Dsh Plugin Creator logo

Dsh Plugin Creator

Organization
lovstudio
dsh-plugin-creator

Create a @deepseek-ai/dsh-* plugin package end-to-end — choose the extension point or capability seam, scaffold the package, implement the tool/hook/service, and run the repo gates. 触发:新增插件 / 加工具 / 开发 capability。

Overview

Publisherlovstudio
Repositoryskills
Skill namedsh-plugin-creator
Stars
67
Forks
17
Bundled files
3
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 3 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by lovstudio on GitHub. Read the source before you install it.

Installation

Install the Dsh Plugin Creator AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/lovstudio/skills.git /tmp/skills
mkdir -p .claude/skills
cp -r /tmp/skills/skills/dsh-plugin-creator .claude/skills/dsh-plugin-creator
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Dsh Plugin Creator in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Dsh Plugin Creator on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Dsh Plugin Creator is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

DSH 插件工坊 · DSH Plugin Studio

Use this skill when adding or modifying a plugin (@deepseek-ai/dsh-* package) in this repository. It compresses the repo's extension-point map, package checklist, and testing policy into one ordered procedure. Each step names the authoritative file; the full grounded detail with per-step sources lives in references/plugin-dev-sop.md.

Triggers

Activate when the user asks to add a plugin, package, capability seam, tool, hook, or model provider to this repo.

  • "给本项目加一个插件" / "新增一个 capability" / "加一个模型工具" / "封一个 skill"
  • "create a dsh plugin", "add a package", "scaffold a tool", "add an llm adapter"

Do not activate when the task only runs a skill, reviews code, or fixes a bug in an existing package without adding a contribution.

Mental model

  • Everything is a plugin; there is no privileged core (docs/architecture.md). Extending dsh means mounting a plugin beside the others.
  • Two plugin forms (packages/AGENTS.md "Plugin exports"): a function plugin exporting name/inject/Config/apply with no default export, or a Service subclass with a default export. Never mix the forms.
  • Every contribution is an effect: ctx.effect() / ctx.on() / ctx.waterfall(). A registry's register() returns the disposer, so teardown and hot-reload unwind automatically.

Workflow

1. Decide the mechanism and home

  1. Read docs/architecture.md "Where new behavior goes" and pick the mechanism: register a ctx.tools tool, listen on an agent/* / tools/* event, or define a new capability seam.
  2. Check docs/cookbook/extension-cookbook.md "feature → mechanism map" for an existing mechanism before inventing one.
  3. A swappable capability needs all three seam roles — Service Definition, Service Provider, Consumer — never one role alone (docs/glossary.md capability-seam).
  4. Pick an existing group under packages/<group>/; a new group is a pure container.

2. Scaffold the package

Create packages/<group>/<pkg>/ with package.json, tsconfig.json, src/index.ts, src/types.ts (types only), src/invariant.ts, and a bilingual README.md + README.zh.md. Follow the package.json invariants (private, version aligned to root, type: module, cordis in both peer and dev deps, exact files list) and register the package in exactly one aggregate — tsconfig.host.json or tsconfig.client.json. Full checklist: docs/cookbook/adding-a-package.md.

3. Implement the plugin

  • Waterfall listeners must call next() to delegate; returning without it short-circuits (docs/cordis-primer.md).
  • Typed events use declaration merging and carry an @mode tag.
  • Name the current role from the role table (Controller/Store/Registry/Provider/Backend/…), not the first implementation.
  • Deployment-varying choices are Config fields, not DEFAULT_* constants; explicit resolve(request): Spec, no hidden ?? default in run().
  • Optional services use ctx.get(name); cross-boundary ids use Branded<B>.
  • For a tool, follow the execute() contract in docs/cookbook/adding-a-tool.md: typed args, one canonical JSON value, honor exec.signal, output.render owns prose, presenters stay pure.

4. Add the invariant

Every package owns src/invariant.ts registering its manifest name and checking one event or data relation, or stating No runtime invariant: with a reason (packages/AGENTS.md).

5. Test

Follow docs/testing.md: an HMR-safety test for every registry, per-file 100% coverage on packages/*/*/src, a non-unit REAL-composition test for any product-visible plugin, and a keyless snapshot in the same PR for any model- or user-visible change.

6. Document

Write the package README.md with service API, config, events, and extension points first, then the canonical ## Model Experience and ## Known Limitations and Deferred Work sequence. Pair it with README.zh.md, and add an Agent Note in the same PR for any non-trivial change.

7. Verify

sh
pnpm install
pnpm run doc-sync
pnpm run constraints && pnpm run typecheck && pnpm run lint
pnpm run build && pnpm run hygiene

Then run only the checks the changed surface reaches — do not default to the full suite.

8. Commit

Split independent changes; label kind/* plus every material area/*; keep the Agent Note in the same PR.

References

  • references/plugin-dev-sop.md — the full grounded SOP with per-step sources.
  • references/skill-composition.md — how this skill hands off to the dsh-* skill family.

通用反馈闭环

用户在 Skill 驱动任务中提出修改意见时,继续当前产物前必须执行:

  1. 先判断意见是 task-specific(仅本次)还是 reusable(可跨任务复用)。
  2. task-specific 只修改当前任务,不改 Skill。
  3. reusable 先确定作用域:领域规则先更新对应 canonical Skill;适用于所有 Skill 的规则先更新共享规范。
  4. 完成规则更新、版本、lint 与分发核验后,再把修改应用到当前任务。
  5. reusable 修改会使此前的“确认”“继续”“发吧”失效;完成当前产物修改和回读后必须停下,等待用户下一步指示,不自动进入发布、提交或其他外部写入。

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Dsh Plugin Creator AI skill do?

Create a @deepseek-ai/dsh-* plugin package end-to-end — choose the extension point or capability seam, scaffold the package, implement the tool/hook/service, and run the repo gates. 触发:新增插件 / 加工具 / 开发 capability。

Why use Dsh Plugin Creator on TypingMind?

Because you install it once and use it with any model. Dsh Plugin Creator is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Dsh Plugin Creator in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/lovstudio/skills/tree/main/skills/dsh-plugin-creator. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Dsh Plugin Creator?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Dsh Plugin Creator?

As many as you like. As long as a model supports skills, you can use Dsh Plugin Creator with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Dsh Plugin Creator AI skill free?

Yes. It is published on GitHub by lovstudio under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