Skip to content
Back to search

creative-writing-skills

MarketplaceWorks with Claude CodeGitHubIncomplete

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/haowjy-creative-writing-skills/badge.svg)](/listings/haowjy-creative-writing-skills)

HTML

<a href="/listings/haowjy-creative-writing-skills"><img src="/api/listings/haowjy-creative-writing-skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

1. **Install** for your environment: [meridian-cli](#mars-meridian-cli), [Claude Code / Cowork](#claude-code--cowork) or [Claude.ai](#claudeai). 2. **Run project setup once, before you write anything.** Agents never start it for you; you do:

| Environment | Run | |---|---| | meridian-cli | `meridian bootstrap` (runs automatically in your first session) | | Claude Code / Cowork | `/creative-writing-skills:project-setup` | | Claude.ai | turn on `project-setup` and ask it to set up your project |

It interviews you about the project, collects writing samples, and creates your `AGENTS.md`, the `kb/` layout and your first style files. Every agent reads those, so skipping this step makes them guess your conventions. 3. **Start writing with muse**, your creative partner: `meridian` (meridian-cli), `claude --agent creative-writing-skills:muse` (Claude Code), or turn on `creative-writing-muse` (Claude.ai).

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 19 lines
  1. {
  2. "name": "creative-writing-skills",
  3. "owner": {
  4. "name": "Jimmy Yao",
  5. "email": "dev.jimm.py@gmail.com"
  6. },
  7. "plugins": [
  8. {
  9. "name": "creative-writing-skills",
  10. "source": {
  11. "source": "git-subdir",
  12. "url": "haowjy/creative-writing-skills",
  13. "path": "plugin/claude",
  14. "ref": "v0.9.5"
  15. },
  16. "description": "Composable agents and skills for creative writing projects. The muse agent is the primary entry point: a creative partner that brainstorms, drafts, critiques, and maintains your story's knowledge base. On claude.ai (skills only), activate the creative-writing-muse skill instead."
  17. }
  18. ]
  19. }
README.md · README · 233 lines
  1. # Creative Writing Skills
  2. [![CI](https://github.com/haowjy/creative-writing-skills/actions/workflows/ci.yml/badge.svg)](https://github.com/haowjy/creative-writing-skills/actions/workflows/ci.yml)
  3. [![License](https://img.shields.io/badge/license-Apache%202.0-blue.svg)](LICENSE)
  4. Write novels, short stories, and serial fiction with AI that maintains your voice, tracks your continuity, and gets better the more you use it. From first brainstorm to polished draft: specialized agents handle each mode of work (writing, critiquing, revising, exploring) while shared skills carry the craft methodology.
  5. **What you get:**
  6. - **Brainstorm without committing**: explore plot options, character arcs, and world mechanics with multiple AI perspectives before deciding anything
  7. - **Write in your voice**: create style files from your existing prose, then draft new scenes that match
  8. - **Catch your own mistakes**: structured critique, continuity checks, and simulated reader reactions
  9. - **Keep everything in sync**: knowledge base updates as your story evolves
  10. ## Quick Start
  11. 1. **Install** for your environment: [meridian-cli](#mars-meridian-cli), [Claude Code / Cowork](#claude-code--cowork) or [Claude.ai](#claudeai).
  12. 2. **Run project setup once, before you write anything.** Agents never start it for you; you do:
  13. | Environment | Run |
  14. |---|---|
  15. | meridian-cli | `meridian bootstrap` (runs automatically in your first session) |
  16. | Claude Code / Cowork | `/creative-writing-skills:project-setup` |
  17. | Claude.ai | turn on `project-setup` and ask it to set up your project |
  18. It interviews you about the project, collects writing samples, and creates your `AGENTS.md`, the `kb/` layout and your first style files. Every agent reads those, so skipping this step makes them guess your conventions.
  19. 3. **Start writing with muse**, your creative partner: `meridian` (meridian-cli), `claude --agent creative-writing-skills:muse` (Claude Code), or turn on `creative-writing-muse` (Claude.ai).
