#946 by GitHub stars
LiPu-jpg/Openwrite
dsh-Openwrite:OpenWrite 的 DeepSeek Harness 小说创作插件,含统一创作 Agent、90 个小说工具、原生工作台与标准审稿 DAG
History
Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.
View weekly changes
presets/dante/skills/dialoguequality/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/dog-delivery-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/dog-import-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/dog-review-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/novel-reviewer/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/oh-story-deslop/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/oh-story-long-write/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/oh-story-review/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/truth-validation/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/dante/skills/workflow-manager/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/foreshadowing-system/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/novel-creator/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/oh-story-long-analyze/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/oh-story-long-scan/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/style-system/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/goethe/skills/world-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/canon/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/dialoguequality/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/dog-delivery-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/dog-import-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/dog-review-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/export-book/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/foreshadow/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/foreshadowing-system/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/learn/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/novel-creator/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/novel-reviewer/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/oh-story-deslop/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/oh-story-long-analyze/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/oh-story-long-scan/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/oh-story-long-write/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/oh-story-review/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/progress/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/review-chapter/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/revise-span/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/style-system/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/truth-validation/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/workflow-manager/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/world-query/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
presets/openwrite/skills/write-next/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
vendor/dsh-dog/docs/skills/dog-v02-agentic-ci/SKILL.md
Week of Oct 26, 2026 · +0 / −0 lines
Before Not checked yet → After Instructions only
edf0e55f4e5a → a8b9e782a9d2
Diff text is unavailable under the recorded licenses. Line counts and the GitHub link remain available.
Heat
7.8- Star growth, last 7 daysAhead of 92% of tracked repositories
- Star growth, last 24 hoursAhead of 96% of tracked repositories
- Rank movement, last 7 daysAhead of 72% of tracked repositories
- Commits, last 30 daysAhead of 94% of tracked repositories
- Accounts asking about it through MCP, last 7 daysNot available
- Hot votes this weekAhead of 0% of tracked repositories
Heat never changes the order by stars, and sponsorship never touches it. How heat works
Stars over time
Individual skills in this repository
- dialoguequality
presets/dante/skills/dialoguequality - dog-delivery-query
presets/dante/skills/dog-delivery-query - dog-import-query
presets/dante/skills/dog-import-query - dog-review-query
presets/dante/skills/dog-review-query - novel-reviewer
presets/dante/skills/novel-reviewer - oh-story-deslop
presets/dante/skills/oh-story-deslop - oh-story-long-write
presets/dante/skills/oh-story-long-write - oh-story-review
presets/dante/skills/oh-story-review - truth-validation
presets/dante/skills/truth-validation - workflow-manager
presets/dante/skills/workflow-manager - foreshadowing-system
presets/goethe/skills/foreshadowing-system - novel-creator
presets/goethe/skills/novel-creator - oh-story-long-analyze
presets/goethe/skills/oh-story-long-analyze - oh-story-long-scan
presets/goethe/skills/oh-story-long-scan - style-system
presets/goethe/skills/style-system - world-query
presets/goethe/skills/world-query - canon
presets/openwrite/skills/canon - dialoguequality
presets/openwrite/skills/dialoguequality - dog-delivery-query
presets/openwrite/skills/dog-delivery-query - dog-import-query
presets/openwrite/skills/dog-import-query - dog-review-query
presets/openwrite/skills/dog-review-query - export-book
presets/openwrite/skills/export-book - foreshadow
presets/openwrite/skills/foreshadow - foreshadowing-system
presets/openwrite/skills/foreshadowing-system - learn
presets/openwrite/skills/learn - novel-creator
presets/openwrite/skills/novel-creator - novel-reviewer
presets/openwrite/skills/novel-reviewer - oh-story-deslop
presets/openwrite/skills/oh-story-deslop - oh-story-long-analyze
presets/openwrite/skills/oh-story-long-analyze - oh-story-long-scan
presets/openwrite/skills/oh-story-long-scan - oh-story-long-write
presets/openwrite/skills/oh-story-long-write - oh-story-review
presets/openwrite/skills/oh-story-review - progress
presets/openwrite/skills/progress - review-chapter
presets/openwrite/skills/review-chapter - revise-span
presets/openwrite/skills/revise-span - style-system
presets/openwrite/skills/style-system - truth-validation
presets/openwrite/skills/truth-validation - workflow-manager
presets/openwrite/skills/workflow-manager - world-query
presets/openwrite/skills/world-query - write-next
presets/openwrite/skills/write-next - dog-v02-agentic-ci
vendor/dsh-dog/docs/skills/dog-v02-agentic-ci
How it works
dialoguequality
When your agent uses it
Use when user wants to analyze dialogue style, extract speech patterns, compare character voices, or detect AI-sounding dialogue. Triggers include "对话", "口头禅", "角色声音", "对白", "AI味对话".
What it loads into context
Frontmatter
57 tokensAlways available to the agent
SKILL.md
1,331 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 当前数据来源
- 可用工具
- 什么时候用
- 工作方式
- 检查清单
- 输出建议格式
- 示例
- 注意
Files
SKILL.mdcore4,555 bytes
Read SKILL.md
# 对话质量系统
用于分析角色对白、提取说话习惯、检查角色声音是否区分明显,以及识别 AI 味对话问题。
## 当前数据来源
对话分析不要只盯着对白片段本身。当前推荐读取顺序是:
1. `data/manuscript/arc_*/ch_*.md` — 实际章节正文
2. `src/characters/*.md` — 角色单源文档
3. `data/style/composed.md` — 当前作品风格约束
4. `src/outline.md` / canonical packet — 本章戏剧位置与章节目标
如果需要完整判断某章对白是否失真,优先结合 `review_chapter` 或章节 packet,而不是只看几行台词。View the complete source on GitHubdog-delivery-query
When your agent uses it
Query a chapter delivery DoG graph across manuscript, review, revision application, and re-review closure. Use after writing or reviewing a chapter, after revision actions, or when the user asks whether a chapter is truly ready to deliver.
What it loads into context
Frontmatter
55 tokensAlways available to the agent
SKILL.md
533 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 查询流程
- 状态解释
- 约束
Files
SKILL.mdcore2,036 bytes
Read SKILL.md
# 章节交付 DoG 查询
章节写作、评审和修订工具会生成:
`data/novels/{id}/data/dog/deliveries/{chapter}/dog-graph.json`
这张图查询六个阶段:`writing → review → revision → application → rereview → closure`。
它不替代 OpenWrite 的写章、六域累加评审或修订服务。
## 查询流程
1. 从最近一次 `novel_write_chapter`、`novel_review_chapter`、`novel_revision_*`
或 `novel_task_get` 结果的 `dog_delivery.graphPath` 找到图;conductor 输出使用
`dog_delivery_graph_path`。View the complete source on GitHubdog-import-query
When your agent uses it
Query a smart-import DoG graph after a novel has been split into chapters. Use to inspect import completeness, AI precheck errors, and pending outline/asset/canonical setup.
What it loads into context
Frontmatter
44 tokensAlways available to the agent
SKILL.md
404 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 工作流
- 约束
Files
SKILL.mdcore1,490 bytes
Read SKILL.md
# 拆书导入 DoG 验收
`conductor/smart_import.py` 完成导入后会生成 `dog-graph.json`。这张图只验收导入
结果,不负责写入章节、资产或大纲。
## 工作流
1. 从导入输出中找到 `data/novels/{id}/data/dog/imports/{IMPORT_ID}/dog-graph.json`。
2. 依次调用 `dog_validate`、`dog_create`、`dog_run`、`dog_wait`,直到终态。
3. 用 `dog_status` 查看 `manifest` 和每个 `chapter-ch_*` 节点的 evidence。
4. 根节点通过后,读取 manifest 的 `construction.nextActions`,再调用 Goethe 的规划和资产工具建立大纲、角色、世界观、进度和正典事件。
5. 资产/大纲建立完成后重新运行 `novel_review_chapter`,再用 `dog-review-query`
查询六域评审和 37 个 legacy leaf。
View the complete source on GitHubdog-review-query
When your agent uses it
Query an existing hierarchical OpenWrite review-v2 artifact through dsh-dog. Use after novel_review_chapter or when inspecting six quality domains, evidence, coverage, blockers, delivery status, or the 37 legacy checks. Never rerun model review from DoG.
