
Servosity MCP
io.github.Servosity由 GitHub 账号 Servosity 发布v0.6.13更新于 Nov 6, 2026
Fleet-wide backup triage, stale-backup detection, and cross-engine analytics for any MCP agent.
概览
面向 Servosity MSP 合作伙伴的第一方 MCP 服务器,让 AI 助手进行全客户备份分诊、过期备份检测和跨客户分析。
- 可以做什么
- 查询本地客户群镜像,获得按严重程度排序的关注清单、隔夜变化、过期备份集和整体健康度
- 审计保留策略、未配置备份的代理、故障聚类,并核对账单与发票
- 为过期备份客户起草跟进邮件,生成单个客户或季度末的 QBR 报告
- 按游标读取已记录的问题事件,并进行搜索和备份事实查询,不额外调用 Servosity
- 执行经预览的写入操作,如分诊的忽略、归档、重新激活、评论,以及原始资源创建与更新
- 适用场景
- 适合使用 Servosity 备份与灾难恢复、希望助手回答跨客户问题的 MSP 合作伙伴,例如哪些客户的备份已过期、隔夜发生了什么变化,这些信息在合作伙伴门户中分散在各客户页面。它补充而非取代合作伙伴门户。
- 运行要求
- 以本地 stdio 进程运行(Claude Desktop 的 .mcpb 包,或安装器将 servosity-cli 与 servosity-mcp 放入本机);部分安装包需要配套 CLI 或设置 SERVOSITY_MSP_CLI_PATH。需要 Servosity 合作伙伴 API 令牌(SERVOSITY_MSP_TOKEN)和 MSP 合作伙伴账户。可选配置包括 SERVOSITY_MSP_BASE_URL、PRINTING_PRESS_CLIENT_PROFILE、SERVOSITY_MSP_RESELLER_ID、SERVOSITY_MSP_READ_ONLY 和 SERVOSITY_MSP_NO_LOCAL_DATA。仅支持桌面端代理,暂不支持网页版 ChatGPT。
安装
在 SourceWeft 中
- 打开 控制台中的 Servosity MCP,将其添加到工作区。
- 为需要使用其工具的对话启用该服务。
Desktop only,通过 STDIO。 STDIO 服务会启动本地进程,因此需要 SourceWeft 桌面宿主。
其他 MCP 客户端
参照 仓库 中的启动说明。
README
Servosity + AI - for Claude, ChatGPT, Codex, Cursor, and any agent that speaks MCP
Published by Servosity Inc. for MSP partners. A first-party Claude Code Skill and MCP server for the Servosity backup and DR platform. Servosity is a trademark of Servosity Inc. Apache-2.0 licensed.
[Servosity demo - animated preview]
▶ Watch the 30-second demo - demo data is simulated; every command shown exists in the real CLI.
Add fleet-wide backup triage, stale-backup-set detection, overnight drift, per-client situational awareness, and cross-client analytics to the AI you already use - Claude Code, Claude Desktop, ChatGPT (desktop app), Codex, Cursor, Windsurf, Cline, Continue, Gemini, or GitHub Copilot. Free, open source, runs on your laptop. A local fleet mirror means your AI can answer cross-client questions the partner portal can't show on one screen - Friday-email-ready in seconds. Built for MSP partners. No code required.
Works with your agent
The five agents MSP owners actually use:
The ChatGPT desktop app runs servosity-mcp locally and shares MCP configuration with Codex (OpenAI). ChatGPT on the web needs a remote server, which msp-skills does not support yet. If your team works there, a hosted gateway is the better pick; check that your AI app is on the gateway's setup list. Setup steps: mcp-install.md.
Also works with Cursor, Windsurf, Cline, Continue.dev, Zed, GitHub Copilot, and Gemini CLI - plus Hermes and OpenClaw, both via MCP and the pre-wired skill frontmatter. Full per-tool wire-up: docs/which-agent.md.