  20. ## Installation
  21. ### Mars (meridian-cli)
  22. ```bash
  23. meridian mars add haowjy/creative-writing-skills
  24. meridian mars sync
  25. ```
  26. Then run `meridian` to start a meridian-cli session with muse. First-time setup runs the bootstrap automatically.
  27. `@writer` defaults to Opus at high effort.
  28. ### Claude Code / Cowork
  29. Claude Code and Cowork use the same plugin format.
  30. **Claude Code**: add the marketplace and install:
  31. ```bash
  32. /plugin marketplace add haowjy/creative-writing-skills
  33. /plugin install creative-writing-skills@creative-writing-skills
  34. ```
  35. **Cowork**: in the sidebar, **Customize** → **Personal plugins** → **+** → **Create plugins** → **Add marketplace** → **Add from repository**, enter `haowjy/creative-writing-skills`, then install the **creative-writing-skills** plugin. (Same flow as [Claude.ai](#claudeai) below — Cowork just also runs the agents.)
  36. Once installed, start a session with muse as your agent:
  37. ```bash
  38. claude --agent creative-writing-skills:muse
  39. ```
  40. Run the one-time project setup to create your `AGENTS.md`, `kb/` layout, and
  41. style files:
  42. ```
  43. /creative-writing-skills:project-setup
  44. ```
  45. ### Claude.ai
  46. `creative-writing-muse` turns an ordinary Claude chat into a creative writing partner. Turn it on, tell it what you're working on, and it carries you from first idea to finished draft in one conversation — talking through ideas, drafting scenes in your voice, flagging what isn't working, and revising. The other skills (prose, scene craft, critique, structure) are the craft knowledge it draws on; you don't have to invoke them yourself.
  47. First, add the skills to claude.ai — pick one:
  48. - **Add the marketplace (easiest):** in the sidebar, **Customize** → **Personal plugins** → **+** → **Create plugins** → **Add marketplace** → **Add from repository**, enter `haowjy/creative-writing-skills`, and click **Sync**.
  49. ![Adding the marketplace in claude.ai: the plugins icon and + in the sidebar, with the repository entered in the Add marketplace dialog](docs/images/claudeai-add-marketplace.png)
  50. - **Upload the files:** download the `.skill` files from the [latest release](https://github.com/haowjy/creative-writing-skills/releases/latest), then **Customize** → **Skills** → **"+"** → **Upload skill** for each one.
  51. Then start a chat, turn on **`creative-writing-muse`**, and describe what you want to write — it leads from there. Adding skills one at a time instead? Start with **creative-writing-muse**, **writing-principles**, **creative-writing-craft**, and **story-review**.
  52. Run **`project-setup`** once before writing. On claude.ai it returns the `AGENTS.md`, kb layout, and style guidance as text to save; in Cowork it writes them into the project.
  53. > Want to build the files yourself? Clone the repo and run `python scripts/create_skill_zips.py` to regenerate them in `zips/`. It packages the committed `plugin/claude/skills` tree; run `mars build plugin` first if you changed a source skill.
  54. ## How It Works
  55. ```mermaid
  56. flowchart TB
  57. You([You]) --> M[muse]
  58. M --> Explore
  59. M --> Draft
  60. M --> Maintain
  61. subgraph Explore ["Explore & Plan"]
  62. direction LR
  63. B[brainstormer] ~~~ CS[character-sim]
  64. CS ~~~ O[outliner]
  65. end
  66. subgraph Draft ["Draft & Revise"]
  67. direction TB
  68. W[muse drafts, or writer] --> CR[critic lanes]
  69. CR -->|findings or editorial memo| W
  70. M -.->|pivotal scenes| RS[reader-sim]
  71. M -.->|canon check| CC[continuity-checker]
  72. end
  73. subgraph Maintain ["Knowledge"]
  74. direction LR
  75. KL[kb-lead]
  76. SC[style-creator]
  77. end
  78. Explore -->|direction confirmed| Draft
  79. Draft -->|decisions & facts| Maintain
  80. Maintain -.->|context for next session| Explore
  81. ```
  82. **Explore:** Fan out brainstormers for creative variety. Spawn character-sims to discover voices. Use outliners to shape structure once a direction is chosen.
