Orgonaut × Model Context Protocol

Your AI, connected
to your organisation.

Bring Orgonaut into ChatGPT, Claude, Claude Code, Codex or another compatible MCP client. Ask about your teams. Compare your plans. Keep the conversation grounded in your organisation.

Connect in a few steps Explore sample prompts
Your remote MCP URL
https://app.orgonaut.co/mcp/orgonaut

Available on every Orgonaut plan, including trials.
Sign in. Choose your organisation. Choose your access.

ChatGPT MCP Orgonaut
ChatGPT Orgonaut
Acme · Q3 Restructure · Read only

What changes in Q3 Restructure compared with its source?

Orgonaut · scenario.delta

The scenario shifts capacity towards Product while keeping the same people.

People 24 → 24
Agent placements 2 → 4
  • Two people move from Platform to Product.
  • Two agent placements are added to Customer Operations.
  • The source organisation is unchanged.
Illustrative interaction · fictional organisation

One connection. Your choice of assistant.

ChatGPT Claude Claude Code Claude Code Codex
Connect once. Then just ask.

A URL. A sign-in.
A much better starting point.

MCP gives your assistant tools to work with Orgonaut. You don’t need to run a local server for the remote connection.

Add Orgonaut

Paste the remote MCP URL into your client, or use its command-line setup.

Choose your access

Sign in through OAuth. Select the organisation, then Read only or Full access.

Ask a real question

Start with the structure, a team roster or a scenario comparison. Your assistant can discover the available tools.

ChatGPT

Connect from ChatGPT

  1. Open ChatGPT’s settings for apps or custom MCP connections. Enable developer mode if your setup requires it.
  2. Add a connection named Orgonaut, paste the URL below and use OAuth authentication.
  3. Sign in to Orgonaut, choose the organisation and access level, then approve the connection.
  4. Use Orgonaut in a conversation and try a prompt below.
https://app.orgonaut.co/mcp/orgonaut

Custom MCP availability depends on your ChatGPT plan and workspace policy. Your administrator may need to enable it.

Follow the connection guide
Claude

Add a custom connector in Claude

  1. Open Claude’s connector settings and add a custom connector.
  2. Name it Orgonaut and enter the remote MCP URL.
  3. Connect, sign in to Orgonaut and choose your organisation and access level.
  4. Enable Orgonaut for the conversation and start asking.
https://app.orgonaut.co/mcp/orgonaut

Connector availability and administrative controls depend on your Claude plan and workspace.

Follow the connection guide
Claude Code Claude Code

Connect from your terminal

Add the remote HTTP server to Claude Code:

Claude Code 
claude mcp add --transport http orgonaut https://app.orgonaut.co/mcp/orgonaut

Open Claude Code and run /mcp to authenticate Orgonaut in your browser. Choose your organisation and access level, then return to the terminal.

Orgonaut connection guide
Codex

Add Orgonaut to Codex

Register the remote server, then start its OAuth sign-in:

Codex CLI 
codex mcp add orgonaut --url https://app.orgonaut.co/mcp/orgonaut
codex mcp login orgonaut

Complete sign-in in your browser, choose the organisation and access level, then ask Codex to use Orgonaut.

Orgonaut connection guide

Using another client? Add the same URL with remote HTTP and OAuth support. Tenant-hosted connections are also documented in the setup guide .

Things you can ask

From “who’s on that team?”
to “what would change?”

Copy a prompt, change the example names and let your assistant use the organisation behind the question.

Explore

Understand the organisation

Show our department structure and the teams in each department. Tell me which organisation and context you are using.

Get your bearings without assembling a spreadsheet.

Staffing

See who is working where

Show the Platform team’s roster, roles and allocations. Highlight people working across more than one team.

See staffing in the context of the work.

Scenarios

Compare a proposed change

Compare the Q3 Restructure scenario with its source. Summarise the changes to teams, headcount, FTE and monthly cost where available.

Turn a future structure into a clear decision.

Planning

Plan a move safely

In Q3 Restructure, draft a proposal to move two people from Platform to Product. Show the affected records and impact before asking me to confirm.

Explore a change before it touches the organisation.

AI operating model

Put agents in the model

Show the agents and robots in our organisation, their team placements and allocations. Separate represented data from missing information.

Plan human and machine capacity together.

Projections

Look along the timeline

For Q3 Restructure, compare headcount, FTE, people cost and agent cost with its source on 1 January each year for the next three years. Label future figures as projections.

See the effect of dated changes over time.

Review

Prepare a change summary

Review recent activity in Q3 Restructure and its scenario delta. Draft a short update for the leadership team with open questions and assumptions.

Bring the evidence into your next conversation.

Studio · Enterprise / Agency

Inspect a transformation case

List our Transformation Cases. For the case I choose, explain its readiness findings and source-to-target changes without editing it.

Review the case and its evidence from your assistant.

Cost and velocity answers require the relevant permissions. Future figures are projections from represented, dated changes. Studio prompts require Enterprise or Agency.

Real tools, real context

More than a document
attached to a chat.

Your assistant can discover structured tools, resolve entities and inspect the current model. Here’s a selection of what the MCP surface supports.

