Windows-only disk tools for DSH: scan, rank usage, estimate cleanup and migrate supported development caches with backups and confirmation.
DSH integration
Native runtime
Author-claimed
Safety audit
Unaudited
Last verified
2026-10-23
License
MIT
01What can it help you accomplish?
Analyze Windows disks, estimate and clean predefined caches, and plan confirmed development-environment migrations
Visual scan reports, Top directories/files, cleanup estimates and migration plans
Windows DSH users managing local disk space and development caches
02How to install into DeepSeek Harness
Prerequisites
- Windows and DeepSeek Harness (dsh); GitHub installation does not require pnpm or a build.
Installation steps
- 01
Run dsh plugin --profile web add git+https://github.com/wee-b/disk-sentinel.git.
- 02
Run dsh web; the 磁盘哨兵 entry appears near the sidebar settings area.
- 03
For source development: git clone, cd, pnpm install, pnpm run build:client, then dsh plugin --profile web add .
Verify the integration
- Open the sidebar and confirm 磁盘哨兵; select a disk, start analysis and review capacity, Top directories/files and elapsed time.
Rollback
- dsh plugin --profile web remove @dsh-plugin/disk-sentinel
03DSH integration and capability boundaries
Windows DSH disk analysis, cleanup and development-environment migration UI
disk_scan
Directory tree, Top directories, Top files and elapsed time
clean_disk
list, estimate and clean phases for predefined junk categories
clean deletes data only after confirmation.dev_env_migrate
A one-time migration plan that executes only after user confirmation
04Who is it for? When not to use it?
Good for
- Windows DSH users managing local disk space and development caches
Not for
- Windows only; the current version uses PowerShell, drive letters and Windows user environment variables.
- clean is irreversible; follow list → estimate → user confirmation → clean and keep backups.
- Scanning reads selected disks; protected or busy paths may be skipped.
- Migration requires another disk with at least 105% of source size free; existing target contents require confirmation and are backed up first.
05Compatibility, maintenance and safety notes
- Windows only; the current version uses PowerShell, drive letters and Windows user environment variables.
- clean is irreversible; follow list → estimate → user confirmation → clean and keep backups.
- Scanning reads selected disks; protected or busy paths may be skipped.
- Migration requires another disk with at least 105% of source size free; existing target contents require confirmation and are backed up first.
Public GitHub repository; README claims
06Frequently asked questions
Can clean_disk delete files immediately?
No. Follow list → estimate → user confirmation → clean. estimate only predicts space and clean performs the irreversible deletion.
Which operating systems are supported?
Windows only. The README says macOS and Linux are not supported by the current PowerShell/drive-letter implementation.
What must be true before a development-environment migration?
The target must be on another disk with at least 105% of source size free. Existing target content requires confirmation and is backed up before replacement.
07Related DSH workflows
weknora
by tencent
Open-source LLM knowledge platform: turn raw documents into a queryable RAG, an autonomous reasoning agent, and a self-maintaining Wiki.
honcho
by plastic-labs
Memory library for building stateful agents
webnovel-writer
by lingfengqaq
A long-form Chinese web-novel writing assistant for Claude Code and DeepSeek Harness, designed to reduce memory loss and hallucinations in serialized projects of up to 2 million characters.
mirage
by strukto-ai
The World's First Virtual Terminal for AI Agents
08Data and sources
DeepSeek Harness 的 Windows 磁盘分析与安全清理插件。安装后会在 DSH Web UI 中增加
“磁盘哨兵”入口
This page is generated from the project’s public documentation, repository metadata and a structured parse of DSH Plugins; last verified on 2026-10-23. Found an error? Submit a correction.
Best DeepSeek Harness Plugins
Twelve plugins worth installing first — picked from the whole catalog, across every category.