What it loads into context
Frontmatter
62 tokensAlways available to the agent
SKILL.md
650 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 工作流
- 状态约定
- 边界
Files
SKILL.mdcore2,400 bytes
Read SKILL.md
# 分层评审 DoG 查询
这个 Skill 只查询 OpenWrite 已物化的 review-v2 artifact。OpenWrite 的
`novel_review_chapter` / conductor 是评审真源;DoG 读取 47 节点分层图:上下文、
六个质量域、硬门禁、确定性聚合和 37 个兼容查询 leaf。DoG 不调用评审模型。
## 工作流
1. 如果本轮还没有评审结果,先按 Dante 正常流程调用 `novel_review_chapter`。
2. 找到评审输出目录中的 `dog-graph.json`。graph 与 `review.json`、`context.json`、
`domain-*.json`、`gate.json`、`aggregate.json` 和 `dim_01.json` …
`dim_37.json` 在同一目录。
3. 读取 graph 后依次调用 `dog_validate`、`dog_create`、`dog_run`、`dog_wait`。
所有节点均使用 programmatic `review-record` verifier;`dog_run` 后必须等到终态。View the complete source on GitHubnovel-reviewer
When your agent uses it
Use when user wants to check logic consistency, polish style, find plot holes, or review existing chapters. Triggers include "检查", "润色", "逻辑", "伏笔漏洞", "审查".
What it loads into context
Frontmatter
52 tokensAlways available to the agent
SKILL.md
1,120 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 触发条件
- 工作流程
- 工具
- 六域(计分)与硬门禁
- 报告给作者看的结构
- 错误处理
- 与其他技能
Files
SKILL.mdcore3,713 bytes
Read SKILL.md
# 小说审查润色
通过 openwrite-bridge 的 `novel_*` 工具评审与修订。不要调用 Python 审查类或本地 33 维脚本。
正式评审是六域累加:质量分、覆盖率、硬门禁、交付状态分开判断;37 项只作展开查询。
## 触发条件
- "检查逻辑" / "逻辑检查"
- "润色" / "润色风格"
- "找伏笔漏洞" / "伏笔检查"
- "审查章节" / "审查"
- "风格问题"
## 工作流程View the complete source on GitHuboh-story-deslop
When your agent uses it
诊断并精修网络小说中的机械 AI 读感,包括重复解释、句式过度齐整、对话同声、抽象判断、万能身体反应、空泛氛围和无功能金句。用于去 AI 味、自然化润色或在保留作者声音的前提下压缩文本。
What it loads into context
Frontmatter
160 tokensAlways available to the agent
SKILL.md
554 tokensWhen the skill is activated
Reference documents
365 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore1,942 bytesagents/openai.yamlasset259 bytesreferences/method.mddoc1,313 bytes
Read SKILL.md
# Oh Story 去 AI 味
把 AI 味视为读感和叙事功能问题,不把词表命中当作错误。先确认用户要扫描、局部示范还是全文精修;默认先报告问题和样例,不擅自覆盖正文。
## 执行流程
1. 读取作者意图、人物声音、题材气质和当前段落功能,建立必须保留的内容。
2. 使用 `references/method.md` 扫描重复解释、句式齐整、对话同声、抽象判断和装饰性表达。
3. 按对读感的影响分级,优先修改高频且无功能的模式,不追求把所有句子变得不同。
4. 用最小改动恢复动作、选择、潜台词和节奏;结构或事实问题转交 `@oh-story-review`,不在润色中暗改。
5. 输出修改统计和少量前后对照;用户明确要求全文精修时才给完整候选稿。
## OpenWrite 边界
View the complete source on GitHuboh-story-long-write
When your agent uses it
规划和创作长篇网络小说,覆盖选题定位、读者契约、人物与世界设定、卷纲细纲、单章写作、日更续写和结构修订。用于开书、补纲、写指定章节、续写或回炉长篇正文。
What it loads into context
Frontmatter
152 tokensAlways available to the agent
SKILL.md
628 tokensWhen the skill is activated
Reference documents
622 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,204 bytesagents/openai.yamlasset279 bytesreferences/method.mddoc2,261 bytes
Read SKILL.md
# Oh Story 长篇写作
先读取当前作品状态和用户本轮目标,再选择开书、补纲、写指定章、续写或修订。裸调用时只诊断所处阶段并给出下一步,不自动生成正文。
## 执行流程
1. 明确本轮交付范围和停靠点。用户没有指定正文数量时最多处理一章;只要求规划时停在候选或大纲交付。
2. 让 Goethe 处理读者契约、人物、世界、关系、全书阶段、卷纲和细纲;让 Dante 根据已确认资产写正文或做局部修订。
3. 使用 `references/method.md` 检查情绪承诺、主角代理权、冲突升级、剧情单元和章末钩子。
4. 写章前只召回不知道就会写错的资料:当前细纲、最近正文、出场人物状态、相关规则、伏笔和当前关注点。
5. 写完检查事实连续性、角色选择、信息变化、情绪结算和章节状态;结构问题优先于句子润色。
## OpenWrite 边界
View the complete source on GitHuboh-story-review
When your agent uses it
对网络小说大纲、章节或短篇成稿做证据化商业审稿,按事实连续性、因果、人物选择、读者契约、节奏、信息、对话和表达分级问题。用于章节验收、整稿诊断、投稿前审查或生成优先修订清单。
What it loads into context
Frontmatter
158 tokensAlways available to the agent
SKILL.md
539 tokensWhen the skill is activated
Reference documents
468 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore1,900 bytesagents/openai.yamlasset264 bytesreferences/method.mddoc1,665 bytes
Read SKILL.md
# Oh Story 商业审稿
先确认审查对象、范围、目标读者和用户要“只诊断”还是“诊断后给修改稿”。默认只输出审查报告,不直接改动正文。
## 执行流程
1. 读取当前项目已确认资产、相关正文和审稿范围,先做缺失资料与冲突预检。
2. 使用 `references/method.md` 按阻断、严重、一般、润色四级检查,不让文风偏好掩盖事实或结构问题。
3. 每条发现给出位置或引文、问题、读者影响、证据和最小修复方向;证据不足时标记“需补充”。
4. 汇总最高优先级问题、可保留优点、建议修订顺序和仍需作者裁定的取舍。
5. 只有用户明确要求修改时才生成候选稿;涉及正式正文提交时继续走 OpenWrite workflow。
## OpenWrite 边界
View the complete source on GitHubtruth-validation
When your agent uses it
Use when user wants to validate story consistency, check for plot holes, or verify state changes. Triggers include "验证", "一致性", "检查逻辑", "状态冲突".
