Microsoft Graph MCP

Microsoft Graph + AI - for Claude, the ChatGPT desktop app, GitHub Copilot, Gemini CLI, and any agent that speaks MCP

Unofficial. Community-built Claude Code Skill and MCP server for the Microsoft Graph API. Not affiliated with, endorsed by, or sponsored by Microsoft Corporation.

[Microsoft Graph demo - animated preview]

_{▶ Watch the 30-second demo - demo data is simulated; every command shown exists in the real CLI.}

The maintained single-binary successor to the retiring mgc - every MSP-relevant Microsoft Graph surface, plus an offline store that finds wasted licenses, privileged-access risks, and stale devices no single API call can. Works with the AI you already use - ChatGPT (desktop app), Claude Desktop, Codex, Claude Code, Claude Cowork, and GitHub Copilot - plus Gemini CLI. Free, open source, runs on your laptop. Built for MSP owners. No code required.

Works with your agent

The six agents MSP owners actually use (self-serve, works today):

The ChatGPT desktop app runs microsoft-graph-mcp locally and shares MCP configuration with Codex (OpenAI). ChatGPT on the web needs a remote (HTTP) server, which msp-skills does not support yet (see mcp-install.md). 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 for the Microsoft and Google stacks

Big install base, but an honest heads-up: Gemini CLI runs the local binary you just installed, while Microsoft 365 Copilot and the Gemini app need a remote server, which msp-skills does not support yet.

Skill-native agents (also covered): Hermes and OpenClaw read this skill's SKILL.md directly and speak MCP - see their install sections below. Also works with Cursor, Windsurf, Cline, Continue.dev, and Zed via MCP. 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 .mcpb launches on macOS (Intel and Apple Silicon), Windows x64 and Linux x64. Every tool shells out to the companion microsoft-graph-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 set MICROSOFT_GRAPH_CLI_PATH to an existing binary). Details: #331.

Download Microsoft Graph MCP (.mcpb) - then open Claude Desktop > Settings > Extensions and select the file. One click, no JSON, no shell. (Browse every Microsoft Graph release on the releases page.)

Prefer the Claude Code plugin? Add the marketplace once, then install - works immediately, no directory listing required:

/plugin marketplace add Servosity/msp-skills/plugin install microsoft-graph@msp-skills

Path A - paste one prompt into your AI agent (recommended)

Copy this into Claude Code, Codex CLI, or Claude Cowork:

Install the Microsoft Graph 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/microsoft-graph/install.sh). If it uses Windows PowerShell, run iwr -useb https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/microsoft-graph/install.ps1 | iex. Then authenticate per the README and run microsoft-graph-cli --help to explore.

The same prompt works in any agent that can run shell.

Path B - run the installer yourself

Windows (PowerShell):

iwr -useb https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/microsoft-graph/install.ps1 | iex

macOS / Linux:

bash <(curl -fsSL https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/microsoft-graph/install.sh)

The installer drops both microsoft-graph-cli (the CLI) and microsoft-graph-mcp (the MCP server) into your user bin path. Claude Code, Codex, and Cowork discover the Skill via SKILL.md in this directory.

Verify:

microsoft-graph-cli --version

Upgrade to the latest version

The installer always fetches the current release - re-run it to upgrade:

macOS / Linux:

bash <(curl -fsSL https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/microsoft-graph/install.sh)

Windows (PowerShell):

iwr -useb https://raw.githubusercontent.com/Servosity/msp-skills/main/skills/microsoft-graph/install.ps1 | iex

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 microsoft-graph@msp-skills.

Add to Claude Desktop, the ChatGPT desktop app, GitHub Copilot, Gemini CLI, or another MCP client

After the installer runs, see mcp-install.md and docs/which-agent.md for the per-agent wire-up - one section per agent, including the GitHub Copilot servers key and the ChatGPT desktop app. Claude Desktop's Settings > Extensions panel is the simplest path; the MCP config block (for users who prefer editing JSON) is documented in mcp-install.md.

Install for Hermes

From the Hermes CLI:

hermes skills install servosity/msp-skills/skills/microsoft-graph --force

Inside a Hermes chat session:

/skills install servosity/msp-skills/skills/microsoft-graph --force

Hermes speaks MCP natively, so it can also use the microsoft-graph-mcp server directly - same install path, same env vars.

Install for OpenClaw

Tell your OpenClaw agent (copy this):

Install the microsoft-graph skill from https://github.com/servosity/msp-skills/tree/main/skills/microsoft-graph. The skill defines how its required CLI (microsoft-graph-cli) can be installed via the openclaw: frontmatter block.

