AI Agent Hub
Back to plugins
🖥️

dsh-ide-context

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install LSAI2023/dsh-ide-context

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

Run dsh plugin install LSAI2023/dsh-ide-context inside DeepSeek Harness to install this plugin; the full source code is available at https://github.com/LSAI2023/dsh-ide-context

About this plugin

Prompting an AI copilot while writing code comes with a recurring annoyance: the model has no idea what you are looking at. You have just selected a block of code and opened three related files, yet you must copy-paste all of it by hand. dsh-ide-context removes that friction. On every model turn it automatically injects the current IDE context — the list of open files, the active file path, the selection starting line and column, and the selected text — so the model already knows where you are without you having to explain.

Under the hood it reads the Claude Code IDE-integration bridge, the same .lock files and MCP-over-WebSocket protocol that the Claude Code CLI uses, which means a single bundle works with both IntelliJ IDEA and Visual Studio Code. Context is scoped to the session: the plugin matches the IDE whose workspace contains the working directory, filters open files and the selection to that project, and drops files from unrelated workspaces or virtual documents. The refresh cadence is tunable — set it to zero to inject only when the IDE state actually changes, or to a positive integer to suppress injections within that many milliseconds of the last one.

dsh-ide-context is aimed at developers who already work inside DeepSeek Harness and use IntelliJ IDEA or VS Code day to day. There is no new workflow to learn; boot a profile that lists the bundle and the model will simply see what is on your screen with every turn.

Use Cases

  • Ask questions about a selected code block without copying it into the chat box manually
  • Let the model reference all open files in your workspace when they work together
  • Drive DeepSeek Harness conversations seamlessly inside IntelliJ IDEA or VS Code

Best For

  • Developers who code daily in IntelliJ IDEA or VS Code
  • Engineers working in DeepSeek Harness who want less manual context-switching overhead
  • Programmers who want their AI assistant to automatically see the current editor state and selection