
HaloPSA MCP
io.github.Servosity由 GitHub 账号 Servosity 发布v0.2.22更新于 Nov 6, 2026
HaloPSA ticket triage, SLA-breach pre-emption, and cross-client analytics for any MCP agent.
概览
非官方本地 MCP 服务器,让助手基于本地 SQLite 镜像对 HaloPSA 工单做分诊、提前预警 SLA 违约并进行跨客户分析。
- 可以做什么
- 分诊与 SLA:查看指定时间窗内即将违约的工单、按团队查看调度视图、查看工程师工作负载。
- 客户与合同:生成单屏客户卡片、跟踪合同工时消耗、找出长期滞留的工单。
- 工单与资产:列出某时间点之后变更的工单,查看某台设备的历史记录。
- 知识库:为某张工单推荐应关联的知识库文章。
- 本地镜像:将 HaloPSA 数据全量同步到带全文检索的本地 SQLite 镜像,之后离线执行跨客户聚合查询。
- 适用场景
- 适合托管服务商和支持团队,希望助手基于 HaloPSA、HaloITSM 或 HaloCRM 数据回答跨客户问题,例如 SLA 风险、队列负载和合同消耗,同时避免频繁调用实时 API 触发限流。
- 运行要求
- 以本地进程(stdio)方式运行,可通过 .mcpb 包或安装脚本安装,工具会调用配套的 CLI 可执行文件。需要 HaloPSA 租户和 OAuth2 client_credentials 类型的 API 应用,通过 HALOPSA_TENANT、HALOPSA_CLIENT_ID、HALOPSA_CLIENT_SECRET 和 HALOPSA_OAUTH_SCOPE 提供,可选 HALOPSA_BASE_URL、HALOPSA_DOMAIN、HALOPSA_SCOPE、HALOPSA_TOKEN 和 HALOPSA_TOKEN_URL。多代理安装命令需要 Node.js。
安装
在 SourceWeft 中
- 打开 控制台中的 HaloPSA 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
HaloPSA + AI - for Claude, ChatGPT, Codex, Cursor, and any agent that speaks MCP
Unofficial. Community-built Claude Code Skill and MCP server for the HaloPSA, HaloITSM, and HaloCRM APIs. Not affiliated with, endorsed by, or sponsored by Halo Service Solutions Ltd. HaloPSA, HaloITSM, and HaloCRM are trademarks of Halo Service Solutions Ltd.
[HaloPSA demo - animated preview]
▶ Watch the 30-second demo - demo data is simulated; every command shown exists in the real CLI.
Add HaloPSA ticket triage, SLA-breach pre-emption, 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 SQLite mirror means your AI can answer cross-client questions the live Halo API can't return in one shot - no rate-limit hits during QBR prep. Built for MSP owners. No code required.
Works with your agent
Security note (until this connector is reprinted). Do not run HTTP mode (
--transport http), not even on localhost. This build has no password, listens on every network interface, and predates the browser-rebinding guard, so a web page you open can reach it. The MCP tools can also make changes with your credentials. Use the default stdio transport. If you set up a tunnel for a remote agent, shut it down. A fix is on the way.
The five agents MSP owners actually use:
The ChatGPT desktop app runs halopsa-mcp locally and shares MCP configuration with Codex (OpenAI). ChatGPT on the web needs a remote (HTTP) server, and this connector's HTTP mode is paused until it is reprinted (see the security note in mcp-install.md), so msp-skills does not support it 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.
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 companionhalopsa-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 setHALOPSA_CLI_PATHto an existing binary). Details: #331.
Download HaloPSA MCP (.mcpb) - then open Claude Desktop > Settings > Extensions and select the file. One click, no JSON, no shell. (Browse every HaloPSA 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 HaloPSA 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/halopsa/install.sh). If it uses Windows PowerShell, runiwr -useb https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/halopsa/install.ps1 | iex. Then authenticate withhalopsa-cli auth login(Configuration > Integrations > Halo PSA API in your tenant gives you the credentials) and runhalopsa-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 halopsa-cli (the CLI) and halopsa-mcp (the MCP server) into your user bin path. Claude Code, Codex, and Cowork discover the Skill via SKILL.md in this directory; the binary is what the Skill actually drives.
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 halopsa@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 halopsa-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 halopsa skill from https://github.com/Servosity/msp-skills/tree/main/skills/halopsa. The
metadata.openclawfrontmatter declares the required CLI (halopsa-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 halopsa-mcp as an MCP server alongside the skill.
Authenticate
HaloPSA uses OAuth2 client_credentials. Create an API application in your tenant under Configuration → Integrations → Halo PSA API (Authentication Method: Client ID and Secret, Services), then:
The CLI caches the access token and auto-refreshes before expiry.
