Drive every Valprovia function
programmatically
REST API, CLI, MCP server and a Microsoft 365 Copilot agent - programmatic access to every governance function. Not a separate module. The way every module is exposed.
What Valprovia Headless is
Programmatic access to every governance function
Valprovia Headless exposes all four Valprovia modules through a REST API, a CLI, an MCP server and a Microsoft 365 Copilot agent - inside your own Azure tenant. One MCP server across all four modules - not a separate add-on, but the way every module is exposed. So governance can run in scripts, pipelines and AI agents.
Trusted by
The agent already wired into Teams and Microsoft 365 Copilot
End users describe what they need in plain language. The agent resolves the right template, owner and policy, then asks for confirmation before acting - so self-service stays inside the guardrails.
Drop Valprovia into Claude, ChatGPT Enterprise and your in-house agents
Every API operation exposed as Model Context Protocol tools. Token-bound per agent, scope-checked per call, recorded in the same audit log - so agents can act without becoming a back door.
Wire governance into the systems your agents and apps already trust
Token-authenticated endpoints for every workspace, permission, template and lifecycle operation - and the same audit log as the UI.
Scriptable governance your terminal, runbooks and CI can drive Coming soon
Native binaries for macOS, Linux and Windows with pipe-friendly output for jq, yq and standard shell tooling. A service-account login lets CI runners write to the same audit log. The CLI is coming soon - the REST API, MCP server and Copilot agent are available today.
Built for agents and CI from day one - not a UI bolt-on
Most governance tools were built for the admin console. Valprovia is built for both - REST, CLI, MCP and the Copilot agent are the same product as the UI.
| Capability | Valprovia Headless | Other governance tools |
|---|---|---|
| Full REST API across every workspace, permission, template and lifecycle operation | Read-only / partial | |
| Native CLI binaries (macOS, Linux, Windows) with pipe-friendly output | Coming soon | |
| MCP server - every API operation exposed as an agent tool | ||
| Microsoft 365 Copilot agent for end-user self-service | ||
| Token-bound scope per agent and per call | ||
| Included in every license tier - not a paid add-on |
Valprovia Headless is the same product as the UI - exposed across REST, CLI, MCP and Copilot. Agents, scripts and end users act inside the same guardrails, against the same audit log.
Frequently asked questions about Valprovia Headless
Answers on the MCP server, supported LLMs, hosting and licensing.
-
What is an MCP server for Microsoft 365?
An MCP server (Model Context Protocol) exposes operations as standardised tools that an LLM or agent can call. The Valprovia MCP server makes every API operation across all four Valprovia modules available as an agent tool - token-bound per agent, scope-checked per call, recorded in the same audit log as the UI. This lets agents run Microsoft 365 governance actions without becoming a back door.
-
Which LLMs are supported?
Claude, ChatGPT/OpenAI, Microsoft Copilot and your own (custom) agents. The MCP server follows the open Model Context Protocol standard, so any MCP-capable client can plug Valprovia in; on top of that, a native Microsoft 365 Copilot agent is already wired into Teams.
-
Is MCP access self-hosted?
Yes. The MCP server runs self-hosted inside your own Azure tenant, just like the rest of Valprovia. Your data never leaves your tenant, and every agent call is gated by existing permissions and captured in the same audit log.
-
What does Headless cost?
Headless is included in every license tier - not a separate module and not a paid add-on. The REST API, CLI, MCP server and the Copilot agent are part of the same product as the UI.
Make Microsoft 365 governance
agent-ready
Valprovia Headless is part of every license tier. Talk to us about your agentic roadmap and integration with the stacks you already run.