What it loads into context
Frontmatter
46 tokensAlways available to the agent
SKILL.md
877 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心概念
- 验证问题类型
- 可用工具
- 使用示例
- 常见问题及修复
- 验证时机
- 与伏笔系统的关系
Files
SKILL.mdcore2,869 bytes
Read SKILL.md
# 真相文件验证系统
验证真相文件与章节内容的一致性,确保故事逻辑连贯。
## 核心概念
真相文件是记录世界当前状态的 Markdown 文件,包括:
- `current_state.md` - 世界当前状态
- `ledger.md` - 资源账本
- `relationships.md` - 角色关系矩阵
相关运行时文件(用于一致性校验):
- `data/foreshadowing/dag.yaml` - 伏笔状态
- `data/hierarchy.yaml` 与 `data/compressed/` - 章节摘要与压缩信息View the complete source on GitHubworkflow-manager
When your agent uses it
Use when user wants to check writing progress, manage workflow stages, or resume interrupted tasks. Triggers include "工作流", "进度", "阶段", "写作流程".
What it loads into context
Frontmatter
44 tokensAlways available to the agent
SKILL.md
940 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 作者可见阶段
- 工具
- 用法
- 状态怎么说
- 与其他技能
Files
SKILL.mdcore3,306 bytes
Read SKILL.md
# 工作流管理
跟踪写章 → 评审 → 修订 → 应用 → 复评 → 交付。全部状态来自 OpenWrite 任务与章节工作简报,不要虚构本地阶段机,也不要调用已废弃的工作流命令。
## 作者可见阶段
1. 上下文预检 — `novel_context_preview`
2. 写章 — `novel_write_chapter` 或后台 `novel_task_create(type=chapter_write)` / `novel_multi_write`
3. 评审 — `novel_review_chapter` 或 `novel_task_create(type=chapter_review)`
4. 作者确认修订 — `novel_revisions_list` → apply / reject
5. 复评与交付 — 应用后复评;交付图为 writing → review → revision → application → rereview → closure
## 工具
View the complete source on GitHubforeshadowing-system
When your agent uses it
Use when user wants to manage foreshadowing (伏笔), track story hooks, or check pending reveals. Triggers include "伏笔", "埋伏笔", "待回收", "伏笔状态".
What it loads into context
Frontmatter
56 tokensAlways available to the agent
SKILL.md
1,004 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心概念
- 可用工具
- 使用示例
- 创建伏笔的时机
- 伏笔设计原则
- DAG验证规则
Files
SKILL.mdcore2,989 bytes
Read SKILL.md
# 伏笔管理系统
管理小说中的伏笔和悬念,确保故事线索连贯。
## 核心概念
伏笔(foreshadowing)是作者在前面章节埋下的线索,后面再回收。
**伏笔状态**:
- `埋伏` - 已埋下,等待回收
- `待收` - 即将回收
- `已收` - 已回收
- `废弃` - 放弃的伏笔
View the complete source on GitHubnovel-creator
When your agent uses it
Use when user wants to write chapters, generate drafts, or continue the story. Triggers include "写第", "生成章节", "续写", "草稿", "创作".
What it loads into context
Frontmatter
47 tokensAlways available to the agent
SKILL.md
1,148 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 触发条件
- 工作流程
- 工具
- 节拍
- 用户审核(强制)
- 前置条件
- 错误处理
- 与其他技能
Files
SKILL.mdcore3,820 bytestemplates/beat_templates.yamlasset5,424 bytestemplates/example_usage.yamlasset3,446 bytes
Read SKILL.md
# 小说创作编排
通过 openwrite-bridge 的 `novel_*` 工具写章。不要调用 Python 模块、内置 Writer 类或已废弃的大纲命令。
所有正文写入走预览/确认门控;已有正文不得覆盖。
## 触发条件
- "写第X章" / "写章节"
- "生成章节" / "生成草稿"
- "续写" / "继续创作"
- "创作" + 章节相关
## 工作流程
View the complete source on GitHuboh-story-long-analyze
When your agent uses it
证据化拆解用户合法提供的长篇小说,分析开篇契约、逐章推进、剧情单元、人物关系、世界设定、情绪节奏、伏笔和表达技法。用于拆书、分析黄金开篇、建立对标报告或提炼可迁移模块。
What it loads into context
Frontmatter
153 tokensAlways available to the agent
SKILL.md
596 tokensWhen the skill is activated
Reference documents
445 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,115 bytesagents/openai.yamlasset252 bytesreferences/method.mddoc1,614 bytes
Read SKILL.md
# Oh Story 长篇拆解
把拆解视为只读、转化性的文学分析。先确认材料来源、用户有权使用的范围、章节边界和期望深度;材料不完整时标注限制,不用常识补齐原作事实。
## 执行流程
1. 先做作品概览与开篇样本,产出快速预览;除非用户明确要求完整拆解,否则在扩大到全书前停靠确认。
2. 使用 `references/method.md` 依次处理章节边界、开篇、逐章摘要、剧情聚合、人物设定、汇总报告和表达技法。
3. 每条硬事实标注可定位证据,区分原文事实、分析推断、迁移建议和待验证假设。
4. 把具体桥段抽象为功能位、情绪机制和输入输出条件,不生成可替代原作的逐章复述。
5. 数据量过大时分批推进并报告已覆盖范围、失败项和下一停靠点,不口头宣称全书完成。
## OpenWrite 边界
View the complete source on GitHuboh-story-long-scan
When your agent uses it
研究长篇网文平台与榜单样本,分析读者契约、题材组合、主角机制、开篇兑现、更新阶段、升级空间和作者适配度。用于长篇扫榜、市场趋势观察、选题比较或开书前验证方向。
What it loads into context
Frontmatter
147 tokensAlways available to the agent
SKILL.md
560 tokensWhen the skill is activated
Reference documents
358 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- 数据边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,030 bytesagents/openai.yamlasset261 bytesreferences/method.mddoc1,341 bytes
Read SKILL.md
# Oh Story 长篇市场研究
把榜单当作有偏差的市场样本,不把排名本身当作选题答案。先确认平台、榜单口径、目标读者、题材范围、采样日期和作者自身优势。
## 执行流程
1. 收集可复核样本,并为每组数据记录平台、榜单类型、链接、访问日期、可见口径和样本量。
2. 使用 `references/method.md` 编码题材组合、情绪承诺、主角起点、核心机制、首个冲突、首次兑现和长期升级空间。
3. 先报告样本频次,再解释可能机制;把样本少的新组合标为信号,不直接称为趋势。
4. 输出多个选题候选,分别说明市场证据、作者适配、差异化、开篇验证方式、持续写作空间和风险。
5. 市场结论只作为 Goethe 的讨论输入,不直接改写项目资产。
## 数据边界
View the complete source on GitHubstyle-system
When your agent uses it
Use when user wants to initialize style, extract reusable writing signals from user-supplied text, compose a project style guide, or analyze style drift.
What it loads into context
Frontmatter
37 tokensAlways available to the agent
SKILL.md
903 tokensWhen the skill is activated
Reference documents
4,334 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心原则
- 目标
- Source Pack
- 支持的动作
- 提取规则
- 推荐格式
- 当前命令口径
Files
SKILL.mdcore3,215 bytesSKILL.mdcore3,215 bytesSKILL.mdcore3,215 bytesSKILL.mdcore3,215 bytesSKILL.mdcore3,215 bytesprompts/analyze_style.mddoc5,136 bytesprompts/batch_extract_style.mddoc1,210 bytesprompts/compose_style.mddoc1,592 bytesprompts/extract_style.mddoc1,606 bytesprompts/initialize_style.mddoc963 bytestemplates/analysis_report.yamlasset3,587 bytestemplates/composed_style.mddoc2,475 bytestemplates/extraction_report.yamlasset3,015 bytestemplates/fingerprint.yamlasset2,979 bytestools/README.mddoc1,079 bytes
Read SKILL.md
# 风格系统
## 核心原则
OpenWrite 不再内置任何参考作品风格库。
风格来源只分三层:
```text
Layer 1: craft/ -> 通用技法与去模板化规则
Layer 2: data/novels/{id}/data/sources/{sid}/ -> 从用户提供文本提取出的 source pack
Layer 3: data/novels/{id}/src/** -> 本书确认版设定与约束
```
最终写作时,这三层会被合成为:View the complete source on GitHubworld-query
When your agent uses it
Use when querying or maintaining novel world entities, character relationships, or the Studio relationship topology. Triggers include "世界观", "实体", "人物关系", "关系图", "拓扑", and "查询世界".
