dsh-plugin-dev
Skill24 starsChanged 38 days agoMITView on GitHub
DeepSeek Harness (DSH) 插件、扩展与技能开发权威规范。当需要为 DSH 编写自定义工具 (Tool)、拦截钩子 (Hook)、UI 扩展、服务提供方或打包发布 DSH 插件时使用。提供官方 Cordis 微内核架构契约、执行流水线、隔离调试与质量校验 SOP。
14 other projects use a version of this.
Adopted by
Most recently taken up by
- TrueHOOHA/dsh-plugin-dev-skilltheir own version, first seen Nov 7, 2026
- mahingbun-dev/dsh-plugin-dev-kittheir own version, first seen Nov 1, 2026
- caparisen/dsh-plugin-devtheir own version, first seen Oct 25, 2026
- wula1223/dsh-plugin-devtheir own version, first seen Oct 24, 2026
- Fabian-698/dsh-plugin-devtheir own version, first seen Oct 20, 2026
- birdmanmandbir/skillstheir own version, first seen Oct 9, 2026
Other versions14 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 14
- zimodzh/dsh-plugin-dev-skills★ 14same name
- hackerFish/awesome-dsh-skills★ 4same name
- abab996/dsh-plugin-dev★ 2same name
- TrueHOOHA/dsh-plugin-dev-skill★ 2same name
- wula1223/dsh-plugin-dev★ 2same name
- birdmanmandbir/skills★ 1same name
- BlueChonk/BlueChonk-skills★ 1same name
- Wenaixi/dsh-plugin-dev★ 1same name
- benz-ai-x/dsh-plugin-dev★ 0same name
- caparisen/dsh-plugin-dev★ 0same name
4 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
Files
DeepSeek Harness (DSH) 插件与扩展开发权威技能指南
本技能向 AI 编程助手交付开发、调试、测试与发布 DeepSeek Harness 插件的完整官方架构契约与标准作业程序(SOP)。
1. 核心铁律(违反必翻车)
- 绝对隔离测试:严禁直接在用户的日常宿主目录
~/.dsh中进行测试或写入配置文件。必须通过DSH_HOME="$(pwd)/.dsh-dev"配合本地.dsh-dev/目录进行隔离开发与冒烟测试。 - 严格外部化依赖:在构建打包脚本(esbuild / tsup)中,必须将所有
@deepseek-ai/*命名空间及其衍生包(@deepseek-ai/cordis、@deepseek-ai/dsh-tools等)声明为external。绝不能将运行时内置包打包进发布产物中。 - 规范输出与展示分离:工具主体
execute必须返回符合output.schema的纯数据结构;面向模型的自然语言由output.render交付;UI 客户端卡片由纯函数presentCall/presentResult交付。严禁在execute中直接返回格式化自然语言。 - 纯函数卡片原则:
presentCall和presentResult会在实时输出和历史日志回放时重复执行,严禁在其中进行 I/O 操作、异步读取或调用时钟/随机数。 - 协作式取消响应:所有异步 I/O、HTTP 请求与外部子进程执行必须透传并监听
exec.signal,确保超时或用户取消操作能被及时终止。
2. 任务路由与决策树
根据用户的具体需求,查阅对应参考文档与执行 SOP:
| 开发任务 | 核心入口 | 关键规范参考 |
| 从零初始化插件项目 | playbooks/01-scaffolding.md | 包含目录骨架、package.json、tsconfig.json 与 esbuild 外部化配置 |
| 开发模型工具 (Tool) | playbooks/02-tool-plugin.md | 参阅 references/tools-contract.md 查看参数校验与卡片规范 |
| 开发权限门禁 / 安全拦截 (Hook) | playbooks/03-hook-plugin.md | 参阅 references/execution-pipeline.md 查看流水线 5 阶段 |
| 编写单元测试 (Unit Test) | references/plugin-testing.md | 基于 Node 原生测试运行器的 Mock 上下文与取消测试 |
| 扩展配方与子代理 (Presets/Subagents) | references/presets-and-subagents.md | 包含 ctx.preset 注册与 ctx.subagents 隔离容器驱动 |
| 开发 UI 或事件流监听器 | references/ui-and-events.md | 包含 session/event 监听、assistant/chunk 流式处理与 followup/steer 驱动 |
| 开发/组织 DSH 技能包 (Skills) | references/skills-subsystem.md | 包含 6 级 Rank 发现优先级、<name>/SKILL.md 规范与会话目录按需披露机制 |
| 本地运行与隔离测试 | playbooks/04-isolated-dev.md | 包含 .dsh-dev 目录与 cordis.yml 装配清单配置 |
| 构建校验与发版前质检 | playbooks/05-verification.md | 包含 TypeScript 检查、产物审查、外部依赖泄漏排查与发布清单 |
| 生态目录收录与市场提交 | playbooks/06-ecosystem-listings.md | 包含 awesome-dsh-plugin 与 dsh-market 自动化收录提交 |
3. 标准开发工作流(Agent 必须遵循的步骤)
当用户要求「写一个 DSH 插件」时,Agent 必须按以下顺序推展:
[需求确认] -> [SOP 01: 创建脚手架] -> [SOP 02/03: 实现功能] -> [SOP 04: 隔离测试] -> [SOP 05: 质检与交付]- 第一步(选型与脚手架):
- 确定是纯后台插件(Host-only)还是全栈插件(Host + Client)。
- 按
playbooks/01-scaffolding.md生成基础工程结构,配置好构建脚本与@deepseek-ai/*外部化规则。
- 第二步(功能实现):
- 编写
src/index.ts,严格遵循 Cordis 插件规范(name,inject,apply)。 - 工具插件遵循
defineTool的强类型输入与输出 Schema。
- 编写
- 第三步(本地隔离验证):
- 运行构建生成
dist/index.js。 - 按
playbooks/04-isolated-dev.md在本地.dsh-dev中编写最小测试配置并验证加载,确保没有污染宿主。
- 运行构建生成
- 第四步(发版前质检):
- 按
playbooks/05-verification.md执行pnpm verify和pack --dry-run,确认依赖无泄漏后向用户交付。
- 按
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

