WorkBuddy 跨端同步

SKILL运维和基础设施
运维和基础设施 未标注权限 许可证不明确
安装这个技能

通过云盘 + 软链接 + NTFS Junction,实现 WorkBuddy 身份文件和会话工作区的跨设备同步

一句话选型结论

按资料字段整理

WorkBuddy 跨端同步是运维和基础设施类的 Agent Skill,安装方式未标注,权限级别未标注,许可证不明确,有可直接复制的安装命令。技能本质是给模型读的说明书,效果取决于模型与客户端;装完先用一个小任务验证,再放进正式流程。

分类运维和基础设施 安装方式未标注 权限级别未标注权限 许可证许可证不明确
累计安装 未标注 本周 未标注
GitHub 星标 未标注 仓库 未标注
下载量 未标注 技能市场下载数
权限级别 未标注权限 安装前请确认风险范围
许可证 许可证不明确 未说明

安装

未标注
在支持 Skills 的客户端里执行,装完问一句「你有哪些技能」验证
skillhub install user_37f992cc/workbuddy-cross-device-sync
  1. 1
    确认客户端支持

    确认你的客户端支持 Skills(Claude、Codex、Cursor 等),并升级到较新的版本。

  2. 2
    拿到技能文件

    执行上面的命令安装;也可以把仓库里的 SKILL.md 手动放进客户端的 skills 目录。

  3. 3
    刷新技能列表

    重新打开客户端,或刷新技能列表,让新放进来的技能被加载。

  4. 4
    验证是否生效

    问一句「你有哪些技能」,确认列表里出现它,再拿一个小任务试跑。

技能档案

运维和基础设施
字段来自技能仓库与公开索引,中文说明由本站标注
作者 / 维护方user_37f992cc仓库 未标注 · 版本 未标注
宿主客户端未标注技能可被这些客户端加载
安装方式未标注安装包 community
权限级别未标注权限安装前请确认风险范围
许可证未说明许可证不明确,商用前请自行确认条款
仓库可用性available技能仓库当前是否可访问
最近核验未标注上游最后核验时间 · 数据更新 1791734504
正文体量11.9 KBSKILL.md 原文大小,越大说明说明书写得越细

典型使用场景

3 类
按分类整理的常见用法,具体以 SKILL.md 为准
查服务与日志

按现象定位服务状态与日志线索。

写部署与巡检脚本

把重复的部署、巡检写成脚本。

故障排查

按步骤排查网络、依赖与配置。

选择建议

装之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有可直接复制的安装命令,不用自己翻文档找入口。

先想清楚的地方

  • 没有标注权限级别,实际能做什么要装好后确认。
  • 许可证不明确,商用或二次分发前先到仓库确认条款。
  • 没有做站内核验,内容以作者仓库为准。
  • 技能只是说明书,实际效果取决于模型与客户端,装完先用小任务试一遍。

更适合

  • 需要在运维和基础设施场景里给模型补一套做法的人。
  • 愿意先读一遍 SKILL.md、确认它会让模型做什么的人。

不太适合

  • 对许可证要求很严、必须明确可商用的人。
  • 只想装完就自动变强、不打算看说明书的人。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
技能权限级别安装方式累计安装宿主客户端安装命令
WorkBuddy 跨端同步当前 未标注权限 未标注 未标注 未标注 有
Azure 部署 调用外部服务 GitHub 仓库 664.2 万 Codex、Claude、Cursor 有
Azure diagnostics 问题管理 调用外部服务 GitHub 仓库 657.6 万 Codex、Claude、Cursor 有
Azure aigateway 部署 调用外部服务 GitHub 仓库 655.3 万 Codex、Claude、Cursor 有
Azure AI 搜索 调用外部服务 GitHub 仓库 651.4 万 Codex、Claude、Cursor 有
Azure Cloud migrate 部署 调用外部服务 GitHub 仓库 650.7 万 Codex、Claude、Cursor 有

SKILL.md 全文

11.9 KB
模型实际读到的说明书原文,来自技能仓库

WorkBuddy Cross-Device Sync

Overview

WorkBuddy stores conversation history and identity files locally by default — they do not automatically sync across devices. This skill implements a two-layer sync architecture using a cloud drive's local sync folder combined with Windows symbolic links (for identity files) and NTFS junctions (for session directories), enabling cross-device continuity of WorkBuddy's memory and identity.