What it loads into context
Frontmatter
52 tokensAlways available to the agent
SKILL.md
642 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 查询
- 增量修改关系
Files
SKILL.mdcore2,453 bytes
Read SKILL.md
# 世界查询系统
把 `data/novels/{novel_id}/src/` 下的人物与世界实体文件作为唯一关系真源。Studio 拓扑和 ReAct 的 `get_world_relations` 必须调用同一个拓扑查询,不维护 `graph.yaml` 或第二套关系数据库。
## 查询
- 用 `query_world` 查询实体列表或详情。
- 用 `get_world_relations` 读取与 Studio 连续性页面同源的节点和边。
- Studio 关系拓扑支持按节点名称、ID、类型、摘要和关系文字搜索;命中节点会保留一跳相邻节点和连线作为上下文。
- ReAct 没有第二套拓扑搜索索引;需要搜索时读取 `get_world_relations` 后在同一份节点/边结果中按上述字段过滤。
- 先用 ID 定位;用户只提供名称时允许按 front matter 名称或 Markdown H1 定位。
兼容读取以下关系来源:
View the complete source on GitHubcanon
When your agent uses it
查设定。Use when the author asks 查设定, 人物卡, 世界观, or types /canon. Read registered assets and search; do not create a second lorebook.
What it loads into context
Frontmatter
45 tokensAlways available to the agent
SKILL.md
210 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore839 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 查设定
接到已有 `novel_*`。人物、地点、世界观以 OpenWrite 资产为准。
1. `novel_assets_list` 列出人物/世界资产(含名称与别名)。
2. 指定实体用 `novel_asset_read`。
3. 按名称或别名找不到时 `novel_search`(scope 可 `assets`)。
4. 关系与状态漂移用 `novel_continuity`。
向作者报告规范名、别名、摘要。改设定走 `novel_structured_change_plan(change_kind="asset")`;普通讨论不得写入。View the complete source on GitHubdialoguequality
When your agent uses it
Use when user wants to analyze dialogue style, extract speech patterns, compare character voices, or detect AI-sounding dialogue. Triggers include "对话", "口头禅", "角色声音", "对白", "AI味对话".
What it loads into context
Frontmatter
63 tokensAlways available to the agent
SKILL.md
1,363 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 当前数据来源
- 可用工具
- 什么时候用
- 工作方式
- 检查清单
- 输出建议格式
- 示例
- 注意
Files
SKILL.mdcore4,747 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 对话质量系统
用于分析角色对白、提取说话习惯、检查角色声音是否区分明显,以及识别 AI 味对话问题。
## 当前数据来源
对话分析不要只盯着对白片段本身。当前推荐读取顺序是:
1. `data/manuscript/arc_*/ch_*.md` — 实际章节正文
2. `src/characters/*.md` — 角色单源文档
3. `data/style/composed.md` — 当前作品风格约束
4. `src/outline.md` / canonical packet — 本章戏剧位置与章节目标View the complete source on GitHubdog-delivery-query
When your agent uses it
Query a chapter delivery DoG graph across manuscript, review, revision application, and re-review closure. Use after writing or reviewing a chapter, after revision actions, or when the user asks whether a chapter is truly ready to deliver.
What it loads into context
Frontmatter
61 tokensAlways available to the agent
SKILL.md
569 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 查询流程
- 状态解释
- 约束
Files
SKILL.mdcore2,245 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 章节交付 DoG 查询
章节写作、评审和修订工具会生成:
`data/novels/{id}/data/dog/deliveries/{chapter}/dog-graph.json`
这张图查询六个阶段:`writing → review → revision → application → rereview → closure`。
它不替代 OpenWrite 的写章、六域累加评审或修订服务。
## 查询流程
1. 从最近一次 `novel_write_chapter`、`novel_review_chapter`、`novel_revision_*`View the complete source on GitHubdog-import-query
When your agent uses it
Query a smart-import DoG graph after a novel has been split into chapters. Use to inspect import completeness, AI precheck errors, and pending outline/asset/canonical setup.
What it loads into context
Frontmatter
50 tokensAlways available to the agent
SKILL.md
434 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 工作流
- 约束
Files
SKILL.mdcore1,671 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 拆书导入 DoG 验收
`conductor/smart_import.py` 完成导入后会生成 `dog-graph.json`。这张图只验收导入
结果,不负责写入章节、资产或大纲。
## 工作流
1. 从导入输出中找到 `data/novels/{id}/data/dog/imports/{IMPORT_ID}/dog-graph.json`。
2. 依次调用 `dog_validate`、`dog_create`、`dog_run`、`dog_wait`,直到终态。
3. 用 `dog_status` 查看 `manifest` 和每个 `chapter-ch_*` 节点的 evidence。
4. 根节点通过后,读取 manifest 的 `construction.nextActions`,再调用规划和资产工具建立大纲、角色、世界观、进度和正典事件。
5. 资产/大纲建立完成后重新运行 `novel_review_chapter`,再用 `dog-review-query`View the complete source on GitHubdog-review-query
When your agent uses it
Query an existing hierarchical OpenWrite review-v2 artifact through dsh-dog. Use after novel_review_chapter or when inspecting six quality domains, evidence, coverage, blockers, delivery status, or the 37 legacy checks. Never rerun model review from DoG.
What it loads into context
Frontmatter
68 tokensAlways available to the agent
SKILL.md
683 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 工作流
- 状态约定
- 边界
Files
SKILL.mdcore2,596 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 分层评审 DoG 查询
这个 Skill 只查询 OpenWrite 已物化的 review-v2 artifact。OpenWrite 的
`novel_review_chapter` / conductor 是评审真源;DoG 读取 47 节点分层图:上下文、
六个质量域、硬门禁、确定性聚合和 37 个兼容查询 leaf。DoG 不调用评审模型。
## 工作流
1. 如果本轮还没有评审结果,先按 OpenWrite 写作流程调用 `novel_review_chapter`。
2. 找到评审输出目录中的 `dog-graph.json`。graph 与 `review.json`、`context.json`、
`domain-*.json`、`gate.json`、`aggregate.json` 和 `dim_01.json` …
`dim_37.json` 在同一目录。View the complete source on GitHubexport-book
When your agent uses it
导出。Use when the author asks 导出, 备份, 出书, EPUB, or types /export-book. Preflight first; never use the host /export session-log command for the manuscript.
What it loads into context
Frontmatter
50 tokensAlways available to the agent
SKILL.md
236 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore930 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 导出
接到已有 `novel_*`。成稿导出与会话日志下载不是同一条命令。
1. `novel_export_preflight`:阅读顺序、结构缺陷、写作目标、评审新鲜度、正文接纳、阻断项与警告。
2. 向作者说明 `backup`(可带警告下载)与 `delivery`(门禁阻断则不可交付)。
3. 作者确认后 `novel_export`,原样传入同一 `format`、`purpose` 与 `preflight_revision`。
不要用宿主 `/export` 下载会话日志来代替成稿。完整项目档案走 `novel_project_archive_action`。View the complete source on GitHubforeshadow
When your agent uses it
查伏笔。Use when the author asks 查伏笔, 待回收, 超期伏笔, or types /foreshadow. Read chapter obligations and continuity; do not keep a second foreshadow store.