What this skill does
Outcome-first, with the single command that answers each question:
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 HaloPSA MCP servers and AI integrations proxy each question into a live API call. That's fine for one ticket. It dies at QBR time, when you're asking "how many backup-failure tickets across all 47 clients last quarter, grouped by contract type?" - that's 47 paginated REST calls, HaloPSA rate-limit headaches (limits aren't publicly documented; they vary by hosting), and a context window full of raw JSON the model has to re-read.
This skill syncs HaloPSA into a local SQLite mirror with full-text search. Aggregate questions become one local SQL join: instant, offline, and the AI sees the answer, not the raw data. Compound commands like triage, client card, and contracts burn join across tickets, clients, contracts, time, and assets - work a stateless API wrapper can't do.
It complements HaloPSA's built-in ChatGPT integration (which is great for per-ticket work like rewriting replies and sentiment-flagging). The two are non-overlapping.
The pain this closes
Two pains MSP owners name repeatedly:
- Cost-to-serve is rising faster than contract value. The squeeze is operational: queue overload, SLA fires, unpriced work eating margin on every ticket.
- Key-person dependency. "If your best technician left tomorrow, would your clients notice within 30 days?" For most MSPs the answer is yes, because situational awareness lives in one person's head - and in a PSA nobody else triages quickly.
This skill puts the PSA's cross-entity truth one sentence away from any team member or their agent: triage / sla breaching surface what will breach before it does; client card gives any teammate one situational-awareness panel on demand; contracts burn catches unpriced work early; agent workload makes the queue legible across the whole team. The knowledge of how to read the PSA moves from a person into a Skill any teammate (or their agent) can run.
Frequently asked questions
Does this work with ChatGPT?
Yes, in the ChatGPT desktop app. It runs MCP servers locally (STDIO), the way halopsa-mcp runs, and shares MCP configuration with Codex (OpenAI): add halopsa-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 (HTTP) server, and this connector's HTTP mode is paused until it is reprinted; the security note in mcp-install.md explains why. If your team works there, a hosted gateway is the better pick; check that your AI app is on the gateway's setup list.
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 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 HaloPSA API credentials once.
Is my HaloPSA data safe?
Your data stays on your machine. The CLI and MCP server are local binaries. The SQLite 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. Credentials are read from your environment or your agent's config; never bundled into this repo or transmitted anywhere by MSP Skills.
Will this hit my HaloPSA API rate limits?
Almost never. HaloPSA's rate limits aren't publicly documented and vary between cloud-hosted and self-hosted instances. This skill syncs once with sync --full, then incrementally; subsequent triage, SLA, client-card, and cross-client analytics queries run against the local mirror, not the live API. The big-batch queries that get you 429'd with API-passthrough tools become one SQL join here.
How is this different from HaloPSA's built-in ChatGPT integration?
HaloPSA's built-in ChatGPT integration is great for single-ticket work - rewriting replies, summarizing one ticket, sentiment-flagging. MSP Skills is the cross-client analytics + MSP-owner-on-the-couch layer: questions across thousands of tickets, multi-system queries that join HaloPSA with Servosity, ad-hoc reports HaloPSA's UI doesn't surface. The two complement; you don't pick one.
Can I run this on Windows?
Yes. The PowerShell installer above is the Windows path. CLI and MCP binaries are native Windows builds. Cline users on Windows may need a small npx workaround documented in docs/which-agent.md.
What does it cost?
Free. Apache-2.0 licensed. You pay only for whichever AI agent you use (Claude, ChatGPT, Codex, etc.), and that's billed by your AI provider, not by us.
Safety model
The skill authenticates to your own Halo tenant with an OAuth application you create and scope. It defaults to discovery and dry-run before any mutation.
The strongest control is the scope you grant the Halo OAuth application - the CLI can only do what that application is permitted to do. Full details, including how to lock it down, are in governance.md.
Status
Live-verified. Confirmed by Servosity (maintainer) against their own production tenant. More MSPs run it live in our weekly 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 for the MSP community. Apache-2.0 licensed. Built with CLI Printing Press. See TRADEMARKS.md for vendor non-affiliation. Last updated: 2026-05-28.
来源:skills/halopsa/README.md,提交 061ec05
工具
0版本历史
7- v0.2.22最新Oct 23, 2026
- v0.2.21Oct 23, 2026
- v0.2.20Oct 22, 2026
- v0.2.19Oct 14, 2026
- v0.2.18Oct 2, 2026
- v0.2.17Sep 29, 2026
- v0.2.16Sep 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。