By benjaroy
Manage a complete personal essay writing workflow with Claude: checkpoint versioned Markdown drafts, sort unstructured notes into key points, propose essay structures, compose in your calibrated voice, critique with ranked feedback, revise point-by-point preserving style, copyedit for polish, and generate targeted titles.
Save, view, and restore named draft versions of an essay as markdown files with incremental version numbering. Use this skill whenever someone wants to save their current draft, see revision history, go back to an earlier draft, or check what they've saved. Trigger when the user mentions checkpoints, saving a version, version history, previous drafts, going back, or undoing changes, or says "save this," "checkpoint this," "go back to the previous draft," or "can I go back."
Write a full personal essay from structured notes in the writer's own voice, calibrated to platform and word count. Build and save a persistent style profile from the writer's own writing samples. Use this skill whenever someone wants to draft or compose an essay, article, or piece of personal writing from notes, an outline, or structured points. Trigger when the user mentions composing, drafting, writing an essay, turning notes into prose, or asks to "write this up," "turn this into an essay," "draft this," or "write this for me." If a full draft already exists, use /revise instead.
Do a close read for grammar, punctuation, and sentence rhythm. Offer specific suggestions that quote the passage, state the change, and give one sentence of reasoning. Strictly surface-level mechanics. Use this skill whenever someone wants a final polish on their writing. Trigger when the user mentions copy editing, proofreading, grammar check, punctuation review, polishing prose, or final pass, or asks to "clean this up," "check this for errors," "do a final read," or "proofread this." If the writer needs substantive changes to argument, structure, or voice, use /revise instead.
Give a thorough, honest critique of a complete or near-complete essay, ranked by severity and capped at 600 words. Open with the single most important thing to fix. Close with simulated social media reactions. Use this skill whenever someone has a finished draft and wants a detailed assessment. Trigger when the user mentions critiquing, reviewing, getting feedback on, assessing, or evaluating writing. Also trigger for "how is this?", "what's working and what isn't?", "be honest with me about this," or "give me a thorough read." For quick check-ins or vague requests, use /riff instead.
Produce a stronger draft of an essay by implementing specific feedback point by point while preserving the writer's voice and strengths. Push back on feedback when warranted and explain why. Use this skill whenever someone wants to revise writing based on feedback, whether from /critique, a human editor, or their own notes. Trigger when the user mentions revising, implementing feedback, rewriting, making changes, or producing a new draft, or says "apply this feedback" or "fix this based on these notes." Distinct from /compose (which starts from notes, not a draft) and /copyedit (which handles mechanics only).
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
What is this?
This plugin is a toolkit to help people with the process of writing personal essays in collaboration with Claude. There are nine skills that are part of it: Sort, Sequence, Compose, Critique, Revise, Copyedit, Title, Checkpoint, and Riff. Together, they contain almost everything I know about writing in general, framed specifically to make writing with an LLM more intuitive through different steps. Each skill works on its own, though seven of the nine skills are also designed to work together in a chain.
Why does this exist?
The idea behind this plugin is pretty simple. I wanted to distill the ways I've found LLMs helpful in my personal writing process over the years, and formalize those flows into a set of skills for myself, then open-source everything for other people to use.
What do the skills actually do?
This skill takes unstructured material like notes, bullet points, or voice memos and distills that into a set of core points. It assesses the type of material you're working with, consolidates overlapping ideas, and preserves your original language and phrasing. Anything that doesn't fit the main structure goes into a "junk drawer" so nothing gets dropped. When it's done, it'll nudge you toward /sequence if you want to explore how those points might evolve into an essay, or it will suggest that you take time to think about the sorted points first if the underlying material is complex.
This skill takes sorted points and helps you figure out how to arrange them into an essay. Before proposing structures, it asks what's driving you to write i.e. whether you want to make an argument, tell a story, explain something, or aren't sure yet. If it detects a genuine structural tension in the material (like competing throughlines), it'll ask a follow up to help you choose a direction. Then it proposes one to three structural options depending on what the material supports. When you've picked a direction, it points you toward /compose.
This skill writes a full personal essay from structured notes in your voice. If you don't have a style profile yet, it asks for a writing sample to build one (you can upload a file, share a link, or paste text directly). This profile is saved and reused across sessions. It then asks where the essay will be published (Substack, LinkedIn, X, or other) and how long it should be, and those answers are used to calibrate tone and length. After the draft is written, this skill will either suggest sitting with what you've got or it will point you toward /critique for an honest assessment.
This skill gives an honest assessment of a piece of writing. It opens with the single most important thing to fix, then works through remaining observations in descending order of severity. It evaluates originality, argument quality, fluff, tone, and vulnerability to criticism, focusing on the dimensions that are most relevant to your specific piece. It closes with three simulated social media reactions to show how the piece might land in public. The skill will direct you from here as it sees fit, either to the next step with /revise, or if the feedback points to structural problems, it'll suggest going back to /sort or /sequence rather than pushing forward.
This skill takes a draft and a set of feedback and produces a stronger draft. It implements feedback point by point while preserving your voice and everything that was already working in the underlying essay. It can push back on feedback it disagrees with rather than blindly implementing everything, and it surfaces those disagreements clearly in its change summary. If the feedback reveals foundational problems, it'll say so and recommend stepping back in the pipeline rather than polishing something that isn't ready. When it's done, it asks whether you'd like to move to /copyedit for a final polish, take another revision pass, or step away and read the draft on your own first.
This skill does a close read of a piece of writing to check for grammar issues, punctuation errors, and sentence rhythm. If the piece still needs structural work, it'll say so and point you toward other skills for continued iteration. It verifies every issue before flagging it to avoid false positives, and if you have a style profile, it reads that first so it doesn't flatten your voice. Each suggestion cites an existing section, states its proposed change to that section, and gives one sentence of reasoning for that change. It can implement whatever changes you like, then when you're happy with the edits, it points you toward /title to help name the piece.
npx claudepluginhub benjaroy/riff --plugin riffWriting domain plugin - Content analysis, idea generation, and writing-kit workflow
10 specialized writing skills implementing Betty Flowers' Madman-Architect-Carpenter-Judge framework with Bryan Garner's whirlybird technique, SEO optimization, quality rubric, and multi-article content strategy. Features collaborative oscillation between AI and human across all phases.
Research-backed writing toolkit: anti-AI prose principles, iterative review loop, author voice capture, and voice-matched content generation
Audit & rewrite content to remove AI writing patterns ("AI-isms"). Supports detect-only and edit-in-place modes, voice profiles, and iterate-to-convergence.
AI-native publishing system — autonomous multi-phase workflows from ideation to EPUB/PDF. Fiction, non-fiction, technical, screenplay, poetry, game, academic, and custom genres.
Complete creative writing suite with 10 specialized agents covering the full writing process: research gathering, character development, story architecture, world-building, dialogue coaching, editing/review, outlining, content strategy, believability auditing, and prose style/voice analysis. Includes genre-specific guides, templates, and quality checklists.