What it loads into context
Frontmatter
54 tokensAlways available to the agent
SKILL.md
239 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore890 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 查伏笔
接到已有 `novel_*`。伏笔真源是 OpenWrite DAG 与章节工作简报。
1. `novel_status` 看全书待办。
2. 当前章用 `novel_chapter_work`:到期 `must_resolve`、超期 `overdue`、待埋入 `to_plant`。
3. 全书未回收与漂移用 `novel_continuity`。
向作者报告分类清单,不要输出内部 JSON 当主界面。写入走 `novel_structured_change_plan(change_kind="foreshadowing")`;仅在作者本轮已授权精确变更时才用 `novel_foreshadowing`。View the complete source on GitHubforeshadowing-system
When your agent uses it
Use when user wants to manage foreshadowing (伏笔), track story hooks, or check pending reveals. Triggers include "伏笔", "埋伏笔", "待回收", "伏笔状态".
What it loads into context
Frontmatter
62 tokensAlways available to the agent
SKILL.md
1,036 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心概念
- 可用工具
- 使用示例
- 创建伏笔的时机
- 伏笔设计原则
- DAG验证规则
Files
SKILL.mdcore3,181 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 伏笔管理系统
管理小说中的伏笔和悬念,确保故事线索连贯。
## 核心概念
伏笔(foreshadowing)是作者在前面章节埋下的线索,后面再回收。
**伏笔状态**:
- `埋伏` - 已埋下,等待回收
- `待收` - 即将回收
- `已收` - 已回收View the complete source on GitHublearn
When your agent uses it
写法记忆。Use when the author asks 记住写法, 成功钩子, 作者偏好, or types /learn. Store style, hooks, and preferences as Workspace-scoped OpenWrite assets; do not create a second memory store.
What it loads into context
Frontmatter
60 tokensAlways available to the agent
SKILL.md
354 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 读取
- 写入(预览/确认)
Files
SKILL.mdcore1,348 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 写法记忆
接到已有 `novel_*`。风格、成功钩子和作者偏好是当前 Workspace 作品资产,不是会话本地文件,也不是第二套 RAG / lorebook。
## 读取
1. `novel_status` 确认当前作品。
2. 风格与提取源:`novel_source_action`(`status_v2` / `profile_v2`)。
3. 已登记钩子:`novel_assets_list`,必要时 `novel_asset_read`。
没有 Workspace(session cwd 缺失)时这些工具返回 `WORKSPACE_CONTEXT_MISSING`,不要改写全局或别的项目。
View the complete source on GitHubnovel-creator
When your agent uses it
Use when user wants to write chapters, generate drafts, or continue the story. Triggers include "写第", "生成章节", "续写", "草稿", "创作".
What it loads into context
Frontmatter
53 tokensAlways available to the agent
SKILL.md
1,180 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 触发条件
- 工作流程
- 工具
- 节拍
- 用户审核(强制)
- 前置条件
- 错误处理
- 与其他技能
Files
SKILL.mdcore4,012 bytestemplates/beat_templates.yamlasset5,424 bytestemplates/example_usage.yamlasset3,446 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 小说创作编排
通过 openwrite-bridge 的 `novel_*` 工具写章。不要调用 Python 模块、内置 Writer 类或已废弃的大纲命令。
所有正文写入走预览/确认门控;已有正文不得覆盖。
## 触发条件
- "写第X章" / "写章节"
- "生成章节" / "生成草稿"
- "续写" / "继续创作"
- "创作" + 章节相关
View the complete source on GitHubnovel-reviewer
When your agent uses it
Use when user wants to check logic consistency, polish style, find plot holes, or review existing chapters. Triggers include "检查", "润色", "逻辑", "伏笔漏洞", "审查".
What it loads into context
Frontmatter
58 tokensAlways available to the agent
SKILL.md
1,219 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 触发条件
- 工作流程
- 工具
- 六域(计分)与硬门禁
- 报告给作者看的结构
- 错误处理
- 与其他技能
Files
SKILL.mdcore4,105 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 小说审查润色
通过 openwrite-bridge 的 `novel_*` 工具评审与修订。不要调用 Python 审查类或本地 33 维脚本。
正式评审是六域累加:质量分、覆盖率、硬门禁、交付状态分开判断;37 项只作展开查询。
`novel_config.yaml` 的 `form: web_novel` 时,情节/节奏域额外检查钩子、黄金三章、追读力;这些准则 `max=0`,不另起一套评分。黄金三章只适用于第 1–3 章。
## 触发条件
- "检查逻辑" / "逻辑检查"
- "润色" / "润色风格"
- "找伏笔漏洞" / "伏笔检查"
- "审查章节" / "审查"View the complete source on GitHuboh-story-deslop
When your agent uses it
诊断并精修网络小说中的机械 AI 读感,包括重复解释、句式过度齐整、对话同声、抽象判断、万能身体反应、空泛氛围和无功能金句。用于去 AI 味、自然化润色或在保留作者声音的前提下压缩文本。
What it loads into context
Frontmatter
167 tokensAlways available to the agent
SKILL.md
586 tokensWhen the skill is activated
Reference documents
365 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,138 bytesagents/openai.yamlasset259 bytesreferences/method.mddoc1,313 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# Oh Story 去 AI 味
把 AI 味视为读感和叙事功能问题,不把词表命中当作错误。先确认用户要扫描、局部示范还是全文精修;默认先报告问题和样例,不擅自覆盖正文。
## 执行流程
1. 读取作者意图、人物声音、题材气质和当前段落功能,建立必须保留的内容。
2. 使用 `references/method.md` 扫描重复解释、句式齐整、对话同声、抽象判断和装饰性表达。
3. 按对读感的影响分级,优先修改高频且无功能的模式,不追求把所有句子变得不同。
4. 用最小改动恢复动作、选择、潜台词和节奏;结构或事实问题转交 `@oh-story-review`,不在润色中暗改。
5. 输出修改统计和少量前后对照;用户明确要求全文精修时才给完整候选稿。
View the complete source on GitHuboh-story-long-analyze
When your agent uses it
证据化拆解用户合法提供的长篇小说,分析开篇契约、逐章推进、剧情单元、人物关系、世界设定、情绪节奏、伏笔和表达技法。用于拆书、分析黄金开篇、建立对标报告或提炼可迁移模块。
What it loads into context
Frontmatter
160 tokensAlways available to the agent
SKILL.md
627 tokensWhen the skill is activated
Reference documents
445 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,308 bytesagents/openai.yamlasset252 bytesreferences/method.mddoc1,614 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# Oh Story 长篇拆解
把拆解视为只读、转化性的文学分析。先确认材料来源、用户有权使用的范围、章节边界和期望深度;材料不完整时标注限制,不用常识补齐原作事实。
## 执行流程
1. 先做作品概览与开篇样本,产出快速预览;除非用户明确要求完整拆解,否则在扩大到全书前停靠确认。
2. 使用 `references/method.md` 依次处理章节边界、开篇、逐章摘要、剧情聚合、人物设定、汇总报告和表达技法。
3. 每条硬事实标注可定位证据,区分原文事实、分析推断、迁移建议和待验证假设。
4. 把具体桥段抽象为功能位、情绪机制和输入输出条件,不生成可替代原作的逐章复述。
5. 数据量过大时分批推进并报告已覆盖范围、失败项和下一停靠点,不口头宣称全书完成。
View the complete source on GitHuboh-story-long-scan
When your agent uses it
研究长篇网文平台与榜单样本,分析读者契约、题材组合、主角机制、开篇兑现、更新阶段、升级空间和作者适配度。用于长篇扫榜、市场趋势观察、选题比较或开书前验证方向。
What it loads into context
Frontmatter
154 tokensAlways available to the agent
SKILL.md
593 tokensWhen the skill is activated
Reference documents
358 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- 数据边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,229 bytesagents/openai.yamlasset261 bytesreferences/method.mddoc1,341 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# Oh Story 长篇市场研究