Run more than one agent? Install across all 51+ supported agents in one command:
npx skills add Servosity/msp-skills@latest(requires Node.js, then run the per-skill installer for the CLI/MCP binaries). See docs/which-agent.md.
Install in 60 seconds
Fastest for Claude Desktop - one-click .mcpb
Claude Desktop bundle: the
.mcpblaunches on macOS (Intel and Apple Silicon), Windows x64 and Linux x64. Every tool shells out to the companionservosity-cli. Releases cut from 2026-09-18 on ship that CLI inside the bundle; older bundles contain only the MCP server, so if yours lacks the companion, run the Path A or Path B installer below first (or setSERVOSITY_MSP_CLI_PATHto an existing binary). Details: #331.
Download Servosity MCP (.mcpb) - then open Claude Desktop > Settings > Extensions and select the file. One click, no JSON, no shell. (Browse every Servosity release on the releases page.)
Prefer the Claude Code plugin? Add the marketplace once, then install - works immediately, no directory listing required:
Path A - paste one prompt into your AI agent (recommended)
Copy this into Claude Code, Codex CLI, or Claude Cowork:
Install the Servosity Skill and MCP server from Servosity/msp-skills in this agent workspace. If this workspace uses a POSIX shell (macOS, Linux, WSL, or Bash), run
bash <(curl -fsSL https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/servosity/install.sh). If it uses Windows PowerShell, runiwr -useb https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/servosity/install.ps1 | iex. Then authenticate withSERVOSITY_MSP_TOKEN=<your-partner-token> servosity-cli doctorand runservosity-cli --helpto explore.
The same prompt works in any agent that can run shell.
Path B - run the installer yourself
Windows (PowerShell):
macOS / Linux:
The installer drops both servosity-cli and servosity-mcp into your user bin path. Claude Code, Codex, and Cowork discover the Skill via SKILL.md in this directory.
Verify:
Upgrade to the latest version
The installer always fetches the current release - re-run it to upgrade:
macOS / Linux:
Windows (PowerShell):
Claude Desktop .mcpb users: download the latest .mcpb (top of this section) and re-select it in Settings > Extensions. Claude Code plugin users: /plugin update servosity@msp-skills.
Add to Claude Desktop, Cursor, Windsurf, Cline, Continue, Gemini, or Copilot
After the installer runs, see mcp-install.md and docs/which-agent.md for the per-agent wire-up. Claude Desktop's Settings > Extensions panel is the simplest path.
Install for Hermes
From the Hermes CLI:
Inside a Hermes chat session:
Hermes speaks MCP natively, so it can also use the servosity-mcp server directly - same install path, same env vars.
Install for OpenClaw
OpenClaw is GA. Install the skill directly from this repo:
Or tell your OpenClaw agent (copy this):
Install the servosity skill from https://github.com/Servosity/msp-skills/tree/main/skills/servosity. The
metadata.openclawfrontmatter declares the required CLI (servosity-cli) and env vars; runopenclaw doctorif anything is missing.
openclaw also speaks MCP natively (see openclaw mcp set in the OpenClaw MCP docs) so you can wire servosity-mcp as an MCP server alongside the skill.
Authenticate
Get your partner API token from the Servosity partner portal, then:
doctor confirms the token works and the local mirror is reachable before you run anything that touches client data. It exits 0 even when the credential is rejected, so scripts should add --fail-on error.
What this skill does
Outcome-first, with the single command that answers each question:
Issue changes as events
Servosity has no issue webhooks yet, so issues watch polls your reseller's issue list (read-only, every 5 minutes by default) and records each change as an event your agent reads by cursor.
