claude

novelist

小说写作全流程插件 - 从爆款题材选择/世界观/人物/大纲设定到章节编写、一致性检查、校对、去AIGC、重写、批量流水线的端到端工作流。13 skill: init/design/character/worldview/outline/write/check/rewrite/proofread/craft/trending/humanize/pipeline; 7 agent: continuity-auditor(双模式)/chapter-writer/proofreader/humanizer/outliner/worldbuilder/indexer。编写前四要素确认, 生成后收尾链(检查→校对→去AIGC→定稿)+评分门控; pipeline 用 Workflow 逐章串行批量写多章(各阶段派专用 agent); craft 三流派叙事镜片, trending 选题方法论, humanize 去AI味

5 on GitHubby lazygophersAGPL-3.0-or-laterupdated 3 days ago
/

Plugin novelist from marketplace ccplugin-market in lazygophers/ccplugin. Commands read from the repo's manifest — 2 days ago.

What it installs

Skills (14)

  • novelist-character

    创建或更新单个小说人物的详细设定。当用户说"加个人物/写某角色设定/完善某人物/这个角色的经历关系"时调用。每个人物在 人物/<人物名>/ 下独立文件夹, 拆为三文件: 简介.md(简介+性格)、经历.md(各时间段经历/事件)、关系.md(与他人交集/是否见过)。本 skill 独占 人物/ 目录。

  • novelist-check

    检查小说整条故事线的一致性(18 子项)。当用户说"检查一致性/查冲突/审故事线/有没有矛盾/设定对不对"时调用; 也被 novelist-write 在编写前引用取事实快照。mode=detect(默认只读审查产冲突清单) / mode=fix(单点冲突直接 Edit 修正, 段落级派 chapter-writer)。派 continuity-auditor agent 审查, 报告落 元数据/检查报告/。

  • novelist-craft

    小说创作大师视角——叙事心智模型库与镜片选择引擎。蒸馏三大叙事传统(中文类型/网文、世界结构、纯文学)的可操作创作启发式, 按用户自己小说的题材选用对应"叙事镜片"。当用户问"这章人物怎么写更立得住/情节怎么推/力量体系怎么设/节奏怎么调/用什么写法"时调用; 也被 novelist-design 与 novelist-write 引用作为创作引擎。这是知识框架 skill, 不替代用户决策——它给镜片, 用户选镜片。

  • novelist-design

    小说的整体核心设计——主要剧情与主要人物一起定(二者必须同处本 skill, 因为主角与主线相互定义)。当用户说"设计这本书的核心/定主线和主角/搭整体框架/核心设定"时调用。本 skill 是统筹者: 自己产出核心骨架, 并引用 novelist-outline / novelist-character / novelist-worldview 落地各部分细节, 不与它们抢目录。

  • novelist-humanize

    去 AIGC / 去 AI 味——把读起来"像 AI 生成"的小说正文改回"像人写的", 降低 AIGC 检测特征同时提升可读性。当用户说"去AI味/降AIGC/这段太AI了/去机器味/人味润色/过检测"时调用; 也被 novelist-write / novelist-rewrite 的生成后收尾链引用。去味时**先运行时发现并依次执行用户已安装的所有 humanizer 类 skill(动态发现, 不硬编码路径)**, 再用本插件 humanizer agent 补小说特化去味(对话破坏/情感呈现)。诊断三层 AI 味(统计指纹/通用tells/小说特有)。只改文字纹理, 不动剧情设定。

  • novelist-init

    初始化一部新小说的目录环境(scale)。当用户说"新建小说/开一本书/初始化小说项目/搭小说目录"时调用, 创建中文目录骨架(人物/世界观/设定/大纲/情节/章节/元数据)+模板文件。这是 novelist 插件所有其他 skill 的前置——没有目录环境, design/write/check 无处落盘。

  • novelist-lint

    校验 novelist 插件目录结构/文件名/文件夹组织是否符合插件规范 (plugin.json 合法 / commands-agents-skills 在根 / SKILL.md 大写 / kebab 命名 / 无产物泄漏 / plugin.json 路径真实 / frontmatter name 对齐)。novelist 插件改动后、加新 skill/agent 后、发布前、或怀疑结构不合规时调用。调用即跑 scripts/lint.py 扫 7 规则违规, 按 fix-hints 移/改/删到合规

  • novelist-outline

    创建或更新小说的大纲与情节线。当用户说"写大纲/规划剧情/排章节/理主线支线/管伏笔"时调用。管理 大纲/(总纲+分卷) 与 情节/(主线+支线+伏笔追踪) 目录, 把宏观结构与逐章节奏落到可核对的表格。本 skill 独占 大纲/ 与 情节/ 目录。

