MIT DSH plugin that derives task cards from sessions and manages them on one board (full, right-docked, or floating) with acceptance, projects, and context carry-over.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-11-20
License
MIT
01What can it help you accomplish?
See which DSH sessions are running, blocked, or waiting on you
A kanban of task cards with goals, progress, blockers and deliverables, plus a Pending view that jumps back to the session
People running many DeepSeek Harness sessions across several workspaces
Accept deliverables and continue work with session context
Task detail with acceptance criteria and artifacts; after confirm-acceptance, a draft you place into a related or new session
Users who hand work between sessions or want a human acceptance trail
Group workspaces into projects and track due dates
Projects spanning workspaces with notes, due dates, a timeline of deliveries, and a schedule view
Users planning several parallel workstreams inside DSH
02How to install into DeepSeek Harness
Prerequisites
- DSH `>= 0.2.1-rc.5 < 0.3.0` with the bundled Node.js `>= 22.19.0`
- README lists only the macOS desktop build as verified; other desktop platforms are not verified one by one
Installation steps
- 01
In DSH open **Plugins → Discover**, add index `https://topmindspace.github.io/dsh-super-kanban/index.json`, open **Super Kanban**, pick a version and install
- 02
Or in **Plugins → Add plugin** enter `dsh-super-kanban@0.10.5` (any target version works)
- 03
Or add the Git repository URL or a local directory on the plugin page; check out a tag first to pin a version
- 04
Enable the plugin and refresh DSH
Verify the integration
- After installation, enable the plugin and refresh DSH
- Open the task board from the left sidebar; in settings confirm the active period, then press **Scan**
03DSH integration and capability boundaries
Native DSH plugin from the marketplace index, `dsh-super-kanban` on npm, or a Git/local path; it scans sessions to maintain task cards and shows one board full-page, docked right, or floating.
Incremental scan into task cards
DSH sessions in workspaces on the board→Task cards with goal, progress, blockers and deliverables; one task can link several sessions
Writes task and project notes under `~/.dsh/storages/super-kanban/`With AI reasoning on, the selected model receives compressed session summaries and related task info; polish and quick chat send your inputOne board, three placements
The same kanban data→Full main area, docked to the right of a session, or floating above it (only one at a time)
The floating entry stays on the session UI by default; you can drag it or hide it from the menu or settingsDeliverable evidence and acceptance
A task with acceptance criteria and artifacts→Preview artifacts, copy paths, locate files, and keep a record after you confirm acceptance
Continue with context
Acceptance criteria, decisions, open items and artifacts→A draft you review and place into a related session or a new session in the same workspace
You send the message yourself in the sessionShared models and streaming quick chat
Your prompt or the open task→Scan, re-analysis, polish and quick chat share the board model; Markdown answers you can copy, push to the session, or save as a task
Quick-chat history lives only in the current window until refresh or **New chat**
04Who is it for? When not to use it?
Good for
- People running many DeepSeek Harness sessions across several workspaces
- Users who hand work between sessions or want a human acceptance trail
- Users planning several parallel workstreams inside DSH
Not for
- Requires DSH `>= 0.2.1-rc.5 < 0.3.0` and Node.js `>= 22.19.0` bundled with DSH; no third-party runtime dependencies declared.
- Local notes under `~/.dsh/storages/super-kanban/` with no telemetry; with AI reasoning on, your model provider receives compressed session summaries and task context, and polish or quick chat sends your input.
05Compatibility, maintenance and safety notes
- Requires DSH `>= 0.2.1-rc.5 < 0.3.0` and Node.js `>= 22.19.0` bundled with DSH; no third-party runtime dependencies declared.
- Only the macOS desktop build is listed as verified; other desktop platforms are not verified individually.
- Scans default to at most 150 reasoning units per day; adjust the per-run budget and unfinished sessions continue in later scans.
- Local notes under `~/.dsh/storages/super-kanban/` with no telemetry; with AI reasoning on, your model provider receives compressed session summaries and task context, and polish or quick chat sends your input.
- The board extracts progress from sessions, but approval and sending still happen inside the session; locked fields keep manual edits across scans.
MIT · last push 2026-11-19 · release v0.10.5
06Frequently asked questions
How do I install Super Kanban in DSH?
Use the marketplace index in **Plugins → Discover** (`https://topmindspace.github.io/dsh-super-kanban/index.json`), enter `dsh-super-kanban@0.10.5` under **Plugins → Add plugin**, or add the Git URL or a local folder. Then enable the plugin and refresh DSH.
What data leaves my machine when scanning?
Task and project notes stay under `~/.dsh/storages/super-kanban/` and the plugin has no telemetry. With AI reasoning enabled, your chosen model receives compressed session summaries and some task context; polish and quick chat send the text you type.
Which DSH and Node versions does it need?
DSH `>= 0.2.1-rc.5 < 0.3.0` with Node.js `>= 22.19.0` that DSH ships. The README only lists the macOS desktop build as verified.
Why are some sessions still unscanned?
The default daily cap is 150 reasoning units. You can change the per-run budget, and sessions left over are picked up in the next scans.
How do acceptance and continue-task work?
Open task detail to review acceptance criteria and artifacts, then **Confirm acceptance** to keep a record. **Continue this task** builds a draft from criteria, decisions and open items; you place it in a related or new session and send it yourself.
07Related DSH workflows
dsh-agent-teams
by nanmicoder
AgentTeams 多智能体团队协作插件:自然语言驱动建团队、拉成员、拆任务声明依赖、成员间消息直达,Web 右上角活动面板
dsh-blender-plugin
by sixtysevenlf
Adds tools for Blender scene editing, viewport capture, render profiling, headless work, and verification.
dsh-free-search
by dddmuc
DSH 免费搜索插件,执行联网检索。
dsh-bili-scroll
by c1overzz
Schedules Bilibili discovery, reads subtitles, writes daily reports, and optionally posts a comment tagging you.
08Data and sources
Super Kanban 从 DeepSeek Harness(DSH)的多条会话中提炼任务,生成任务卡,并随会话推进更新进度、阻塞和交付物。
- **全局、右侧与悬浮** · 同一块看板可以占满主区域、停靠在会话右侧,或悬浮在会话之上,同一时刻只显示一种。
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-11-20. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.