- Events:
servosity.issue.created,servosity.issue.updated(state, severity, priority,updated_atorlast_seen_atchanged),servosity.issue.archived,servosity.issue.ignored,servosity.issue.reactivated.--typenarrowsevents listto one. - Each event carries issue and company ids and triage fields only, never the description or comments. Run
issues read <id>for the rest. - The issue list holds active issues only, so an active issue that leaves it is re-read to learn whether it was archived or ignored, and an ignored issue is re-read at most hourly to catch it being archived or changed. An issue you can no longer read is dropped.
- Only GET requests reach Servosity. The first poll must read every page before it records the baseline. A failed poll waits twice as long each time, at most an hour; a refused token stops the watch.
- Over MCP,
events_listreads the recorded events (read-only, no Servosity call).issues watchis CLI-only.
Full command reference: guide.md. For the AI-agent operating contract (--agent, --dry-run, when to confirm before mutating), see AGENTS.md.
What makes this different
Most AI integrations for backup and DR vendors proxy each question into a live API call against the partner portal. That's fine for one client. It fails when you're asking "which of my 47 clients have a backup set that hasn't succeeded in 7 days?" - the partner portal scatters that across per-client pages, per-type views (M365, DR Server, DR Desktop), and an alert queue full of noise.
This skill syncs Servosity into a local fleet mirror with snapshot history and FTS5 search. Cross-client questions become one local query: instant, offline, and the AI sees the answer, not a context window full of paginated JSON. Compound commands like attention, drift, stale-backups, and backup-facts join across companies, backups, and issues - and the snapshot history they persist powers drift, fleet-health deltas, and storage-trend forecasts no stateless API wrapper can offer.
The pain this closes
Backup and DR is where "silent failure" hurts most. The pains MSP owners name:
- Silent backup failures discovered too late. A backup that quietly stopped succeeding is invisible until a client needs a restore - the worst possible moment to find out.
- No fleet-wide view. Each client's backup state lives in its own portal view; there is no single screen that says "across my whole book, here's what's stale, failing, or in-flight right now."
- Alert-queue noise buries the real failure. Dozens of repeat and known-safe issues pile up per client; the one that matters hides in the pile.
- Per-client questions mean portal archaeology. Answering "is this client OK?" means clicking through metadata, every backup type, contracts, and issues by hand.
This skill turns per-client portal views into fleet-wide, offline-fast intelligence: attention is one screen across every client; stale-backups finds the Friday-email list and email-draft writes the emails; drift opens Monday with situational awareness instead of a blank slate; triage batch-suppresses known-safe noise (ignore/archive/comment) so the queue shows only what's new; qbr and qbr-all turn quarter-end report-building into one command.
Frequently asked questions
Does this work with ChatGPT?
Yes, in the ChatGPT desktop app. It runs MCP servers locally (STDIO), the way servosity-mcp runs, and shares MCP configuration with Codex (OpenAI): add servosity-mcp under Settings > MCP servers as a STDIO server, or add it once to Codex's ~/.codex/config.toml. ChatGPT on the web is not supported by msp-skills yet: it needs a remote server. If your team works there, a hosted gateway is the better pick; check that your AI app is on the gateway's setup list. Setup steps: mcp-install.md.
Does this work with Codex, Cursor, Windsurf, Cline, Copilot, or Gemini?
Yes - all of them speak MCP. Cross-tool install commands are in the matrix above and the deep-dive in docs/which-agent.md. Tool-specific gotchas (Copilot uses servers not mcpServers, Cline's Windows npx issue) are documented there.
Do I need to be a Servosity MSP partner?
Yes. The Servosity API surface this skill wraps is the one available to authenticated MSP partners. You'll need a partner API token from the Servosity partner portal. The skill itself is free; the token is part of your Servosity partner agreement. Not a Servosity partner yet? Reach out via servosity.com.
Will this replace my Servosity partner portal?
No. The portal is still where you do administrative work, configure backups, and onboard clients. This skill makes the portal's data answerable by AI - so morning triage, Friday stale-backup reports, drift checks, and ad-hoc cross-client questions don't require portal archaeology. The portal is unchanged; you're adding a new way to ask the data questions.