  • novelist-pipeline

    批量写小说的流水线编排: 默认 mode=write 两流水线并行批量写(Writer 流 ‖ 检测定稿流, 每章 write→三环并行检测(check/humanize/proofread)→fix串行改→定稿)。加 mode 入参支持 review(评审)/humanize(去AI味)/proofread(校对)/polish(润色=三环收尾)/rewrite(重写)/outline(大纲)。载体: write 默认 Workflow(workflow.js), 其余默认 trellisx subagent 编排(main DAG 调度派 novelist 系列 skill, 并发上限2)

  • novelist-proofread

    校对小说正文的文字层问题(12 子项)——错别字(形近/音近/多漏字)、语法(残缺/搭配/语序)、标点、用词、用字统一。支持 mode=detect(只读扫描产报告, 不改正文) 与 mode=fix(默认, 报告+就地 Edit 修正)。当用户说"校对/查错别字/润色文字/检查语法/通顺一下"时调用。区别于 novelist-check(查剧情设定一致性), 本 skill 只管文字表达正确通顺, 不动剧情。派 proofreader agent 校对, 报告落 元数据/校对报告/。

  • novelist-rewrite

    重写小说章节。入参 mode=detect|fix(默认 fix): detect=只读诊断, 扫描全书产「建议重写章清单+理由+建议模式A/B/C」, 不改文件; fix=执行三模式——(A) 按 novelist-check 报告定点修复冲突; (B) 从指定章起清空该章及之后重写; (C) 连续重写多个指定章节。当用户说"重写第N章/从第N章开始重写/这几章重写/按检查报告改/第N章之后全删重写/诊断哪些章需重写"时调用。清空是破坏性操作, 执行前必须显式确认。本 skill 独占 章节/ 目录的覆盖写入。

  • novelist-trending

    爆款题材选择——网文/小说选题方法论框架 + 实时数据协议。当用户问"写什么题材火/选个爆款题材/这个题材现在能写吗/选题/题材有没有市场/红海还是蓝海"时调用; 也被 novelist-design 在立项前引用以定题材方向。给的是稳定的选题判断镜片(榜单逻辑/生命周期/红蓝海/作者优势匹配/避坑), 当下"什么火"靠 agentic protocol 联网取——不钉死会过时的热点。这是决策辅助框架, 不替用户拍板, 给镜片+利弊由用户选。

  • novelist-worldview

    创建或更新小说的世界观与各类设定。当用户说"设定世界观/写背景设定/定魔法体系/加个组织或物品设定/历史背景"时调用。世界观拆为 地理/势力/规则(力量体系)/历史 四文件; 其他设定(物品/组织/术语)按一设定一文件存于 设定/ 目录。本 skill 独占 世界观/ 与 设定/ 目录。

  • novelist-write

    小说写作一站式入口——用户只需调用 write, 即跑完整链: 前置就绪(目录环境+核心设计)→ 四要素确认 → 生成正文 → 收尾链(一致性检查→校对→去AIGC→定稿)。当用户说"写第N章/继续写/接着写剧情/写下一章/写这本书"时调用。设计未就绪会自动引导补(design), 全流程由本 skill 编排, 用户不必分别调各 skill。本 skill 独占 章节/ 目录的新增写入。

Subagents (7)

  • chapter-writer

    Use this agent to write or rewrite a novel chapter's prose. Dispatched by novelist-write (new chapters, after four-element confirmation) and novelist-rewrite (fixes / re-drafts, modes A/B/C). Also shared by novelist-check (mode=fix paragraph-level fixes — rewrites only the conflict paragraph, never

  • continuity-auditor

    Use this agent for novel consistency checking across six dimensions / eighteen sub-items (setting conflicts, character contradictions, world-rule violations, timeline errors, unresolved foreshadowing, logic gaps). Three modes — 审查-detect (read-only report), 审查-fix (single-point Edit on prose; paragr

  • humanizer

    Use this agent to remove the "AI taste" (AIGC fingerprint) from novel prose — even out the over-uniform rhythm, cut cliché transition words, replace abstraction with concrete detail, restore human burstiness. Dispatched by the novelist-humanize skill. It rewrites at the prose-texture level only; nev

  • indexer

    Use this agent to finalize a chapter — update the chapter index (章节/_索引.md) and progress (元数据/进度.md) after a chapter passes the closing chain. Dispatched by the novelist-pipeline workflow's 定稿 stage. Mechanical bookkeeping only; does not write or fix prose.

  • outliner

    Use this agent to generate a per-batch chapter route map (路线图) for a novel — reads 大纲/总纲+分卷, 情节/主线+伏笔, 元数据/进度, and outputs a structured route map (每章: 核心事件/人物变化/伏笔推进/收尾钩子/字数目标). Dispatched by the novelist-pipeline workflow''s 路线图 stage. Plans only; does not write chapter prose.

