AI Agent Hub
Back to plugins
🖥️

dsh-task-board

Client Updated 2026.09.08

Run the following command in DeepSeek Harness:

dsh plugin install linsibin/dsh-task-board

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

Run dsh plugin install linsibin/dsh-task-board in your DeepSeek Harness terminal to install this plugin; the full source is at https://github.com/linsibin/dsh-task-board

About this plugin

DSH Desktop conversations are built around a linear Q&A flow, which makes it hard to keep track of multiple pending, in-progress, and finished work items at a glance. dsh-task-board injects a Tasks kanban tab into the conversation header, sitting side by side with Chat and Trajectory, so you can manage your backlog without leaving the window.

The plugin does one thing for real: it turns a kanban card into an executable DSH session. Drag a card into the In Progress column and it creates a brand-new session bound to the card work directory, sends the description as a queued prompt, and the agent actually does the work. Assistant output streams back onto the card in real time; when the turn ends and the session idles, the card slides to Done with a result summary. Failures and timeouts send it back to Backlog with a visible error. Cards carry a UUID, title, group, Markdown description, work directory, and file list, are collapsible by task group, and persist in browser local storage across reloads.

Built for developers who juggle multiple engineering tasks inside DSH Desktop - batch refactors, script debugging, doc generation - and want a single surface to see what is running, what finished, and what broke, with one drag to re-run.

Use Cases

  • Juggle multiple engineering tasks in a DSH conversation and drag a card to trigger real agent execution
  • Group refactors and debug sessions by project and watch live streaming output and status on each card
  • Retry a failed task with a single drag without re-filling description, workdir, or files

Best For

  • Developers running parallel tasks inside DSH Desktop
  • Users who prefer kanban-style workflows embedded in the conversation window
  • Engineering teams batch-running agent tasks and tracking their status