Do I need to know how to code?
No. The recommended install is to paste one sentence into Claude Code or Codex - your agent reads SKILL.md and does the install. The fallback is a one-line installer per OS (bash or PowerShell). Neither path requires writing code. You'll enter your Servosity partner token once.
Is my Servosity / client data safe?
Your data stays on your machine. The CLI and MCP server are local binaries. The fleet mirror sits in a directory under your user account. The AI agent only sees what the CLI returns - typically a query result, not raw bulk data. The partner token is read from your environment or your agent's config; never bundled into this repo or transmitted anywhere by MSP Skills. The token is scoped to your reseller account only (no cross-reseller access).
Can my agent hear about new Servosity issues?
Yes. Servosity does not send webhooks yet, so servosity-cli issues watch polls your issue list read-only (every 5 minutes by default) and records each new, changed, archived, ignored or reactivated issue as an event. Your agent reads them with servosity-cli events list --since <cursor> or the read-only MCP tool events_list, which costs no Servosity call. Events carry ids and triage fields, never issue text.
Can I run this on Windows?
Yes. The PowerShell installer above is the Windows path. CLI and MCP binaries are native Windows builds.
What does it cost?
Free. Apache-2.0 licensed. Servosity does not charge for the API access required to run this skill - the partner-portal API is part of your existing partner agreement. You pay only for whichever AI agent you use (Claude, ChatGPT, Codex, etc.), billed by your AI provider.
Safety model
The skill authenticates with one partner token, scoped to your reseller account only (no cross-reseller access). Know how writes are gated - the binary does NOT gate them for you: --dry-run is an opt-in preview, and where a command prompts interactively, only an explicit --yes skips the prompt - --agent does not imply --yes. The gate lives in your agent's policy: preview with --dry-run, show the exact command, get approval, then run.
Keep autonomous agents to Read plus previewed writes; gate everything below that behind a human. Full matrix and lock-down guidance in governance.md.
Status
Already used inside Servosity's own backup / DR operations; published here for MSP partners. The public surface is in beta and being validated with MSPs in live Build Sessions. RSVP at compoundingteams.com/build-sessions.
Standards. Conforms to the open Agent Skills spec (Anthropic, Dec 2025; 40+ agents). MCP-compatible - works with any MCP-capable agent including Hermes and OpenClaw, with metadata.openclaw frontmatter pre-wired so openclaw skills install works directly.
Maintained by Servosity. Apache-2.0 licensed. Built with CLI Printing Press. Last updated: 2026-10-23.
来源:skills/servosity/README.md,提交 8e88535
工具
0版本历史
10- v0.6.13最新Nov 6, 2026
- v0.6.12Nov 5, 2026
- v0.6.11Oct 24, 2026
- v0.6.10Oct 23, 2026
- v0.6.9Oct 23, 2026
- v0.6.8Oct 22, 2026
- v0.6.7Oct 16, 2026
- v0.6.6Oct 2, 2026
- v0.6.5Sep 29, 2026
- v0.6.4Sep 16, 2026
评分
已满足适用检查项的 40%
该评分衡量此收录条目的完整程度,依据注册中心条目、端点与公开来源计算得出。它不是对服务器本身的评价,也不表示服务器安全或当前可用:可连通性是带日期的检查结果,评分也不包含安装量或用户评价。
- 近 30 天内可连通25 分
请保持端点能够响应 MCP 握手;SourceWeft 每 30 天探测一次。
- 已声明认证方式10 分
请在注册中心条目中声明凭据请求头,或保持端点能够响应探测,以便判定其认证方式。
- 已列出工具15 分
请让服务器在无凭据时响应 tools/list,或发布可在沙箱中无凭据启动的软件包。
- 描述必需15 分
- 图标10 分
- 文档15 分
- 许可证10 分
请在仓库或注册中心条目中声明许可证。