把榜单当作有偏差的市场样本,不把排名本身当作选题答案。先确认平台、榜单口径、目标读者、题材范围、采样日期和作者自身优势。
## 执行流程
1. 收集可复核样本,并为每组数据记录平台、榜单类型、链接、访问日期、可见口径和样本量。
2. 使用 `references/method.md` 编码题材组合、情绪承诺、主角起点、核心机制、首个冲突、首次兑现和长期升级空间。
3. 先报告样本频次,再解释可能机制;把样本少的新组合标为信号,不直接称为趋势。
4. 输出多个选题候选,分别说明市场证据、作者适配、差异化、开篇验证方式、持续写作空间和风险。
5. 市场结论只作为规划阶段的讨论输入,不直接改写项目资产。
View the complete source on GitHuboh-story-long-write
When your agent uses it
规划和创作长篇网络小说,覆盖选题定位、读者契约、人物与世界设定、卷纲细纲、单章写作、日更续写和结构修订。用于开书、补纲、写指定章节、续写或回炉长篇正文。
What it loads into context
Frontmatter
156 tokensAlways available to the agent
SKILL.md
653 tokensWhen the skill is activated
Reference documents
622 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,381 bytesagents/openai.yamlasset279 bytesreferences/method.mddoc2,261 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# Oh Story 长篇写作
先读取当前作品状态和用户本轮目标,再选择开书、补纲、写指定章、续写或修订。裸调用时只诊断所处阶段并给出下一步,不自动生成正文。
## 执行流程
1. 明确本轮交付范围和停靠点。用户没有指定正文数量时最多处理一章;只要求规划时停在候选或大纲交付。
2. 先处理读者契约、人物、世界、关系、全书阶段、卷纲和细纲,再根据已确认资产写正文或做局部修订。
3. 使用 `references/method.md` 检查情绪承诺、主角代理权、冲突升级、剧情单元和章末钩子。
4. 写章前只召回不知道就会写错的资料:当前细纲、最近正文、出场人物状态、相关规则、伏笔和当前关注点。
5. 写完检查事实连续性、角色选择、信息变化、情绪结算和章节状态;结构问题优先于句子润色。
View the complete source on GitHuboh-story-review
When your agent uses it
对网络小说大纲、章节或短篇成稿做证据化商业审稿,按事实连续性、因果、人物选择、读者契约、节奏、信息、对话和表达分级问题。用于章节验收、整稿诊断、投稿前审查或生成优先修订清单。
What it loads into context
Frontmatter
165 tokensAlways available to the agent
SKILL.md
571 tokensWhen the skill is activated
Reference documents
468 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 执行流程
- OpenWrite 边界
Files
LICENSEasset1,081 bytesSKILL.mdcore2,096 bytesagents/openai.yamlasset264 bytesreferences/method.mddoc1,665 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# Oh Story 商业审稿
先确认审查对象、范围、目标读者和用户要“只诊断”还是“诊断后给修改稿”。默认只输出审查报告,不直接改动正文。
## 执行流程
1. 读取当前项目已确认资产、相关正文和审稿范围,先做缺失资料与冲突预检。
2. 使用 `references/method.md` 按阻断、严重、一般、润色四级检查,不让文风偏好掩盖事实或结构问题。
3. 每条发现给出位置或引文、问题、读者影响、证据和最小修复方向;证据不足时标记“需补充”。
4. 汇总最高优先级问题、可保留优点、建议修订顺序和仍需作者裁定的取舍。
5. 只有用户明确要求修改时才生成候选稿;涉及正式正文提交时继续走 OpenWrite workflow。
View the complete source on GitHubprogress
When your agent uses it
看进度。Use when the author asks 看进度, 写到哪, 现在哪一章, or types /progress. Read OpenWrite status; do not invent a local workflow store.
What it loads into context
Frontmatter
47 tokensAlways available to the agent
SKILL.md
212 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore849 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 看进度
接到已有 `novel_*`。不要另建进度文件或阶段机。
1. `novel_status`:当前作品、阶段、资产缺口、待确认项。
2. 需要章级细节时 `novel_chapter_work`。
3. 后台工作 `novel_tasks_list`;进行中或待确认的任务 `novel_task_get`。
向作者报告:当前章、正文是否存在、评审是否新鲜、任务 phase、伏笔到期/超期/待埋入计数。不要把内部 JSON 当主界面。
细则见 workflow-manager。View the complete source on GitHubreview-chapter
When your agent uses it
审这一章。Use when the author asks 审这一章, 审查, 检查逻辑, or types /review-chapter. Run six-domain novel_review_chapter; do not rerun review from DoG.
What it loads into context
Frontmatter
57 tokensAlways available to the agent
SKILL.md
299 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore1,101 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 审这一章
接到已有 `novel_*`。正式评审是六域累加;37 项只作展开查询。
网文项目(`form: web_novel`)另附钩子 / 黄金三章 / 追读力可选准则,不计分。
1. `novel_outline_read` 与 `novel_status` 确认章节。
2. `novel_chapter_work` 读取正文 revision、评审新鲜度、本章伏笔义务。
3. 尚无正文则提示先写章,不要空审。评审 stale 时先复评。
4. `novel_review_chapter` 跑六域评审。
5. 向作者报告质量分、覆盖率、门禁、交付状态,以及带正文证据的问题。
作者要改稿时转到 /revise-span。查询已物化评审图用 dog-review-query,不要让 DoG 再调模型。细则见 novel-reviewer。View the complete source on GitHubrevise-span
When your agent uses it
改这段。Use when the author asks 改这段, 扩写, 缩写, 去 AI 味, 按审稿改, or types /revise-span. Stage revision hunks; never PUT the manuscript.
What it loads into context
Frontmatter
58 tokensAlways available to the agent
SKILL.md
277 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore1,044 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 改这段
接到已有 `novel_*`。改稿只走修订提案,不直接整章覆盖。
1. 选区扩写/缩写/去 AI 味:`novel_revision_create_selection`。
2. 按审稿改:`novel_revision_create_from_review`(带 issue 与期望 revision)。
3. 展示 before/after;作者确认后 `novel_revision_apply`(可带 `selected_hunk_ids`)。
4. 提案 `source_revision` 与当前正文不一致时拒绝覆盖,改为 `novel_revision_regenerate` 或重新预览。
5. 应用后必须 `novel_review_chapter` 复评。
手工/润色类正文预览用 `novel_document_change_plan`。没有当前评审时,按审稿改失败闭合。细则见 novel-reviewer。View the complete source on GitHubstyle-system
When your agent uses it
Use when user wants to initialize style, extract reusable writing signals from user-supplied text, compose a project style guide, or analyze style drift.
What it loads into context
Frontmatter
43 tokensAlways available to the agent
SKILL.md
935 tokensWhen the skill is activated
Reference documents
4,334 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心原则
- 目标
- Source Pack
- 支持的动作
- 提取规则
- 推荐格式
- 当前命令口径
Files
SKILL.mdcore3,407 bytesSKILL.mdcore3,407 bytesSKILL.mdcore3,407 bytesSKILL.mdcore3,407 bytesSKILL.mdcore3,407 bytesprompts/analyze_style.mddoc5,136 bytesprompts/batch_extract_style.mddoc1,210 bytesprompts/compose_style.mddoc1,592 bytesprompts/extract_style.mddoc1,606 bytesprompts/initialize_style.mddoc963 bytestemplates/analysis_report.yamlasset3,587 bytestemplates/composed_style.mddoc2,475 bytestemplates/extraction_report.yamlasset3,015 bytestemplates/fingerprint.yamlasset2,979 bytestools/README.mddoc1,079 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 风格系统
## 核心原则
OpenWrite 不再内置任何参考作品风格库。
风格来源只分三层:
```text
Layer 1: craft/ -> 通用技法与去模板化规则
Layer 2: data/novels/{id}/data/sources/{sid}/ -> 从用户提供文本提取出的 source pack
Layer 3: data/novels/{id}/src/** -> 本书确认版设定与约束
```View the complete source on GitHubtruth-validation
When your agent uses it
Use when user wants to validate story consistency, check for plot holes, or verify state changes. Triggers include "验证", "一致性", "检查逻辑", "状态冲突".
