Back to directory

dsh-wechat-article

In a curated listMaintenance: Active

aiworkskills/dsh-wechat-article

Runs a WeChat Official Account workflow from topic selection through drafting, layout, images, and publication.

View on GitHubHomepage
$ git clone https://github.com/deepseek-ai/DeepSeek-Harness.git "$HOME/code/deepseek-harness"

14

stars

5

forks

TypeScript

Language

Apache-2.0

License

2026-08-17

Created

2026-09-10

Last push

Apache-2.0 DSH plugin that carries a WeChat Official Account article from topic selection to the WeChat draft box, keeping Markdown, layout HTML and images in your own project folder.

DSH integration

Native runtime

Author-claimed

Safety audit

Unaudited

Last verified

2026-11-21

License

Apache-2.0

01What can it help you accomplish?

  • Take a WeChat Official Account article from topic selection all the way to the WeChat draft box inside the chat

    Topic directions to choose from first, then a Markdown final draft, a layout HTML file, a cover and body images — all left in the project workspace, with a stop-and-ask whenever a decision belongs to you

    Developers and account owners who publish to a WeChat Official Account and would rather describe the article than click through a writer UI

  • Describe the account once and keep long-lived material instead of re-pasting it for every article

    Saved positioning, target readers, default author, content strategy, writing conventions and layout/cover/image style, plus reusable content directions that hold documents and pictures the agent fetches on demand

    Teams running one or more accounts whose product material, cases and images stay the same across articles

  • Review a finished article and revise one paragraph or one picture without leaving the conversation

    A creation panel listing every article by last update, with the Markdown final, the layout HTML, the cover and the body images previewable and quotable back into the main chat for a targeted redo

    Editors who want a specific paragraph or image regenerated instead of the whole piece

02How to install into DeepSeek Harness

Prerequisites

  • DeepSeek Harness `0.1.1-rc.2` — the version this community preview targets
  • Node.js `22.19`, or Node.js 24 and newer
  • pnpm 11 and git, in a macOS or Linux terminal (the only platform the README's script covers)

Installation steps

  1. 01

    Clone both repositories: `git clone https://github.com/deepseek-ai/DeepSeek-Harness.git "$HOME/code/deepseek-harness"` and `git clone https://github.com/aiworkskills/dsh-wechat-article.git "$HOME/code/aiworkskills/dsh-plugin"`.

    $ git clone https://github.com/deepseek-ai/DeepSeek-Harness.git "$HOME/code/deepseek-harness"

  2. 02

    Build Harness first: `cd "$HOME/code/deepseek-harness"`, then `pnpm install` and `pnpm run build`.

  3. 03

    Build the plugin: `cd "$HOME/code/aiworkskills/dsh-plugin"`, then `pnpm install` and `pnpm build`.

  4. 04

    Add the plugin to a profile and start the web UI: `pnpm dsh plugin --profile web add "$HOME/code/aiworkskills/dsh-plugin"`, then `pnpm dsh web --port 3081`.

Verify the integration

  • Open `http://127.0.0.1:3081/`: a “公众号” entry appears at the bottom of the left sidebar, and the plugin management page lists a single plugin, `aws-wechat-article`.

03DSH integration and capability boundaries

DSH integrationNative runtime

Native DSH plugin: built from source and added to a Harness profile with `pnpm dsh plugin --profile web add`, it puts a “公众号” entry at the bottom of the left sidebar that opens a WeChat workbench in the right panel, and the plugin management page lists it as `aws-wechat-article`

  • Topic-to-draft-box pipeline

    One sentence in the main chat saying what to write, plus the account configuration already saved in the project→Topic directions for you to confirm, then a Markdown final draft, a layout HTML file, a cover and body images, ending in the WeChat draft box

    writes the Markdown final, the layout HTML, the cover and the body images into the current project directorycreates a WeChat draft once account credentials are configured
  • Skill installation and updates

    One click on 安装公众号写作 Skill the first time the workbench opens→The nine writing Skills from aiworkskills/wechat-article-skills installed from GitHub, updatable later from the 更新 Skill button in the top-right corner

    fetches the Skill repository from GitHub over the networkthe Skill stays outside the plugin source, so plugin and Skill update independently
  • Account configuration pages

    Account positioning, target readers, default author, content strategy, writing conventions, layout/cover/body-image style and publishing credentials→A configuration applied to the current project, with a status check and a 技能配置 entry point for later changes

    writes the configuration into the current project directory
  • Article file browser with quoting

    Articles already produced in the project workspace→Each article listed by last update with its Markdown final, layout HTML, cover and body images previewable, and a quote button that puts one file or one image into the main chat

  • Writing material library

    Product introductions, business documents, cases and pictures grouped under named content directions→Material the agent reads on demand for the direction you name, preferring the matching images as body illustrations

    uploads and stores documents and images inside the project's material directions

04Who is it for? When not to use it?