Problem Background

WorkBuddy's local storage architecture means that switching computers results in:

  • Loss of AI identity (SOUL.md, IDENTITY.md, USER.md, MEMORY.md) — the AI does not "recognize" the user
  • Loss of project workspace memory (.workbuddy/memory/ daily logs and project notes)
  • Inability to continue previous work context

Conversation history (stored in workbuddy.db) cannot be synced safely due to database file-lock risks. This skill syncs everything except the raw conversation database — identity, memory, and workspace files are all covered.

Solution Architecture

Cloud Drive Sync Folder (e.g., D:\BaiduSyncdisk\)
|
+-- WorkBuddy-Config\          <- Layer 1: Identity files
|   |-- SOUL.md                   -> symlink -> ~\.workbuddy\SOUL.md
|   |-- IDENTITY.md               -> symlink -> ~\.workbuddy\IDENTITY.md
|   |-- USER.md                   -> symlink -> ~\.workbuddy\USER.md
|   |-- MEMORY.md                 -> symlink -> ~\.workbuddy\MEMORY.md
|   +-- models.json               -> symlink -> ~\.workbuddy\models.json (optional, if custom models configured)
|
+-- WorkBuddy-Projects\        <- Layer 2: Session workspace
    +-- Sessions\                 -> junction -> C:\Users\<username>\WorkBuddy
        |-- 2026-07-13-17-16-19\
        |   +-- .workbuddy\memory\
        |       |-- 2026-07-13.md  (daily work log)
        |       +-- MEMORY.md      (project long-term memory)
        |-- 2026-07-19-15-13-10\
        +-- ...

What Gets Synced

Content Synced? Mechanism
Identity files (SOUL/IDENTITY/USER/MEMORY) Yes Symbolic link -> cloud sync folder
Custom model config (models.json) Yes (if exists) Symbolic link -> cloud sync folder
Session workspace (.workbuddy/memory/) Yes NTFS junction -> cloud sync folder
Daily work logs and project notes Yes Via session junction
Output files stored in workspace Yes Via session junction
Conversation history (workbuddy.db) No Local database, file-lock risk, do not sync

Prerequisites

Cloud Drive Requirements

The cloud drive must provide a local real-time sync folder — a folder on the local disk that the cloud client automatically synchronizes bidirectionally.

Cloud Drive Local Sync Folder Suitable?
Baidu Netdisk (sync space, SVIP) Yes Recommended for China users
OneDrive Yes Works but may be slow in China
Nutstore (坚果云) Yes Stable, recommended
WPS Cloud Drive Yes Works
Tencent Weiyun (微云) No (manual upload/download only) Not suitable
Baidu Netdisk (regular, no sync space) No Not suitable

System Requirements

  • Windows 10/11
  • PowerShell 5.1+ (built-in)
  • For symbolic links: either Developer Mode enabled or Administrator PowerShell
  • For NTFS junctions: regular PowerShell (no admin needed)

Finding the Cloud Sync Folder Path

Before starting, identify the cloud drive's local sync folder path. Common locations:

# Baidu Netdisk sync space
D:\BaiduSyncdisk\

# OneDrive
C:\Users\<username>\OneDrive\

# Nutstore
C:\Users\<username>\坚果云\  or  D:\Nutstore Files\

# WPS Cloud Drive
C:\Users\<username>\Documents\WPS Cloud Files\<account_id>\

To verify the sync folder exists and is working, create a test file inside it and check whether it appears on the other computer.

Setup Process

Phase 1: Prepare Cloud Drive (Both Computers)

  1. Install the cloud drive client on both computers and log in with the same account.
  2. Enable the local sync folder feature on both computers.
  3. Verify the sync folder path on each computer (paths can differ between computers).

Phase 2: Set Up Layer 1 — Identity Files (Computer A First)

Perform this on Computer A first (the computer where WorkBuddy identity files already exist).

