AI Agent Hub
Back to plugins
🖥️

dsh-conversation-quote

Client Updated 2026.08.17

Run the following command in DeepSeek Harness:

dsh plugin install xjwwjx/dsh-conversation-quote

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

Run dsh plugin install xjwwjx/dsh-conversation-quote in your terminal to install this plugin; the source repository is https://github.com/xjwwjx/dsh-conversation-quote and no extra configuration is required.

About this plugin

By the middle of a long conversation you find yourself reaching back to an earlier line constantly, wanting the model to reason over that exact text. Copying, pasting, and appending a note is tedious and easy to get wrong. dsh-conversation-quote turns the workflow into two steps: select the text, tap Quote, done.

The whole experience lives in the browser. Select any text inside a chat node (user, assistant, tool, or injected context) and a small Quote pill floats above the selection. Click it and a compact quote tab appears above the composer, left-aligned; hover to expand the full text, click the × to remove that one entry. Stage as many quotes as you like while the input box stays clean for your actual question. Press Enter to send and every staged tab merges into the message as an attributed, citable block in staging order — no extra schema, prompt section, or request-header change, and zero impact on KV cache.

Who it is for: anyone who frequently needs to pin a specific earlier line in long-running conversations — debugging tool output, tracing an assistant reasoning chain, or building follow-up questions on top of injected context. No configuration, no model-side cooperation; install and go.

Use Cases

  • Pin a specific earlier line in a long conversation for follow-up
  • Circle a key tool-output fragment and ask the model to analyze it
  • Stage multiple quotes side by side for comparative follow-ups

Best For

  • Heavy conversational users who often reference earlier text
  • Engineers debugging AI tool pipelines
  • Developers who value precision and flow in multi-turn chat