OpenClaw isn't generally available yet; the frontmatter wiring is pre-shipped and will activate the moment OpenClaw launches.

Authenticate

Set the credentials the CLI needs (from your Microsoft Graph portal):

MICROSOFT_GRAPH_TOKEN=<value> microsoft-graph-cli doctor

doctor confirms the credentials work before you run anything that touches data. It exits 0 even when the credential is rejected, so scripts should add --fail-on error.

What this skill does

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 Microsoft Graph integrations and MCP servers proxy each question into a live API call. That's fine for one record. It dies at scale, when you're asking "how much license spend can we reclaim across the tenant" or "who can administer this tenant right now" - and Graph throttles the bulk reads those questions need.

This skill syncs the MSP-relevant Microsoft Graph surface 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 licenses waste, admins audit, and tenant snapshot join across users, subscribed SKUs, directory roles, security alerts, and Intune devices - work a stateless API wrapper can't do. And it does it as one cross-platform binary with no .NET or PowerShell runtime - the lightweight successor to the mgc CLI Microsoft retires on August 28, 2026.

The pain this closes

Microsoft retires the Microsoft Graph CLI (mgc) on August 28, 2026 - deprecated since September 2025, security fixes only - and steers everyone to the heavier PowerShell SDK (Microsoft 365 Developer Blog, "Microsoft Graph CLI retirement"). MSPs who scripted tenant reporting on a lightweight cross-platform binary now face a .NET/PowerShell dependency on every machine that runs it. And mgc never answered the cross-entity questions an MSP actually asks - recoverable license spend, who holds admin, which devices are drifting - because no single Graph endpoint returns them; each one becomes a CSV export and a spreadsheet join, per client, every month.

This skill is the lightweight successor that also closes that gap:

See pain-point.md for the longer narrative.

Frequently asked questions

Does this work with ChatGPT?

Yes, in the ChatGPT desktop app. It runs MCP servers locally (STDIO), the way microsoft-graph-mcp runs, and shares MCP configuration with Codex (OpenAI): add microsoft-graph-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, which msp-skills does not support yet; see mcp-install.md. 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.

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 Microsoft Graph credentials once.

Is my Microsoft Graph 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.

Is this the replacement for the Microsoft Graph CLI (mgc) that's being retired?

It's built as the lightweight successor for the MSP read-and-report core - directory, licensing, security, and device surfaces - as one cross-platform Go binary with no .NET or PowerShell runtime. Microsoft's own recommended path is the PowerShell SDK; this is the option for teams who want a scriptable single binary and their AI agent instead. It is not affiliated with or endorsed by Microsoft.

Will this hit my Microsoft Graph throttling limits?

The local SQLite mirror exists so reads stop hitting Graph. After the first pull, the cross-entity views (licenses waste/orphans/map, admins audit, security triage, managed-devices drift, groups risk, tenant snapshot) run against local SQLite with zero API calls. Live calls follow @odata.nextLink and respect a --rate-limit throttle, and pull treats resources your token can't reach as warnings, not failures.

Does it use a delegated or app-only token?

Either. Run microsoft-graph-cli auth login --tenant <id> --client-id <id> --client-secret <secret> to mint and cache an app-only (client-credentials) token for unattended MSP use, or export a pre-minted token as MICROSOFT_GRAPH_TOKEN. Read scopes such as Directory.Read.All, RoleManagement.Read.Directory, SecurityAlert.Read.All, and DeviceManagementManagedDevices.Read.All must be granted and admin-consented. App-only tokens have no /me, so users me is delegated-only.

How do I audit which third-party apps have access to a tenant?

Run microsoft-graph-cli apps consent --agent. It inventories every non-Microsoft enterprise application (service principal) consented into the tenant and the access it holds - joining service principals, delegated consent grants (oauth2PermissionGrants), and application/app-only permissions (appRoleAssignments) into one risk-ranked table. It flags over-privileged apps (broad or write-all scopes), admin-consented (tenant-wide) apps, privilege-escalation permissions, and user-consented shadow IT - all read-only. Microsoft's own first-party apps are counted but not listed. Needs read scopes Application.Read.All, Directory.Read.All, and DelegatedPermissionGrant.Read.All.

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

No command writes to your tenant. The strongest control is the scope you grant the Microsoft Graph credentials - the CLI can only read what the credentials are permitted to read. Full details, including how to lock it down, are in governance.md.

Status

Beta. Validated against the Microsoft Graph API surface and being validated with MSPs running it live against their own production tenant 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. OpenClaw-ready (frontmatter pre-wired, awaiting OpenClaw launch).

Maintained by Servosity. Apache-2.0 licensed. Built with CLI Printing Press. Last updated: 2026-10-28.