Step 1: Create Cloud Folder

New-Item -ItemType Directory -Path "<sync_folder>\WorkBuddy-Config" -Force

Step 2: Move Identity Files to Cloud

$workbuddy = "$env:USERPROFILE\.workbuddy"
$cloud = "<sync_folder>\WorkBuddy-Config"

# Back up existing files
$files = @("SOUL.md", "IDENTITY.md", "USER.md", "MEMORY.md")
foreach ($f in $files) {
    $src = Join-Path $workbuddy $f
    if (Test-Path $src) {
        Copy-Item $src "$src.local-backup" -Force
        Move-Item $src (Join-Path $cloud $f) -Force
    }
}

Step 3: Create Symbolic Links

Option A — Administrator PowerShell (recommended):

Open PowerShell as Administrator, then:

$w="$env:USERPROFILE\.workbuddy";$c="<sync_folder>\WorkBuddy-Config";@("SOUL.md","IDENTITY.md","USER.md","MEMORY.md")|%{if(Test-Path "$w\$_"){Remove-Item "$w\$_" -Force};New-Item -ItemType SymbolicLink -Path "$w\$_" -Target "$c\$_"|Out-Null}

Replace <sync_folder> with the actual sync folder path.

Option B — Developer Mode enabled (no admin needed):

Enable Developer Mode in Windows Settings -> Privacy & Security -> For developers, then run the same command in regular PowerShell.

> Note: In WorkBuddy's sandbox environment, symbolic link creation may fail even with Developer Mode enabled. If the AI agent cannot create symlinks, instruct the user to run the command manually in an external PowerShell window.

Step 4: Verify

# Check that symlinks point to the correct targets
Get-Item "$env:USERPROFILE\.workbuddy\SOUL.md" | Select-Object Name, Attributes, Target

Expected output: Attributes should include ReparsePoint, and Target should show the cloud path.

Phase 3: Set Up Layer 1 — Identity Files (Computer B)

  1. Install WorkBuddy and log in with the same account, then close WorkBuddy.
  2. Wait for the cloud drive to sync WorkBuddy-Config folder to this computer.
  3. Verify the files exist: ls "<sync_folder>\WorkBuddy-Config\"
  4. Create symbolic links (same command as Phase 2 Step 3).
  5. If .workbuddy directory does not exist yet, create it first:
    powershell New-Item -ItemType Directory -Path "$env:USERPROFILE\.workbuddy" -Force
  6. Launch WorkBuddy and verify the AI recognizes the user's identity.

Phase 4: Set Up Layer 2 — Session Junction (Both Computers)

This step makes session workspace directories (including .workbuddy/memory/) sync across devices via an NTFS junction.

On Computer A (where existing sessions exist):

  1. Close WorkBuddy completely (including system tray icon).

  2. Run the setup script scripts/setup-cross-device-sync.ps1 with appropriate parameters:

powershell powershell -ExecutionPolicy Bypass -File "<skill_path>\scripts\setup-cross-device-sync.ps1" -Mode junction -SyncFolder "<sync_folder>"

Or run the commands manually:

```powershell
$workbuddy = "$env:USERPROFILE\WorkBuddy"
$target = "<sync_folder>\WorkBuddy-Projects\Sessions"

# Create target directory
if (-not (Test-Path $target)) { New-Item -ItemType Directory -Path $target -Force | Out-Null }

# Copy existing sessions to sync folder (merge, do not overwrite)
robocopy $workbuddy $target /E /XJ /R:3 /W:2 /NP /NDL /NFL

# Remove original directory and create junction
Remove-Item $workbuddy -Recurse -Force
New-Item -ItemType Junction -Path $workbuddy -Target $target
```

  1. Verify the junction:

powershell Get-Item $workbuddy | Select-Object Name, Attributes, Target

Attributes should include ReparsePoint.

