Limosaidentity

FULL AI SUPPORT

The full platform.
Through your AI assistant.

Configure applications, shape access policies, complete assigned work and investigate outcomes. Limosa exposes its full management interface to compatible AI assistants through the Model Context Protocol (MCP).

Abstract connected modules illustrating AI interaction with identity governance
Concept illustration · AI working across a governed platform

AI ACROSS THE WHOLE WORKFLOW

From initial setup
to everyday operation.

Every registered management operation is discoverable and callable by AI. New registered operations become available through the same interface.

01

Applications & accounts

Discover connectors, configure connections, test access, preview accounts and run aggregation.

02

Identities & mappings

Configure identity attributes, source mappings and correlation; explore connected account inventory.

03

Roles & policies

Manage roles and assignments, configure access workflows and maintain governance settings.

04

Requests & decisions

Submit requests, inspect assigned approval context and carry out authorized decisions.

05

Reviews & fulfillment

Configure certification, complete assigned reviews and follow manual fulfillment work.

06

Tasks, reports & history

Run supported jobs, follow progress, query reports and investigate changes and provisioning outcomes.

AI ACCESS ARCHITECTUREA full interface. The same governance.
YOU + YOUR ASSISTANT

Describe the work

A compatible AI assistant discovers available operations and acts on your instructions.

AUTHENTICATED MCP CONNECTION

Connect with consent

Your tenant and current permissions determine what the assistant can do.

SHARED APPLICATION API

The same operations as the web workspace

ValidationPermissionsTenant isolationAudit attribution
Configure

Connections, mappings, roles and workflows

Operate

Requests, reviews, tasks and fulfillment

Investigate

Inventory, reports and operation history

Conceptual architecture. Bulk work runs through background jobs; approvals, execution results and observed access remain separate.

Full capability, scoped authority

The assistant uses the connected user’s current permissions. Ownership, reviewer eligibility, tenant isolation and configured approval rules still apply.

Connect once, act with context

Sign-in and consent remain interactive. External provider setup and credentials are prerequisites; connecting an assistant does not grant rights in those systems.

Follow execution to its outcome

Starting a task or approving access does not prove provisioning succeeded. Assistants can follow job progress and inspect the resulting evidence.

START WITH THE WORK YOU NEED DONE

A conversation with purpose.

01 / CONFIGURE

From intent to configuration.

“Help me connect our HR source and map worker records to identities.”

Your assistant can discover the connector, gather settings, validate configuration and test the saved connection before importing.

Explore the connectors →
02 / OPERATE

Get governance work done.

“Show my pending access reviews and the context for each decision.”

Your assistant can retrieve your assigned work and submit permitted decisions through the same governed operations as the workspace.

Explore access reviews →
03 / INVESTIGATE

Follow the evidence.

“Find failed provisioning operations and help me understand what happened.”

Your assistant can read task progress and operation histories, then help you distinguish an approval from the actual target outcome.

Explore the architecture →

Illustrative requests, not a transcript or autonomous execution guarantee. Available actions follow the connected user’s permissions and the application’s configured capabilities.

START WITH YOUR AI ASSISTANT

One URL.
Your first workspace.

Add Limosa to your assistant, then sign in when prompted. No existing Limosa account or workspace ID needed.

Use an email magic link, or Google or GitHub when offered. Create or choose your own workspace and approve the connection. Your assistant can then help you configure the platform.

Choose your assistant →

REMOTE MCP SERVER URL

https://identity.limosa.io/mcp

  1. Add the URLUse a remote MCP connection with OAuth.
  2. Sign in & choose a workspaceNew here? Create your workspace during sign-in.
  3. Approve & start a conversationTry: “Help me set up my first application.”

An account is created through sign-in; data access always requires authentication and your explicit approval. Assistant plans and organization policies may restrict custom connections.

CONNECT THE ASSISTANT YOU ALREADY USE

A few steps to your first conversation.