  • proofreader

    Use this agent to proofread novel prose for text-level issues only — typos, grammar, punctuation, awkward sentences, wording, and consistent use of names/terms (12 sub-items). Dispatched by the novelist-proofread skill. Supports mode=detect (read-only report) and mode=fix (default; fix text in place

  • worldbuilder

    Use this agent to update a novel's worldview/setting/character files when a batch's route map introduces new elements (new rules, factions, geography, items, characters). Dispatched by the novelist-pipeline workflow's 世界观 stage. Updates settings only; does not write chapter prose or alter the route

View source on GitHub

More from lazygophers/ccplugin

All 9 plugins →

skein

claude

SKEIN — 独立任务管理插件 (零 trellis 依赖, 自带 .skein/ 工作区): 强制 task 闭环 (plan→exec→check→finish) + 动态 DAG 编排调度 (冲突自算边 + depends_on, 并发上限 2, task 级 + subtask 级双层) + worktree 隔离 + task.md 看板; 差异化核心是两层×类目规则记忆 (基于 .claude/rules): core 常驻注入硬规 + recall 按需语义召回, 层内按类目 (git/test/arch/build...) 分子目录, task finish 后经 sediment 判定门沉淀。9 skill: flow/orchestrate/workspace/planning/memory/grill/spec/check/clean + 3 agent: implementer/checker/researcher (无 Agent/Task 递归护栏)

8 skills · 3 commands · 11 agents · hooks

5

cortex

claude

知识库 + 记忆管理插件 - 双层 vault, 5 级记忆 (遗忘曲线), 8 skill: schema/ingest/lint/extract/history-digest/context-digest/evolve/recall

8 skills · 6 agents · hooks

5

深度研究插件 - 基于图思维框架的多智能体深度研究系统。支持多领域深度调查、引用验证、知识合成和专业报告生成

6 skills · 4 agents

5

trellisx

claude

Trellisx — Trellis 增强改造工具: 跑一次 trellisx-apply 把 强推task / subtask拆分 / worktree隔离 / plan→exec→check→finish闭环 / task.md看板 五维度注入 .trellis (workflow + config.yaml 生命周期 hook + scripts); 含 trellisx-flow 强制task闭环 / orchestrate 编排 / workspace 看板 / spec 破坏式重构

8 skills · 1 command · hooks

5

Claude Code skill 集合 plugin——Git、项目工程、交互工具,7 个 skill 一键装齐。

8 skills

5

notify

claude

系统通知插件 - 通过系统通知向用户实时提示会话状态变更、权限请求等重要事件。跨平台支持 macOS、Linux、Windows

1 command · hooks

5

Similar plugins

agentdb

claude

AgentDB (external: maintained in ruvnet/agentdb, pinned to commit 27678c3, plugin 3.6.0): agentdb's own standalone MCP server pinned to [email protected] (vector memory with get/list/update/delete for records, patterns and skills, hybrid keyword + vector episode search and an opt-in local cross-encoder, an opt-in aidefence injection guard on memory writes, reflexion, skills, outcome feedback, capabilities report), a memory-specialist agent, 3 skills, /agentdb-status, and the agentdb mod guard that keeps secrets out of memory and refuses delete/reset unless allowed. Complements ruflo-agentdb, which reaches AgentDB through ruflo's server

3 skills · 1 command · 1 agent · 1 MCP server

ruvnet/ruflo 81,589

humanizer

claude

Rewrite AI-sounding text so it reads like the writer without changing what it says.

1 skill

blader/humanizer 68,461

5 skills + 35 modes + Material Passport pipeline. Includes v3.6.7 cross-model audit gate and v3.6.8 generator-evaluator contract.

5 skills · 16 commands · 4 agents · hooks

Imbad0202/academic-research-skills 56,698

Connect Claude Code directly to your CockroachDB clusters for hands-on database work — explore schemas, write optimized SQL, debug queries, and manage distributed database clusters. This plugin provides 14 tools across two active MCP backends (self-hosted MCP Toolbox and managed CockroachDB Cloud MCP Server), three specialized agents (DBA, Developer, Operator), 32 skills across 6 operational domains, and built-in safety hooks.

34 skills · 3 agents · 3 MCP servers · hooks

anthropics/claude-plugins-official 47,076

OpenAPI specification generation, Mermaid diagram creation, tutorial writing, API reference documentation

3 skills · 1 command · 5 agents

wshobson/agents 42,461

Provision direct cloud access (Vertex AI, Bedrock, or LLM gateway) for the Claude Microsoft 365 add-in. Generates the customized manifest, walks through Azure admin consent, and writes per-user config via Graph extension attributes.

1 skill · 9 commands

anthropics/financial-services 41,259

Maintain novelist?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins9 plugins