  1. Launch WorkBuddy and verify sessions are accessible.

On Computer B:

  1. Close WorkBuddy completely.
  2. Wait for cloud sync to complete — verify Sessions folder exists in the sync folder.
  3. If C:\Users\&lt;username&gt;\WorkBuddy already exists:
    - If it contains local sessions, back them up or merge them first.
    - Remove the directory: Remove-Item &quot;$env:USERPROFILE\WorkBuddy&quot; -Recurse -Force
  4. Create the junction:

powershell $workbuddy = &quot;$env:USERPROFILE\WorkBuddy&quot; $target = &quot;&lt;sync_folder&gt;\WorkBuddy-Projects\Sessions&quot; if (Test-Path $workbuddy) { Remove-Item $workbuddy -Recurse -Force } New-Item -ItemType Junction -Path $workbuddy -Target $target

  1. Launch WorkBuddy and verify sessions from Computer A are visible.

Phase 5: Optional — Sync Custom Model Config

If the user has configured custom models (e.g., DeepSeek API), sync models.json as well:

$w = &quot;$env:USERPROFILE\.workbuddy&quot;
$c = &quot;&lt;sync_folder&gt;\WorkBuddy-Config&quot;
if (Test-Path &quot;$w\models.json&quot;) {
    Move-Item &quot;$w\models.json&quot; &quot;$c\models.json&quot; -Force
    New-Item -ItemType SymbolicLink -Path &quot;$w\models.json&quot; -Target &quot;$c\models.json&quot;
}

Usage Rules

After setup is complete, follow these rules to avoid data corruption:

  1. Never run WorkBuddy on both computers simultaneously. The session database and memory files can conflict if modified from two locations.
  2. Wait for sync to complete before switching. After closing WorkBuddy on Computer A, wait 1-5 minutes for the cloud drive to finish syncing before opening WorkBuddy on Computer B.
  3. The AI writes work logs to maintain continuity. After each substantive work session, the AI should append a note to .workbuddy/memory/YYYY-MM-DD.md. This is the primary mechanism for cross-device context continuation.
  4. Conversation history does not sync. The AI cannot recall the exact dialogue from a previous session on another computer. It relies on work logs and MEMORY.md files for context.

Verification Checklist

After completing setup on both computers, verify:

  • [ ] ~\.workbuddy\SOUL.md is a symbolic link pointing to the cloud folder
  • [ ] ~\.workbuddy\IDENTITY.md is a symbolic link pointing to the cloud folder
  • [ ] ~\.workbuddy\USER.md is a symbolic link pointing to the cloud folder
  • [ ] ~\.workbuddy\MEMORY.md is a symbolic link pointing to the cloud folder
  • [ ] ~\WorkBuddy is a junction pointing to the cloud Sessions folder
  • [ ] WorkBuddy launches normally on both computers
  • [ ] The AI recognizes the user's identity on both computers
  • [ ] Work logs written on Computer A are readable on Computer B after sync

Troubleshooting

For common issues and solutions, see references/troubleshooting.md.

Resources

scripts/

  • setup-cross-device-sync.ps1 — Automated setup script supporting two modes:
  • -Mode identity — Set up Layer 1 (identity file symbolic links)
  • -Mode junction — Set up Layer 2 (session directory NTFS junction)

references/

  • troubleshooting.md — Common problems, diagnostic commands, and solutions including:
  • Developer Mode not working in WorkBuddy sandbox
  • Symbolic link permission errors
  • Junction creation failures
  • Cloud sync conflicts
  • Missing files after sync

常见问题

5 问
答案由本站依据条目字段整理
WorkBuddy 跨端同步怎么安装?

在支持 Skills 的客户端里执行:skillhub install user_37f992cc/workbuddy-cross-device-sync。安装方式为未标注。装完刷新技能列表或重开客户端。

WorkBuddy 跨端同步支持哪些客户端?

资料没有标注宿主客户端,但只要客户端支持 Skills(Claude、Codex、Cursor 等),一般都能加载。

WorkBuddy 跨端同步权限高吗,安全吗?

权限级别为未标注权限。安装前先看它会读取或写入哪些内容。

WorkBuddy 跨端同步收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于技能本身和所用的模型或客户端,正式使用前按作者说明确认。

WorkBuddy 跨端同步用什么许可证?

许可证标注为 未说明(许可证不明确)。商用或二次分发前请到仓库确认条款原文。

热门专题

按场景挑技能
全部专题 →

技能内容以作者仓库为准;分类、权限级别与许可证状态的中文说明为中文整理,最近核验 未标注。