Claude skills for MSPs
422 Claude skills across 14 platforms — every one named, documented, and traced when it runs. Hosted, so there is nothing to install and no API keys to hand your techs. Read-only by default, and on most platforms read-only by construction.
The board
Every platform, every skill counted.
A skill you can't inspect is a claim, not a capability. Each platform below links to its full skill list — what each one reads, in plain language, and what the integration can never do.
How these run
Claude, an MCP server, and a paper trail.
Each platform on the board is an MCP server MSPStuff hosts — the standard way to hand an AI model real tools. Its Claude skills teach the model what that platform's data means to an MSP: what a stale agent looks like, which patch states matter, where the audit trail lives.
When you ask a question, Switchboard runs Anthropic's Claude against those servers and answers with figures traced to the actual tool calls behind them — you can audit every number back to the API response it came from.
Credentials stay yours, scoped by you. Nothing installs on an endpoint, and writes — on the few platforms that allow any — are separate, explicit grants. The rest are read-only by construction.
Common questions
Asked before you had to ask.
What are Claude skills?
Skills are packaged instructions and tools that teach Anthropic's Claude how to do a specific job well. MSPStuff builds them for the platforms an MSP actually runs — 422 skills across 14 platforms so far — each one documented on its platform's page, down to what it reads and what it can never touch.
Do I need Claude Code or an Anthropic account to use these?
No. That's the difference between this page and a skills directory: nothing here is a download. The skills run inside Switchboard, MSPStuff's hosted product — you connect a platform once with credentials you control, and sessions answer in plain English from any browser.
Is there an MCP server for my platform?
Every live platform on the board is an MCP server MSPStuff hosts and operates. That's the architecture: Claude works each question through the platform's MCP server, and every figure in an answer is traced to the real tool call that produced it.
Can I see the full skill list before connecting anything?
Yes — that's the point of the board below. Every platform page lists each skill by name with a one-line description of what it reads, plus an equally plain statement of what the integration cannot do.
Are these official Anthropic products?
No. MSPStuff builds and operates these skills and MCP servers independently; Claude is Anthropic's model, which Switchboard runs under the hood. Anthropic hasn't endorsed or reviewed them.
Patch your stack in.
Start the Switchboard free trial against your own environment — read-only, traced answers, live in an afternoon. The apps are on the shelf when you want them.