AI Agent Hub
Back to plugins
🖥️

dsh-code-collector

Client Updated 2026.08.27

Run the following command in DeepSeek Harness:

dsh plugin install bycall/dsh-code-collector

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

Run dsh plugin install bycall/dsh-code-collector (repo: https://github.com/bycall/dsh-code-collector) in DeepSeek Harness, then add dsh-code-collector to the dsh.profile.bundles array in ~/.dsh/profiles/web/package.json and restart to activate.

About this plugin

In long DeepSeek Harness sessions, particularly when iterating on code, model-generated code blocks scatter across the conversation. Going back to find a specific snippet or compare file versions means manually scrolling through the entire chat. dsh-code-collector gathers every code block produced by the model into a persistent panel, turning them into a searchable, groupable, and exportable resource list.

The panel offers two views - by language (color dot, name, total lines) and by file (same-file iterations automatically form a version history that you can expand to inspect, copy, or jump to any prior version). Built-in real-time search filters by filename, language, or code content, case-insensitive. Batch operations include copying a single entry, copying all with filename header comments, and downloading everything as a single merged Markdown file. Clicking any entry smooth-scrolls to the originating turn and highlights it, eliminating the need to hunt through the conversation.

Ideal for developers who run extended coding sessions in dsh, perform code review, or iterate on refactors - especially when you need to audit model output, compare multiple file revisions across turns, or export code in bulk for downstream use.

Use Cases

  • Quickly locate a specific code snippet in a long session and jump to its originating turn
  • Compare version differences of the same file across multiple iterations and inspect any prior revision
  • Export all code blocks from a session as a single Markdown file with clear delimiters

Best For

  • Developers running extended coding and refactoring sessions in dsh
  • Code reviewers who need to audit and compare multi-turn model output
  • Engineers who want to batch-export AI session code for project reuse