Organisation & people

Explore hierarchy, teams, actors, positions and placements in their organisational context.

org.unit.treeteam.listorg.team.rosteractor.profileposition.list

Metrics & projections

Inspect permitted headcount, FTE, cost and velocity data. See how dated scenario changes play out.

org.unit.metricsteam.velocitytransfer.impactscenario.timeline

Scenarios & history

Compare a future organisation with its source and follow the planning history behind it.

scenario.listscenario.deltascenario.historyactivity.recent

Reviewable changes

Draft a supported action in a mutable scenario, inspect the proposal, then explicitly confirm it.

scenario.action.planscenario.action.confirm

Transformation Studio

Read capabilities, governed AI Solutions and Case readiness on Enterprise and Agency. Studio changes stay in the product.

business.capability.detailai.solution.deploymentstransformation.case.readiness

Discovery & guidance

Let your assistant discover the available tools and review workflows instead of guessing what it can do.

capability.listcapability.explainorg.review.guidescenario.review.guide
Full tool catalog, resources & prompt reference
You set the boundaries

Useful access.
Explicit decisions.

Choose how your assistant can work with the organisation. Orgonaut checks both the connection’s access level and the permissions of the signed-in member.

  • Read onlyExplore permitted data, compare scenarios and prepare written analysis. Scenario changes cannot be drafted or confirmed through this connection.
  • Full accessRead permitted data and stage supported actions in mutable scenarios. Applying a proposal still requires explicit confirmation.
  • One deliberate organisationThe connection stays bound to the organisation you select. Switching organisations in the web app does not switch the connection.

Reconnect to choose a different organisation or access level. Revoke access from your connected-app settings. MCP does not write to Live, promote scenarios or edit Transformation Studio records.

Permissions & supported actions
Astro Orgonaut
Acme · Q3 Restructure · Full access

Draft a move for Alex and Sam from Platform to Product. Show me the proposal first.

Orgonaut · scenario.action.plan
Scenario proposal

Move two placements

Platform Product

Alex Chen · 100% allocation
Sam Rivera · 80% allocation

Awaiting your confirmation
Illustrative interaction · fictional organisation
Plan → review → confirm

The proposal comes
before the change.

Ask for an action in a mutable scenario. Orgonaut prepares the supported change as a proposal, so you can inspect it before it is applied.

Plan

scenario.action.plan prepares the action with the relevant records and a proposal ID.

Review

Read the proposed change, resolve ambiguities and decide whether the scenario should change.

Confirm

scenario.action.confirm uses the proposal ID to apply the approved action in the scenario.

For builders

An open protocol.
A toolkit you can build on.

MCP is the conversational tool surface. The REST API and CLI give you direct routes for integrations, scripts and your own agents.

Discover through MCP

After establishing an authenticated MCP session, your client can request the tool catalog and use Orgonaut’s discovery and review guides.

MCP protocol message · tools/list 
{"jsonrpc":"2.0","id":1,"method":"tools/list"}

Resources include baseline summaries and scenario deltas. Prompt templates support organisation and scenario reviews; guide tools serve clients that only expose tools.

MCP developer reference

Call the REST API directly

Use authenticated endpoints when you need predictable HTTP integration. This example retrieves the organisation-unit tree.

REST API · example 
curl 'https://app.orgonaut.co/api/v1/org-units/tree' \
  -H "Authorization: Bearer $ORGONAUT_TOKEN" \
  -H 'X-Tenant: acme' \
  -H 'Accept: application/json'

Set your own token in ORGONAUT_TOKEN and replace acme with your tenant. Endpoint permissions still apply.

REST API reference
Start with a conversation

No organisation yet?
Build a first draft with AI.

The guest modelling workflow can turn a description or pasted data into a temporary organisation. Explore its structure, model alternatives and compare them before deciding whether to save.

Your guest model stays temporary until you choose to claim it. Saving takes you into Orgonaut’s account and trial flow.

Guest modelling tools & save workflow
GUEST MODELLING / SEPARATE ENDPOINT

“Help me shape this organisation.”

“We have Product, Engineering and Customer Operations. Help me model our teams and reporting lines. Ask for missing information rather than inventing it.”

Guest / agent MCP endpoint
https://app.orgonaut.co/mcp/orgonaut-agent

Use this endpoint for the guest workflow. Keep the guest access key private; the temporary model expires unless you save it.

Your next step

The details, when you need them.

Connect MCP clients Canonical setup guide, OAuth and organisation selection. MCP reference Tool catalog, resources, prompts and permission requirements. CLI guide A second route for terminals, scripts and coding agents. REST API reference Authenticated HTTP endpoints for your own integrations.

Good questions.

What is MCP?

Model Context Protocol is a standard way for AI clients to discover and use external tools and context. Orgonaut’s remote MCP server exposes organisational tools to compatible assistants through an authenticated connection.

Which plans include it?

Every Orgonaut plan includes MCP, including trials. Transformation Studio’s read-only tools require Enterprise or Agency. Your AI provider controls availability of custom connections in its own plans and workspaces.

Can my assistant see data I cannot?

No. Orgonaut enforces the signed-in member’s permissions and connection scopes. Cost, pay and velocity data require their relevant permissions; restricted data does not become available simply because you use an AI client.

What happens if I switch organisations in the web app?

Your MCP connection remains bound to the organisation selected during sign-in. Reconnect deliberately to choose another organisation. Ask the assistant to state the organisation and scenario before analysing or proposing changes.

Does a proposal automatically change my organisation?

No. A supported MCP action is staged in a mutable scenario and requires explicit confirmation to apply. Live changes and scenario promotion are handled through the product workflow.

Can I use the CLI instead?

Yes. The Orgonaut CLI supports terminal workflows, JSON output and a companion skill for Claude Code and Codex. Use the tool that fits how you work.

Your assistant. Your organisation.

Make your next question
a connected one.

Start with the model. Let your assistant help you explore it.

Connect Orgonaut Explore all AI-native tools