What it loads into context
Frontmatter
52 tokensAlways available to the agent
SKILL.md
909 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 核心概念
- 验证问题类型
- 可用工具
- 使用示例
- 常见问题及修复
- 验证时机
- 与伏笔系统的关系
Files
SKILL.mdcore3,061 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 真相文件验证系统
验证真相文件与章节内容的一致性,确保故事逻辑连贯。
## 核心概念
真相文件是记录世界当前状态的 Markdown 文件,包括:
- `current_state.md` - 世界当前状态
- `ledger.md` - 资源账本
- `relationships.md` - 角色关系矩阵
相关运行时文件(用于一致性校验):View the complete source on GitHubworkflow-manager
When your agent uses it
Use when user wants to check writing progress, manage workflow stages, or resume interrupted tasks. Triggers include "工作流", "进度", "阶段", "写作流程".
What it loads into context
Frontmatter
50 tokensAlways available to the agent
SKILL.md
972 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 作者可见阶段
- 工具
- 用法
- 状态怎么说
- 与其他技能
Files
SKILL.mdcore3,498 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 工作流管理
跟踪写章 → 评审 → 修订 → 应用 → 复评 → 交付。全部状态来自 OpenWrite 任务与章节工作简报,不要虚构本地阶段机,也不要调用已废弃的工作流命令。
## 作者可见阶段
1. 上下文预检 — `novel_context_preview`
2. 写章 — `novel_write_chapter` 或后台 `novel_task_create(type=chapter_write)` / `novel_multi_write`
3. 评审 — `novel_review_chapter` 或 `novel_task_create(type=chapter_review)`
4. 作者确认修订 — `novel_revisions_list` → apply / reject
5. 复评与交付 — 应用后复评;交付图为 writing → review → revision → application → rereview → closure
View the complete source on GitHubworld-query
When your agent uses it
Use when querying or maintaining novel world entities, character relationships, or the Studio relationship topology. Triggers include "世界观", "实体", "人物关系", "关系图", "拓扑", and "查询世界".
What it loads into context
Frontmatter
58 tokensAlways available to the agent
SKILL.md
674 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 查询
- 增量修改关系
Files
SKILL.mdcore2,645 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 世界查询系统
把 `data/novels/{novel_id}/src/` 下的人物与世界实体文件作为唯一关系真源。Studio 拓扑和 ReAct 的 `get_world_relations` 必须调用同一个拓扑查询,不维护 `graph.yaml` 或第二套关系数据库。
## 查询
- 用 `query_world` 查询实体列表或详情。
- 用 `get_world_relations` 读取与 Studio 连续性页面同源的节点和边。
- Studio 关系拓扑支持按节点名称、ID、类型、摘要和关系文字搜索;命中节点会保留一跳相邻节点和连线作为上下文。
- ReAct 没有第二套拓扑搜索索引;需要搜索时读取 `get_world_relations` 后在同一份节点/边结果中按上述字段过滤。
- 先用 ID 定位;用户只提供名称时允许按 front matter 名称或 Markdown H1 定位。
View the complete source on GitHubwrite-next
When your agent uses it
写下一章。Use when the author asks 写下一章, 续写, 生成章节, or types /write-next. Preview context then call novel_write_chapter; never overwrite existing prose.
What it loads into context
Frontmatter
48 tokensAlways available to the agent
SKILL.md
270 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
Files
SKILL.mdcore1,055 bytes
Read SKILL.md
Language: follow the author's current conversation language for responses and help; preserve the manuscript language. Translate these instructions when presenting them.
# 写下一章
接到已有 `novel_*`。已有正文不得覆盖。
1. `novel_status` 看阶段与待确认项;`outline_scope` 待确认时先让作者确认大纲范围。
2. `novel_outline_read` 定位章纲:未指定则取大纲顺序中最早尚无正文的章。
3. `novel_context_preview` 预检上下文包。
4. `novel_chapter_work` 读取本章到期/超期/待埋入伏笔与字数目标。
5. `novel_write_chapter` 写下一章或指定章。
6. 草稿出来后暂停,请作者选择:通过进入评审、按意见重写、或到创作页手工编辑。
资产不齐时先规划补齐,不切换会话。节拍与审核口令见 novel-creator。View the complete source on GitHubdog-v02-agentic-ci
When your agent uses it
当一个目标是非形式化的、没有唯一判准的"大目标"时使用(例:"做一份高质量 PPT""写一篇真正的好文章""写一份可读性强的复盘报告")——这种目标不能直接判断"做到了没",需要:①把它拆成多个可以单独评看子目标(子目标本身也大多是非形式化的语义目标——好坏仍要靠判断力,所以由子代理 agentic 判定;只有少数能精确成规则的才用脚本 programmatic);②系统逐项验证;③汇总。效果:得到质量门禁报告——每个子目标通过/不通过/需要人看 + 证据,以及总体结论与失败原因。
What it loads into context
Frontmatter
204 tokensAlways available to the agent
SKILL.md
5,888 tokensWhen the skill is activated
Reference documents
0 tokensLoaded on request
Token counts are estimates. Scripts run separately; they are not counted as reference context.
