AI Agent Hub
Back to plugins
🖥️

dsh-auto-fold-turn

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install ycp424c/dsh-auto-fold-turn

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install ycp424c/dsh-auto-fold-turn in your DeepSeek Harness terminal to install the auto-fold plugin from https://github.com/ycp424c/dsh-auto-fold-turn.

About this plugin

In a DSH conversation, a single agent turn often involves multiple tool calls, intermediate reasoning steps, and model retries. These process nodes can make the chat UI feel cluttered, burying the actual answer beneath a wall of intermediate information. dsh-auto-fold-turn addresses this by automatically collapsing all process nodes that appear before the final reply of each turn into a single summary line (▶ Process · N items), giving users an immediate view of the core answer while keeping details one click away.

The plugin is built entirely on DSH's official Conversation Node extension interface without patching or modifying any DSH source code—disabling or uninstalling it restores the native chat view in full. Each turn folds independently, and any explicit expansion state a user sets is persisted per session and turn in localStorage, surviving page reloads until the user folds it back. A unified fail-open policy ensures that if the final reply cannot be reliably identified or the DOM structure does not match, no content is hidden, so nothing is lost.

Ideal for DSH Web users who rely heavily on tool calls, multi-step reasoning, or simply prefer a clean, answer-first chat experience.

Use Cases

  • Jump straight to the final answer after multi-step tool calls
  • Review intermediate reasoning or model-retry details of a specific turn on demand
  • Keep long conversations visually clean with clear answer-first hierarchy

Best For

  • DSH users who rely heavily on tool calls and multi-step reasoning
  • Developers who prefer an answer-first reading style with on-demand detail expansion
  • Chat users who want to reduce visual noise and focus on core replies