Good for

  • Developers and account owners who publish to a WeChat Official Account and would rather describe the article than click through a writer UI
  • Teams running one or more accounts whose product material, cases and images stay the same across articles
  • Editors who want a specific paragraph or image regenerated instead of the whole piece

Not for

  • It is pinned to DeepSeek Harness `0.1.1-rc.2` and, because Harness is still in RC, installs next to the Harness source rather than as a standalone package.
  • The writing method is not in this repository: the nine Skills come from aiworkskills/wechat-article-skills and are installed from GitHub on first open, so the plugin alone does not make the workflow run.

05Compatibility, maintenance and safety notes

  • It is pinned to DeepSeek Harness `0.1.1-rc.2` and, because Harness is still in RC, installs next to the Harness source rather than as a standalone package.
  • The documented install path is a shell script for macOS or Linux terminals; the README gives no Windows equivalent.
  • The writing method is not in this repository: the nine Skills come from aiworkskills/wechat-article-skills and are installed from GitHub on first open, so the plugin alone does not make the workflow run.
  • Plugin and upstream Skill are both Apache License 2.0, and the plugin does not copy, modify or bundle the Skill source.
2026-08-172026-09-10Not specified by the author

Apache-2.0 · community preview, no release published yet (last push 2026-09-10)

06Frequently asked questions

How do I install it in DeepSeek Harness?

The README's path is from source: clone DeepSeek-Harness and this repository, run `pnpm install` and `pnpm run build` in both, then add the plugin with `pnpm dsh plugin --profile web add "$HOME/code/aiworkskills/dsh-plugin"` and start `pnpm dsh web --port 3081`. It targets DSH `0.1.1-rc.2` and needs Node.js `22.19` (or 24 and newer), pnpm 11 and git.

Where does it show up once DSH is running?

Open `http://127.0.0.1:3081/`: a “公众号” entry appears at the bottom of the left sidebar and opens the WeChat workbench in the right panel. The plugin management page lists a single plugin, `aws-wechat-article`, which is this one.

Do I need to install something else after adding the plugin?

Yes. The first time the workbench opens, click 安装公众号写作 Skill: the plugin pulls aiworkskills/wechat-article-skills from GitHub, because the nine Skills are deliberately kept out of the plugin source so the two can be updated separately. Deployments that already point `skillSourceDir` at a local Skill checkout skip this step.

How do I actually start an article?

Just say it in the main chat — for example, ask for a WeChat article introducing the DeepSeek Harness plugin system, aimed at TypeScript developers, with a few topic directions first. The agent proposes directions, waits for your confirmation, then drafts, reviews, lays out and illustrates the piece, stopping wherever a decision is yours.

Does it upload my articles or material anywhere?

No. Articles, configuration and writing material all stay in the project folder you opened as the workspace, and the plugin keeps no separate cloud library. Only the Skill installation fetches from GitHub, and a WeChat draft is created once you configure account credentials.

08Data and sources

  • Author-claimedgithub.com56e48fc6d1ca…

    **在对话里写完一篇公众号文章 —— 从选题一路到微信草稿箱。**

  • Author-claimedgithub.com56e48fc6d1ca…

    pnpm dsh plugin --profile web add "$HOME/code/aiworkskills/dsh-plugin"

  • Author-claimedgithub.com56e48fc6d1ca…

    安装成功后,左侧栏底部会出现“公众号”,插件管理页中只会显示一个插件:`aws-wechat-article`。

This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-11-21. Found an error? Submit a correction.

🏆

Best DeepSeek Harness Plugins

Twelve plugins worth installing first — picked from the whole catalog, across every category.

DSH Plugins is an independent community directory of DeepSeek Harness plugins. Not affiliated with or endorsed by DeepSeek. Third-party plugins are not security-audited — review the source before installing.

New DeepSeek Harness plugins, weekly. No spam.