The playbook
- 0. 什么时候使用这个 skill(什么时候不用)
- 1. 本质(别当它是"检查清单")
- 2. 图(完整定义)
- 3. 完整示例一:写一篇"真正有质量"的文章
- 4. 完整示例二:做一份"高质量答辩 PPT"
- 5. 怎么让它跑
- 6. 结果里每个状态是什么意思
- 7. 常见错误
- 8. 宿主维护(模型不用动)
- 9. 纪律与实战教训(务必遵守)
Files
SKILL.mdcore18,608 bytesSKILL.mdcore18,608 bytes
Read SKILL.md
# DoG(DAG of Goals)使用说明
> DoG = **DAG of Goals**(目标的有向无环图):把一个大目标拆成一棵"目标树",每个小目标可以被单独判定,系统逐项判定后汇总。
> 本说明对应 **dsh-dog v1.2.0**(协议 schemaVersion `0.9`);协议字段如下文所述,产品版本演进不改协议号。
> **需要完整操作手册(全部工具字段、状态语义、错误诊断、最佳实践)请读 `docs/agent-guide.md`。**
## 0. 什么时候使用这个 skill(什么时候不用)
**用**:你的任务目标是一句"非形式化"的话——"做得高不高/好不好"没有唯一判准,不能一眼说"做到了":
- "做一份高质量的答辩 PPT"
- "写一篇真正有质量的文章"(不是字数够、没有错别字那种)
- "重写一段可读性强的说明"
**不用**:能一条规则直接判断的("运行测试""改这个文件名")。View the complete source on GitHubSkills included
- dialoguequality
- dog-delivery-query
- dog-import-query
- dog-review-query
- novel-reviewer
- oh-story-deslop
- oh-story-long-write
- oh-story-review
- truth-validation
- workflow-manager
- foreshadowing-system
- novel-creator
- oh-story-long-analyze
- oh-story-long-scan
- style-system
- world-query
- canon
- dialoguequality
- dog-delivery-query
- dog-import-query
- dog-review-query
- export-book
- foreshadow
- foreshadowing-system
- learn
- novel-creator
- novel-reviewer
- oh-story-deslop
- oh-story-long-analyze
- oh-story-long-scan
- oh-story-long-write
- oh-story-review
- progress
- review-chapter
- revise-span
- style-system
- truth-validation
- workflow-manager
- world-query
- write-next
- dog-v02-agentic-ci
Repository facts
- License
- Apache-2.0
- Created
- Mar 3, 2026
- Last pushed
- Oct 28, 2026
- Forks
- 131
- Open issues
- 14
- Project activity
- Active
Rank over the last 12 weeks
Now #946 · best #736 · down 102 in 30 days
- #946 on Oct 31
- #947 on Nov 1
- #953 from Nov 2 to Nov 3
- #949 on Nov 4
- #947 from Nov 5 to Nov 6
- #946 since Nov 7
81 earlier periods
- #835 on Aug 2
- #804 on Aug 7
- #802 on Aug 8
- #804 on Aug 9
- #829 on Aug 10
- #825 on Aug 11
- #816 on Aug 12
- #819 on Aug 13
- #800 on Aug 14
- #771 on Aug 15
- #790 on Aug 16
- #798 on Aug 17
- #813 on Aug 18
- #807 on Aug 19
- #797 on Aug 20
- #796 on Aug 21
- #748 on Aug 22
- #796 on Aug 23
- #810 on Aug 24
- #818 on Aug 25
- #796 on Aug 26
- #802 on Aug 27
- #828 on Aug 28
- #758 on Aug 29
- #763 on Aug 30
- #797 on Aug 31
- #816 on Sep 1
- #791 on Sep 2
- #795 on Sep 3
- #774 on Sep 4
- #753 on Sep 5
- #780 on Sep 6
- #795 on Sep 7
- #815 on Sep 8
- #797 on Sep 9
- #805 on Sep 10
- #778 on Sep 11
- #773 on Sep 12
- #772 on Sep 13
- #800 on Sep 14
- #780 on Sep 15
- #792 on Sep 16
- #790 on Sep 17
- #777 on Sep 18
- #769 on Sep 20
- #789 on Sep 21
- #809 on Sep 22
- #794 on Sep 23
- #763 on Sep 24
- #736 on Sep 25
- #738 on Sep 26
- #796 on Sep 28
- #821 on Sep 29
- #805 on Sep 30
- #752 on Oct 1
- #770 on Oct 2
- #746 on Oct 3
- #771 on Oct 4
- #781 on Oct 5
- #788 on Oct 6
- #803 on Oct 7
- #844 on Oct 8
- #806 on Oct 9
- #785 from Oct 10 to Oct 11
- #809 on Oct 12
- #813 on Oct 13
- #822 on Oct 14
- #809 on Oct 15
- #813 on Oct 16
- #792 on Oct 17
- #772 on Oct 18
- #818 on Oct 19
- #828 on Oct 20
- #834 on Oct 21
- #806 on Oct 22
- #812 on Oct 23
- #770 on Oct 24
- #937 on Oct 25
- #944 on Oct 28
- #947 on Oct 29
- #944 on Oct 30
Before you install
Instructions onlyAutomated signals, not a guarantee
Checked Nov 4, 2026
Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.
Show 9 notes
presets/goethe/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/goethe/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/goethe/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/goethe/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/openwrite/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/openwrite/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/openwrite/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedpresets/openwrite/skills/style-system/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzedvendor/dsh-dog/docs/skills/dog-v02-agentic-ci/SKILL.md:1Analysis incomplete: Referenced executable target unavailable; not analyzed
- No broad tool permissions
- No dangerous shell commands
- 10 network referencesin 10 files: oh-story-deslop/SKILL.md:35 oh-story-long-write/SKILL.md:36 oh-story-review/SKILL.md:35 and 7 more
- No hidden or obfuscated text
- No instruction overrides
- No executable scripts
- No access to secrets
- No requests to vote in Hot
Show all 10 signals by file
presets/dante/skills/oh-story-deslop/SKILL.md
- Network accessline 35
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-deslop`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/dante/skills/oh-story-long-write/SKILL.md
- Network accessline 36
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-write`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/dante/skills/oh-story-review/SKILL.md
- Network accessline 35
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-review`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/goethe/skills/oh-story-long-analyze/SKILL.md
- Network accessline 36
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-analyze`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/goethe/skills/oh-story-long-scan/SKILL.md
- Network accessline 35
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-scan`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/openwrite/skills/oh-story-deslop/SKILL.md
- Network accessline 38
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-deslop`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/openwrite/skills/oh-story-long-analyze/SKILL.md
- Network accessline 39
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-analyze`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/openwrite/skills/oh-story-long-scan/SKILL.md
- Network accessline 38
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-scan`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/openwrite/skills/oh-story-long-write/SKILL.md
- Network accessline 39
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-long-write`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
presets/openwrite/skills/oh-story-review/SKILL.md
- Network accessline 38
本 Skill 改编自 [oh-story-claudecode](https://github.com/worldwonderer/oh-story-claudecode) 的 `story-review`,基于提交 `48a4789d1f542eb672addf8ccb5dbdc20e63be46`;MIT 许可见同目录 `LICENSE`。
How the signals fit together
Facts from a static read of the files at the pinned commit. They describe what the code does, not why.
Some files could not be read; the facts below still hold.
Signal confidence: 10 facts
- Network access in prose of the SKILL.mdoh-story-deslop/SKILL.md:35Low confidence
- Network access in prose of the SKILL.mdoh-story-long-write/SKILL.md:36Low confidence
- Network access in prose of the SKILL.mdoh-story-review/SKILL.md:35Low confidence
- Network access in prose of the SKILL.mdoh-story-long-analyze/SKILL.md:36Low confidence
- Network access in prose of the SKILL.mdoh-story-long-scan/SKILL.md:35Low confidence
- Network access in prose of the SKILL.mdoh-story-deslop/SKILL.md:38Low confidence
- Network access in prose of the SKILL.mdoh-story-long-analyze/SKILL.md:39Low confidence
- Network access in prose of the SKILL.mdoh-story-long-scan/SKILL.md:38Low confidence
- Network access in prose of the SKILL.mdoh-story-long-write/SKILL.md:39Low confidence
- Network access in prose of the SKILL.mdoh-story-review/SKILL.md:38Low confidence
Works with
| Agent | Status | Evidence | Install to |
|---|---|---|---|
| Claude Code | Native | Frontmatter field user-invocable in presets/openwrite/skills/dialoguequality/SKILL.md | .claude/skills/<skill-name>/ |
| Codex | Compatible | Standard SKILL.md presets/dante/skills/dialoguequality/SKILL.md | .agents/skills/<skill-name>/ |
| Cursor | Compatible | Standard SKILL.md presets/dante/skills/dialoguequality/SKILL.md | .cursor/skills/<skill-name>/ |
| Gemini CLI | Compatible | Standard SKILL.md presets/dante/skills/dialoguequality/SKILL.md | .gemini/skills/<skill-name>/ |
| GitHub Copilot | Compatible | Standard SKILL.md presets/dante/skills/dialoguequality/SKILL.md | .github/skills/<skill-name>/ |
| OpenCode | Compatible | Standard SKILL.md presets/dante/skills/dialoguequality/SKILL.md | .opencode/skills/<skill-name>/ |
- Native The repository ships a manifest, folder or frontmatter field for this agent.
- Compatible A standard SKILL.md that this agent can read.
- Unknown No evidence either way.
Derived automatically from paths, manifests and frontmatter. Not tested on each agent.
Explore alternatives and comparisons
Add this badge to your README
It updates every night and links to this page.
[](https://hotskills.dev/skill/lipu-jpg-openwrite/)