  83. **Draft & Revise:** Muse drafts and revises with you, and hands writer the passes that need their own session: alternate takes on another model, a polish pass you ask for, or a very long draft. Each critic runs one lane: a focus area for depth on one craft dimension, or an edit level, where editorial review gives the holistic book-editor priority across structure, voice, line quality, and surface. After two critique rounds on a chapter, muse brings you what is still open and you decide. Reader-sim gives experiential signal on pivotal scenes.
  84. **Knowledge:** When a chapter, revision, or ruling settles, muse dispatches `kb-lead` with the `story-memory` skill and what settled, so a trial stays a trial; in harnesses without agents it applies `story-memory` itself. Style-creator captures voice patterns from prose samples. The kb grows as the project evolves, giving every future agent accurate context.
  85. ## Agents
  86. | Agent | Role |
  87. |---|---|
  88. | **muse** | Author-facing creative partner for all story work: talks, plans, and drafts with you, spawns critique and knowledge capture, and owns the verdict |
  89. | **spark** | Budget creative partner with muse's posture: brainstorms, advises and drafts with you in one context, spawning no one |
  90. | **setup** | Runs `/project-setup` with you once, for runtimes where you start a chat with an agent instead of a skill |
  91. | **writer** | Production prose in one mode (draft, revision, bridge, alternate take, line polish) for passes that need their own session |
  92. | **critic** | Diagnoses a draft in one lane: a focus area (structure, character, voice, prose, continuity) or an edit level (editorial review, developmental, line edit, copyedit, proofreading) |
  93. | **reader-sim** | Experiential reader response to a draft, moment by moment |
  94. | **character-sim** | In-character conversation for voice discovery and relationship testing |
  95. | **continuity-checker** | Cross-references content against established canon for contradictions |
  96. | **brainstormer** | Creative option generation for a scoped question or angle |
  97. | **outliner** | Sequences confirmed direction into arc, chapter, and beat-level outlines |
  98. | **style-creator** | Analyzes prose samples to produce style reference files for the project's voice |
  99. | **kb-lead** | Mines settled work for decisions and updates the kb; the agent behind `story-memory` |
  100. Four `meridian-base` delegates ship with the plugin for the work around the writing: `explorer` finds what's related in a large area, `session-miner` finds where past sessions discuss a question, `cold-reader` checks text that outlives the work for what a reader who wasn't there cannot follow, and `kb-maintainer` refactors a documentation tree's structure.
  101. Knowledge extraction into the kb in harnesses without agents stays in the conversation: the muse applies the `story-memory` skill directly.
  102. ## Skills
  103. | Skill | Purpose |
  104. |---|---|
  105. | **creative-writing-craft** | Prose modes (fresh draft, revision, bridge, alternate take, line polish) and craft references for prose, scenes, style, voice, and genre |
  106. | **writing-principles** | Reader reward channels, AI failure modes, and fiction-specific taste discipline |
  107. | **story-planning** | Direction, brainstorming, outlining, and story architecture before pages exist |
  108. | **story-review** | Editorial review, developmental edit, line edit, copyedit, proofreading, craft critique, and reader-signal synthesis |
  109. | **story-memory** | Context, fact extraction, reference writing, artifact layout, and persistent issue tracking |
  110. | **reader-sim** | First-time reader simulation from a specified persona |
  111. | **character-sim** | In-character conversation for voice and relationship testing |
  112. | **creative-writing-muse** | The muse's method: one creative partner that talks, plans, drafts, and spawns what needs its own context; turn it on in any environment |
  113. | **story-workflow** | The lifecycle of story work: intent, direction, drafting, critique rounds, the verdict, knowledge capture, presenting |