Add the public URL above using OAuth and Streamable HTTP. When Limosa opens, sign in, select or create a workspace, then choose Approve & connect. New users can complete these steps without registering beforehand.

ChatGPT

  1. Open Plugins, select +, then Add custom MCP server. In the legacy interface, use Settings → Apps → Create instead.
  2. Name it Limosa Identity, paste the MCP URL and choose OAuth. Review the connection warning and create the plugin/app.
  3. Complete Limosa sign-in and workspace consent, then enable the connection in your conversation.

Custom MCP availability depends on your account and workspace permissions. Your administrator may need to create or approve the connection.

OpenAI setup guide

Claude

  1. Open Customize → Connectors → + Add → Add custom connector. Organization owners can add a connector under Organization settings → Connectors → Add → Custom → Web.
  2. Name it Limosa Identity and paste the MCP URL. Choose Sign in now and Register automatically for the OAuth client when those choices appear.
  3. Complete Limosa sign-in and workspace consent, then enable the connector for your conversation.

Limosa supports automatic client registration; choose that option rather than Claude’s published-identity option. Organization permissions apply.

Claude setup guide

Microsoft 365 Copilot

  1. In Copilot Studio, open your agent’s Tools → Add a tool → New tool → Model Context Protocol.
  2. Enter Limosa Identity and the MCP URL. Select OAuth 2.0 → Dynamic discovery.
  3. Create the connection, sign in to Limosa, choose your workspace and approve access. Add the tool to the agent and test it.
  4. Publish the agent to Microsoft 365 Copilot through your organization’s supported channel and approval process.

This is an agent setup path, not a URL pasted into an ordinary Copilot chat. Copilot Studio access, licensing and administrator policies apply.

Microsoft MCP setup guide

Client instructions checked November 13, 2026. Menus and availability can change; the linked vendor guides are the source for current client requirements. These are setup instructions, not a claim that every client or plan has been certified.

Already working in an organization’s workspace?

Open API & AI connections → AI assistant (MCP) in that workspace to copy its dedicated URL. It uses that workspace’s sign-in and access policies. The common URL is for creating or connecting workspaces you manage; it does not grant membership in somebody else’s organization.

Start with a useful question.

“Show me my workspace context, then help me configure my first application. Explain the settings I need and test the connection before importing.”

Your assistant can discover supported operations and guide configuration. You provide any external system permissions and credentials that the integration needs. You can revoke access from API & AI connections in your workspace.

An architectural loop of glass and ceramic stations illustrating ideas, AI-assisted building, inspection and iteration
Original AI-generated concept illustration · development and iteration

AI IN HOW WE BUILD

Built with AI.
Improved continuously.

AI is the main driver behind our fast development cycles.

We use AI to turn product ideas into implementation, tests and improvements. Human direction, review and validation remain part of the process.

The result is a product that can evolve quickly around real governance needs. Follow the release journal to see the concrete capabilities each update brings.

See what we’re building →

A FEW PRACTICAL DETAILS

What full AI support means.

Can AI configure the platform as well as read it?

Yes. The management interface supports reads and writes through MCP, including connector configuration, identity mappings, roles, workflows and governance administration. Each action is subject to the same validation and authorization as the web interface.

Does Limosa include its own chat assistant?

The integration connects your compatible AI assistant to Limosa through MCP. Your assistant supplies the conversational experience; Limosa supplies the governed operations and workflow guidance.

Does AI bypass approvals or manage external credentials for me?

No. Your current permissions and configured approval policies remain authoritative. Sign-in, consent and external provider authorization still require the appropriate people and credentials.

How does AI influence product development?

AI is the main driver behind our fast development cycles, helping turn ideas into implementation, tests and improvements. Human direction, review and validation remain part of the process. The release journal documents concrete changes without promising a fixed delivery cadence.

YOUR NEXT STEP

See governance in practice.

Explore the public demo or bring your applications and governance questions to a guided conversation.