Skip to content

HuanLinOTO/dsh-plugin-auto-blame

60Last commit Aug 15, 2026

dsh-plugin-auto-blame DSH plugin

After the model completes a turn, this plugin sends the last 3 messages to an LLM and generates 3 critical follow-up requests. These suggestions appear as bubbles above the input bar and can be clicked to send directly. The generation is fire-and-forget, so it doesn't block the conversation.

How to install the dsh-plugin-auto-blame DSH plugin

dsh plugin --profile web add @huanlin/dsh-plugin-auto-blame

Copying does not run this command. Review the repository and version before installing the dsh-plugin-auto-blame DSH plugin.

dsh-plugin-auto-blame DSH plugin data source

dsh-plugin-auto-blame DSH plugin snapshot date: Aug 16, 2026

discovered

What the dsh-plugin-auto-blame DSH plugin can do

  • Automatically generates 3 critical follow-up suggestions after each AI turn
  • Displays suggestions as clickable bubbles above the input bar with a fade-in effect
  • Clicking a bubble sets the draft and submits it immediately
  • Shows a 'Leadership Vision' label with a DeepSeek blue shimmer while generating
  • Clears old suggestions when a new turn starts

Where the dsh-plugin-auto-blame DSH plugin fits

  • Quickly get critical perspectives on AI responses without manual prompting
  • Save time thinking of follow-up questions during research or debate
  • Maintain a critical mindset in long conversations by surfacing alternative angles

Who the dsh-plugin-auto-blame DSH plugin is for

  • DSH users who want automated follow-up suggestions
  • Users who need to critically evaluate AI outputs in real-time

dsh-plugin-auto-blame DSH plugin limitations

  • Only works in DSH web mode (dsh web)
  • Suggestions are silently dropped if LLM generation or parsing fails (no fallback)
  • Suggestions are cleared after each turn and cannot be reused

dsh-plugin-auto-blame DSH plugin: from the repository README

Quoted from the HuanLinOTO/dsh-plugin-auto-blame README, the upstream source of the dsh-plugin-auto-blame DSH plugin. Copyright remains with the original authors.

> 📌 本插件已收录于 [dshfind](https://dshfind.com/zh) 插件超市,点击上方徽章直达主页。 # dsh-auto-blame [![npm version](https://img.shields.io/npm/v/@huanlin/dsh-plugin-auto-blame)](https://www.npmjs.com/package/@huanlin/dsh-plugin-auto-blame) ![preview](docs/preview.webp) 当模型完成当前轮次对话后,将最后 3 条消息发送给 LLM,生成 3 条批判性跟进请求,显示在输入框上方作为可选项,点击直接发送。生成期间"领导视野"标签带 DeepSeek 蓝色流光扫过(同 Deep diving... 特效),建议到达后气泡依次淡入。 ## 工作原理 ``` [host] agent/turn-stopping 触发 → fire-and-forget 调 ctx.llm.stream() 生成 3 条毒舌跟进 → session.append('auto-blame/suggestions', { turn, suggestions }) → projection unit 折叠该事件 → session/projection 推送帧 ↓ [client] useProjection('autoBlame') 收到值 → conversation.composer.dock 渲染 3 个气泡 → 点击 → inputActions.setDraft(text) + submit() ``` - **fire-and-forget**:LLM 调用不阻塞 turn 关闭,建议晚几百 ms 出现 - **非 surface 事件**:`auto-blame/suggestions` 不进入 model-visible 历史,不干扰 agent loop - **失败静默**:LLM 失败 / 解析失败 → 不 append 事件,不显示气泡 - **点击发送**:走 InputBar 同一路径(`inputActions.setDraft + submit`) - **下一轮清空**:新 turn 开始时 projection 归零,旧气泡立即消失 ## 开发 ```sh pnpm run typecheck # tsc --noEmit pnpm test # vitest run(28 个单测) pnpm run build # tsc + tsdown → lib/index.js, lib/invariant.js, lib/client.js pnpm

Read the full READMERepository license: NOASSERTION

dsh-plugin-auto-blame DSH plugin questions

How do I install the Auto Blame plugin?

Run `dsh plugin --profile web add @huanlin/dsh-plugin-auto-blame` in your terminal. After installation, restart `dsh web` and perform a hard refresh in your browser (Ctrl+Shift+R) to load the plugin.

Why are no suggestions showing up after the model responds?

Suggestions are generated asynchronously; they may appear a few hundred milliseconds later. If the LLM call fails or the response cannot be parsed, the plugin silently suppresses the suggestions. Also make sure you are using DSH in web mode, as the plugin only works with the web interface.

Does Auto Blame affect the conversation history?

No. The suggestions event (`auto-blame/suggestions`) is a non-surface event that does not enter the model-visible history. It does not interfere with the agent loop or the conversation log.

Can I use Auto Blame with other DSH profiles (e.g., cli, terminal)?

The installation command uses `--profile web`, indicating the plugin is designed for the web profile. It likely does not work in CLI or terminal modes because it relies on the composer dock and UI elements available only in the web interface.

How do I disable or remove the Auto Blame plugin?

You can remove the plugin by running `dsh plugin remove @huanlin/dsh-plugin-auto-blame` while the web profile is active. Then restart `dsh web` and hard refresh the browser to fully unload it.