  114. | **writing-staffing** | Which agent takes which lane, what to tell it to load, and how many lanes to run |
  115. | **llm-writing** | Intentional language discipline: catches unchosen LLM defaults while preserving deliberate ambiguity, omission, repetition, and rhythm |
  116. | **shared-dao** | Shared vocabulary: canonical story terms, aliases, and ambiguity resolution |
  117. | **project-setup** | One-time guided setup (user-invoked): interviews you, then creates AGENTS.md, the kb layout, and style files |
  118. The plugin also ships the shared `meridian-base` skills its agents load: `knowledge-layers`, `qi-layer`, `qi-maintenance`, `information-hierarchy`, `intent-modeling`, `md-validation`, `shared-workspace`, and `structured-artifact`, for the knowledge and craft discipline.
  119. ## Project Layout
  120. ```text
  121. my-story/
  122. ├── AGENTS.md # Project conventions (created by project-setup)
  123. ├── story/ # Chapters and manuscript
  124. ├── work/ # Current drafting effort
  125. │ ├── outline/
  126. │ ├── drafts/
  127. │ ├── critique-reports/
  128. │ └── brainstorm/
  129. └── kb/ # Durable knowledge base
  130. ├── styles/ # Voice reference files
  131. ├── characters/ # Character state and profiles
  132. ├── world/ # Locations, lore, systems
  133. ├── timeline/ # Chronology
  134. ├── canon/ # Established facts
  135. └── issues/ # Tracked writing problems
  136. ```
  137. ## Compatibility
  138. | Feature | Claude Code | Cowork | Mars (meridian-cli) | Claude.ai |
  139. |---|:---:|:---:|:---:|:---:|
  140. | All agents | Yes (flat) | Yes (flat) | Yes | No (grayed out in chat) |
  141. | All skills | Yes | Yes | Yes | Marketplace add or zip |
  142. | Multi-agent orchestration | Via muse | Via muse | Via muse → focused workers | No |
  143. | Project setup | Yes | Yes | Yes | Text to save |
  144. Claude Code, Cowork, and meridian-cli run muse as the session agent: it drafts with you and spawns a compact worker set: writer, critic, reader-sim, brainstormer, outliner, character-sim, style-creator, continuity-checker, and creative-researcher. You can add the marketplace to the Claude desktop app from GitHub; Cowork runs the agents, but plain claude.ai chat runs skills only (agents grayed out) — there the `creative-writing-muse` skill runs the same muse method in one conversation, backed by the craft skills.
  145. ## Current Experiments
  146. **Rhetorical questions in skill prompts.** The economy section in `writing-principles` uses rhetorical questions ("what can you leave out and still have the scene work?") rather than declarative statements. LLMs can distinguish rhetorical from information-seeking questions internally ([arxiv 2604.14128](https://arxiv.org/abs/2604.14128)), and Self-Ask prompting shows questions improve reasoning, but no research directly tests whether rhetorical questions in system prompts improve task performance vs. equivalent declaratives. Keeping the rhetorical form to see if it activates a self-check loop that declaratives don't.
  147. ## Development
  148. ### Rebuild the plugin
  149. The Claude/Codex/Grok, Cursor, and OpenCode plugin trees in `plugin/`, the
  150. marketplace catalogs, and the root OpenCode `package.json` are generated by
  151. Mars from `agents/`, `skills/`, `bootstrap/`, and the dependency graph. After
  152. editing a source agent or skill, rebuild and commit the output:
  153. ```bash
  154. mars build plugin # Mars 0.26.2 or newer
  155. ```
  156. `mars check` reports stale or hand-edited output; CI rebuilds and fails on any difference.
  157. ### Validate package
  158. ```bash
  159. mars check
  160. ```
  161. ### Release
  162. ```bash
  163. mars version patch # bump, rebuild plugin/, promote CHANGELOG, commit, tag
  164. mars version patch --push # also push the commit and tag
  165. ```
  166. Use `mars` 0.26.2 or newer (`uv tool install mars-agents`); it rebuilds `plugin/` and the marketplace catalogs at the new version inside the release commit.
  167. ## License
  168. Apache License 2.0. See [LICENSE](LICENSE).

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.