Odin Writing

Use when prose is drafted or edited, reviewed for AI tells, or the user asks to remove AI patterns, humanize, or add voice. Offers dedash and detool modes. Not for code debris: use deslop.

Use when a behavioral diff may have left docs or CHANGELOG stale, or when tidy targets stale code comments. Applies only safe fixes. Not for roadmap drift: use drift-detect.

Use when the user knows what they mean but cannot express it completely or clearly. Not for discovery, ideation, or style-only editing: use unslop for style.

Use when reviewing a prose plan, spec, PRD, requirements doc, design doc, or brainstorm, or invoking /doc-review. Not for collaborative drafting: use doc-coauthoring.

Use when documentation has drifted from code changes and the user asks to synchronize. Not for ADRs or architectural rationale (docs-and-adrs). Not for writing an isolated document (docs-writing).

Use when the user asks to create, rewrite, review, or polish an open-source README, including shop-window READMEs. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to write, restructure, or audit documentation with Diataxis types, or to draft a feature docs page or diagram source. Not for ADRs or architectural rationale.

Use when the user supplies French text that reads like AI output and asks to naturalize it. Not for English STE rewriting: use humanizer-en-asd-ste100.

Use when a grounded piece needs user-selected beat-by-beat assembly of verbatim beats. Not for unstructured capture: use writing-fragments. Not for paragraph shaping: use writing-shape.

Use when shaping a source document paragraph by paragraph without modifying it. Not for selected-beat assembly: use writing-beats. Not for fragment capture: use writing-fragments.

Use when the user just edited a durable artifact, or says "clean and true", "run the hygiene pass", or "taste the output". Routes the artifact through the hygiene passes it earned.

Use when the user says "give this rhythm", "vary the structure", or "the sections all read the same". Not for judging an artifact's overall taste: use taste.

Use when the user asks for branded or style-governed output. Not for remote, credential, publish, deploy, or irreversible changes.

Use when drafting a doc, proposal, spec, RFC, design doc, decision doc, or PRD in chat. Not for reviewing an existing plan: use doc-review. No source or remote-system changes.

Use when the user runs /markdown-to-pdf on Markdown to render a publication-quality PDF. Not for remote, credential, publish, deploy, or irreversible changes.

Use when a human asks to purge stale docs, clean Markdown, or reorder the documentation hierarchy. Not for code debris: use deslop.

Use when a trainer wants a training session or workshop documented as a compte rendu with a final .docx. Not for remote, credential, publish, deploy, or irreversible changes.

Use when a document or file has been edited piecemeal and reads as sediment. Not for re-deriving a code subsystem contract: use breaking-driven.

Use when technical prose needs writing, reviewing, or draft editing with real symbols and controlled English. Not for articles or blog posts with external references: use technical-article-writer.

Use when exploration needs heterogeneous noticings captured before structure. Not for beat-by-beat assembly: use writing-beats. Not for paragraph-by-paragraph shaping: use writing-shape.

Use when a user asks to clean up, sync, dedupe, de-noise, or rewrite an iterated artifact. Not for a sediment rewrite of prose: use rewrite-clean-v0.

Use when asked to cold-read an artifact with fresh zero-context eyes and cut whatever a stranger cannot follow. Not for criteria-review of requirements docs: use doc-review.

Use when a user asks to write or announce a press release for any occasion or region, adapting to release type and media format. Read-only. All output in chat.

Use when a user gives a founder story or result and asks for a LinkedIn post with a chosen hook and humanized body. Not for tasks that require source or remote-system changes.

Use when asked to ghostwrite Substack newsletters and web posts from structured intake. Not for tasks that require source or remote-system changes.

Use when asked to rewrite technical English for STE compliance, stripping ambiguity and AI-style patterns. Not for French de-AI-ification: use humaniseur-fr.

Use when a user asks to tighten verbose-but-correct prose without a full rewrite. Cuts in place to load-bearing density, preserving every load-bearing claim.

Use when asked to write, review, or improve a technical article or engineering blog post. Blog mode covers deep dives, postmortems, and launches; article mode cites references. Not for code changes.

Use when a user asks to fix a listing whose order has gone arbitrary while preserving intentionally ranked items. Not for freely reorderable listings: use reorder.

Use when asked to create, refresh, or port a brand tone of voice guide (TONE.md) or measurable prose style rules (PROSE.md). Not for general copywriting: use copywriting.