Use for later drafts when you want hard feedback.
Surfaces counterarguments, objections, and "yeah buts" a reader might have to your argument. Helps you anticipate and address resistance before it derails the piece.
Begin or refresh the separate VOICE.md, STYLE.md, and optional AUDIENCE.md guides in a writer's durable writing home from a short conversation, existing writing, or maintained context. Use during first-run writing-home setup and when the user wants to define,…
Convene a panel of reviewers to analyze a draft from multiple perspectives, then synthesize their feedback into consensus findings, productive tensions, and prioritized recommendations.
Generate 3 promise options that create anticipation for what's to come. Invoke with /cw-promise or /cw-promise [topic/context].
Review copy as an intended reader encountering it for the first time, with no access to the writer's unstated context. Identify confusing language, missing setup, assumed knowledge, broken expectations, trust-eroding or off-putting moments, and points where…
Capture a confirmed writing preference, editorial lesson, recurring failure pattern, or workflow improvement in the right durable context file. Use when the user asks to save, remember, codify, or turn a repeated correction into a rule.
Route open-ended writing work through the smallest useful Compound Writing workflow. Use when the user wants to start, continue, finish, or improve a piece and has not named a more specific skill.
Create or initialize a durable writing home with VOICE.md, STYLE.md, examples/, and drafts/. Use when Scribe identifies a first-time writer with no established writing context after resolving the target, or when the user explicitly asks to set up, scaffold,…
Generate 3 thesis options tailored to your content. Invoke with /cw-thesis or /cw-thesis [topic/context].
Scan or clean prose for AI-generated residue, including unsupported connective tissue, epistemic inflation, over-completed arguments, performed voice, formulaic structure, empty figurative language, stock phrasing, corporate abstraction, hedging, and false…
Generate analogy options to explain a concept. Invoke with /cw-analogy or /cw-analogy [concept to explain].
Diagnose whether the most important idea in an article, essay, post, memo, section, paragraph, or sentence appears where it best serves the reader. Use for BLUF checks, buried or misplaced ledes, slow or misleading openings, intro drift, competing main ideas,…
Help users surface raw material for a piece when they don't have a specific idea yet. Invoke with /cw-brainstorm or when user says they want to explore ideas.
Turn an outline, notes, source material, or partial prose into a complete draft while preserving the writer's voice, thesis ownership, and source provenance. Use when the user asks to draft a piece or section.
Run a publication-readiness check on a nearly complete piece and deliver a clear ready, almost-ready, or needs-work verdict. Use when the user asks for a final pass, quality gate, or pre-publication check.
Cuts ruthlessly. Flags every adjective, adverb, and unnecessary word. Demands you kill your darlings.
Reviews for suspense and tension. Where's the bomb under the table? What does the reader know that someone doesn't?
Draw out and organize the writer's own thinking without supplying meaning, stakes, or an argument for them. Invoke with /cw-interview or when the user has a specific idea they want to explore before outlining.
Revise prose at the sentence and word level while preserving voice, meaning, source accuracy, and useful weirdness. Use when the structure is stable and the user asks for a line edit, polish, tightening pass, or usable rewrite.
Reads your writing from a loving, supportive, but not-quite-getting-it perspective. Surfaces where you've lost the general reader.
Organize interview material into a detailed structure using the 10%/30% framework. Invoke with /cw-outline or after completing an interview.
Finds the funny. Looks for moments that could be mined for humor, absurdity, or self-deprecation.
Rewrite complex text in plainer language. Invoke with /cw-simplify or /cw-simplify [paste text].
Checks pacing and momentum. Is there enough forward motion? Are you walking and talking or standing still?
Finds where the writing shows its tracks—scaffolding, throat-clearing, process narration, and arrival sequences that should be erased now that the destination is clear.
Generate 3 context-aware transition options between sections. Invoke with /cw-transition or /cw-transition [from section X to section Y].
Diagnose whether writing matches the active writer, project, publication, and platform voice context, then offer a closer revision. Use when the user asks whether text sounds like them, feels generic, or has drifted from an established voice.
Applies Vonnegut's 8 rules for writing. Start close to the end, give characters wants, be a sadist, etc.
Repository README source: README.md
Read the source README